#avr | Logs for 2016-09-24

Back
[00:29:29] <Levitator> Excuse me, homies, but what does the notation UCSZn[2:0] mean?
[00:30:00] <Levitator> I guess the n is there in case the bit position is different between different instances of a device or port. But then what are 2 and 0?
[00:35:29] <Casper> Levitator: register UCSZ number whatever, bit 2 to 0
[00:35:40] <Casper> it mean there is more than one of those device
[00:35:46] <Casper> like uart 0 and 1
[00:39:20] <Levitator> Ahhhh.
[00:39:44] <Levitator> That should have been obvious. It didn't occur to me that the range would be MSB-first.
[00:40:51] <Levitator> There's one problem with that, though. That's the frame format register for the USART, and those are the data-length bits. USART handles up to 9-bit data words, but three bits only allows you to count to 7.
[00:41:24] <Casper> the uart is 'main' register is 8 bits wide
[00:41:34] <Casper> the 9th is stored in another one...
[00:41:46] <Casper> it's an hack really
[00:42:00] <Levitator> Yeah, but how are you going to specify 8-9 bit widths when 3-bits can only count to seven?
[00:42:12] <Levitator> That register configures the data width.
[00:42:25] <theBear> Casper, yeah, a hack that sounds "cool" until you implement it and get all disgusted
[00:42:32] <Casper> yup
[00:42:37] <Casper> but no other way...
[00:43:06] <theBear> fortunately needing more than 8 (useful) bits in serial land isn't real common
[00:43:32] <Levitator> All I want is 8, but if you only have three bits to specify the width, then you can only specify 7, so how does that work?
[00:43:34] <Casper> and fortunatelly you can use some function to set the 9th bit and don't have to do it every time
[00:43:36] <theBear> and to be fair, dealing with >8bit values was probly gonna be a bit messy however you got it done
[00:43:57] <theBear> Levitator, maybe you making the mistake of thinking the bits don't have "special meanings" in this case
[00:44:20] <Levitator> I have searched the manual like five times now. It really should specify, if that's the case.
[00:44:44] <Levitator> It says valid sizes are 0-9, and then it tells you that three bits configure that. That makes no sense.
[00:45:16] <theBear> 0-9 ? you SURE ?
[00:45:44] <theBear> without looking, i'd confidently say maybe 7-9 is the range
[00:46:28] <Levitator> Oh. I see, it is saying that n, for the purpose of a specific diagram is 0-8.
[00:46:52] <Levitator> Which is kind of dumb because n can be 9, so it's not a complete example.
[00:47:04] <Levitator> Valid widths are actually 5,6,7,8, or 9.
[00:48:14] <theBear> hmmm, 5 eh ? there sure are some wacky ideas out there :) at least you see
[00:48:51] <Levitator> Ok, so assuming the three bits form an index into that list, then it makes sense, and is consistent with examples that specify binary 11.
[00:49:07] <Levitator> (11 for 8-bits)
[00:49:22] <Levitator> Incidentally, who would ever use 9 bits? That's the first time I've heard of that.
[00:52:52] <theBear> i used to know an example, but it ain't coming to me today .... but for example, if you had some situation where you needed values of 1-512, it would obviously make more sense to use 9 bits than 1 1/8th bytes for each value
[00:53:50] <Levitator> More like 0-511, but that's sort of a tautology. 9 bits in case you need 9 bits.
[00:54:01] <theBear> or you might be implementing your own parity/ec of some kind (hopefully "better" than simple parity, else a waste of effort and bits) and use 1 or several extra bits for that
[00:54:48] <Levitator> Ok. Was just wondering if there was some sort of common use case I wasn't aware of, like some strange IO convention.
[00:54:50] <theBear> Levitator, i decided that i preferred to have a 512 than a 0 <grin> and it might look like a turtle and quack like a turtle, but it's hard to argue with :)
[00:55:42] <theBear> nah, everything outside of 7 and 8 with 1 stop bits and no parity is basically unheard of (even back into historical times,) and even 7 data bits is pretty rare
[00:56:08] <Levitator> 7 makes sense if you know your data is pure ASCII.
[00:56:18] <theBear> from memory the 90s??????? (first avr series/tech) only had 7 and 8 as options
[00:56:38] <Levitator> I cannot remember the last time I ever saw parity used anywhere.
[00:58:02] <theBear> re: pure ascii -> as it often was with modems and printers and stuff to do with english language (encapsulations/data-tunnels like ppp and slip were uncommon until home-net-access became common, so raw ascii was flying all over the place for a while there)
[00:58:03] <Levitator> Oh, yeah, I see how 9N1 makes sense if you want to implement parity in software.
[00:59:44] <Levitator> You can configure 9N1 and it is compatible with 8(O/E)1 as long as you implement parity on your end.
[01:05:00] <theBear> yeah, that kinda thing... you're all over it
[01:06:46] <theBear> ooh, and a personal example cos i kinda proud of remembering one, in the past i've used 8bits of data (tho the number is kinda irrelevant) to move "pairs" of 4 bit/bcd style values... that could easily be a bit more or less that you wanted to squeeze into every byte/frame (they called frames in serial land ? i forgets)
[02:01:27] <Levitator> Ok, so just to be clear, in UCSZn0, the n is the device ordinal?
[02:01:46] <Levitator> So, these constants are defined in case different devices have the same bit in different places?
[03:21:02] <sabor> Levitator: exactly
[03:36:06] <Levitator> Blah. I wonder if that's just a precaution, or if it's necessary sometimes.
[03:37:37] <Levitator> Time for the slepes.
[05:04:03] <theBear> it's necessary sometimes, and also make adding any abstraction/function stuff later likely to be more transparent, also in older chips whole register names/approaches are sometimes different
[05:06:04] <Lambda_Aurigae> too bad he isn't here to see that theBear
[05:10:23] <theBear> meh, i'm ambivolent
[05:10:42] <Lambda_Aurigae> and I'm just a bastard.
[05:10:47] <Lambda_Aurigae> :}
[05:11:08] <theBear> don't be so modest, you're a COMPLETE bastard.... everyone says so <grin>
[05:11:16] <Lambda_Aurigae> thank you!
[05:11:24] <Lambda_Aurigae> even our general manager at work!
[05:11:37] <theBear> we don't share one ! :)
[05:11:59] <Lambda_Aurigae> a couple of months ago the internal IT group wanted me to do some work for them in our office as they don't have any staff on this side of the state and didn't want to send someone over.
[05:12:32] <Lambda_Aurigae> our office general manager told me I wasn't to do it...that I was not an internal IT tech and my job was taking care of the customers.
[05:12:50] <Lambda_Aurigae> last week he asked me to help him get his email on his new phone.
[05:13:07] <Lambda_Aurigae> I told him I wasn't internal IT and that he needed to call compusuite(the internal IT group)
[05:13:33] <theBear> lol
[05:13:46] <Lambda_Aurigae> my supervisor backed me 100% on it too.
[05:14:04] <theBear> that's an impressive grade of oblivious hypocrit style crazy
[05:14:41] <Lambda_Aurigae> my job has changed a lot since I started too...
[05:15:04] <Lambda_Aurigae> when I first started, I was trying to help the internal IT guy(didn't have the big group yet) setup the new server.
[05:15:14] <Lambda_Aurigae> my boss told me I was hired to fix copiers, not computers.
[05:15:31] <Lambda_Aurigae> now 90% of my job is fixing computers...mainly making computers and copiers talk to each other.
[05:16:18] <Lambda_Aurigae> now that we are part of Xerox, my job is very computer oriented. I hardly ever touch a screwdriver these days.
[05:16:24] <theBear> lol, sounds like a dream ! i told bosses many similar things over the years, but you think they ever said "good point ! you go solder and i'll get a minimum wage retard to manually transfer 200 contacts onto my new phone for me !"
[05:16:39] <theBear> i miss my fiery alpha rip
[05:16:52] <Lambda_Aurigae> I HATE Fiery
[05:17:03] <Lambda_Aurigae> when they work, they work well.
[05:17:13] <Lambda_Aurigae> when they start having issues, they are a pain in the ass to fix.
[05:17:24] <Lambda_Aurigae> and expensive too if it is a hardware issue.
[05:17:32] <Lambda_Aurigae> even though 80% of the box is commodity hardware.
[05:18:28] <Lambda_Aurigae> xerox has the freeflow server too...like a fiery only different.
[05:18:53] <Lambda_Aurigae> it is 90% commodity hardware...being a dell workstation with a custom video card in it.
[05:19:09] <Lambda_Aurigae> running either solaris or windows.
[05:19:41] <theBear> i never liked it as a fiery thing, hot damn it was a finely packaged and finished alpha standard/ref. implementation + trimmings when someone eventually got fed up trying to give back a rip he kept being asked for over and over yet couldn't get anyone to accept/take away and naturally turned to me
[05:22:25] <Lambda_Aurigae> hmmm..might have to jump down to harbor freight this weekend.
[05:22:38] <Lambda_Aurigae> they have a cheap rotary tool for 7 dollars.
[05:22:49] <Lambda_Aurigae> would make a good head for my milling machine build.
[05:25:34] <theBear> hmmm, what in some circles might be called a "die grinder" ? or perhaps a "vinyl cutter" ? (two very different potentially cheap rotary tools btw)
[05:26:17] <Lambda_Aurigae> well,,,not really..
[05:26:22] <Lambda_Aurigae> more like a dremel
[05:26:51] <Lambda_Aurigae> http://www.harborfreight.com/Rotary-Tool-Kit-80-Pc-69451.html?ccdenc=eyJjb2RlIjoiMjcxMzQ0MTkiLCJza3UiOiI2OTQ1MSIsImlzIjoiNi45OSIsInByb2R1Y3RfaWQi%0D%0AOiI4OTc3In0%3D%0D%0A&utm_medium=email&utm_campaign=3816b&utm_source=1021&cid=mEmail_s1021_c3816b
[05:27:14] <Lambda_Aurigae> die grinder is much heavier...meant for serious metal cutting.
[05:28:03] <Lambda_Aurigae> uhuo
[05:28:05] <Lambda_Aurigae> I'm in trouble now.
[05:28:12] <Lambda_Aurigae> cat just came running out of the bedroom
[05:28:20] <Lambda_Aurigae> with wife following(walking)
[05:30:49] <theBear> ahh tiny one... and i found in modern times die grinder is more about being dirt-cheap and fast, tho i suppose that is stilll much heavier :)
[05:31:35] <theBear> and of course a vinyl cutter is what a lotta industries call a hopelessly lightweight often largely plastic construction version of a (plunge)router
[05:31:58] <Lambda_Aurigae> to me a vinyl cutter is a plotter with a knife blade on it.
[05:32:24] <theBear> if you remember actually, i'd be rather interested what you find in the thing you linked between the chuck/out shaft and the motor shaft when you seen it
[05:32:48] <theBear> lol, that's EXACTLY what i said when i was first told what i just repeated bout "vinyl cutter"
[05:33:05] <theBear> apparently we're crazy and futuristic and wrong tho :)
[05:33:07] <Lambda_Aurigae> we have one.
[05:33:33] <Lambda_Aurigae> it's called a cricut
[05:34:09] <theBear> way waaaaay back early 90s when i raced pushbikes, one of the local bike shop/teams had one out the back.... very cool for getting things like stickers knocked up whileUwait
[05:34:25] <Lambda_Aurigae> yup.
[05:34:48] <Lambda_Aurigae> lady I work with has one that will do 4 foot wide vinyl.
[05:35:01] <theBear> it could also do reverse-contrast style where ultimately you get say, the letters out of your design as individual (but shared backing for application) characters vs just cut outta a rectangle
[05:35:17] <Lambda_Aurigae> yuppers.
[05:36:32] <Lambda_Aurigae> http://hackaday.com/2016/09/22/electric-train-demonstrator/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+hackaday%2FLgoM+%28Hack+a+Day%29
[05:36:33] <theBear> my little bro used to build modern EXPENSIVE litigation-proof playgrounds for a lotta years... they cut big stuff there for upholstering, ya know, stupidly large chunks of playground so idiots couldn't be nat-selected by them, but that is the also the first time i heard of either undersized plunge routers or vinyl-cutter as a term, 'cos they didn't have a machine :)
[05:36:37] <Lambda_Aurigae> this is neato.
[05:36:42] <Lambda_Aurigae> gonna have to make one.
[05:36:54] <Lambda_Aurigae> only, I wanna make one that's like 3 to 5 meters long.
[05:44:05] <Lambda_Aurigae> http://www.waynesthisandthat.com/How%20To%20Build%20The%20Simplest%20Electric%20Train.html
[05:44:11] <Lambda_Aurigae> that's just too awesome.
[05:45:51] <theBear> heh,. trailer park boys (comedy series) have a few episodes where they using a little traditional toy train and lines etc to smuggle weed and in exchange /return-trip cigarettes, across the canadia border, which i found very entertaining
[05:46:18] <theBear> hmmm, interesting approach on an old phenomenon
[05:46:28] <Lambda_Aurigae> yeah..nothing new..it's just a motor.
[05:46:34] <Lambda_Aurigae> but,,,it's still fun
[05:47:30] <Lambda_Aurigae> problem is getting the magnets
[05:48:04] <Thrashbarg> Magnets. Always with the magnets.
[05:48:10] <Lambda_Aurigae> yup.
[05:48:14] <Lambda_Aurigae> ask the mormons how they work.
[05:49:14] <Thrashbarg> lol
[05:49:26] <Lambda_Aurigae> it's an ooold meme...
[05:49:56] <theBear> mormons did magnets before clowns ? i gotta see this
[05:50:08] <Lambda_Aurigae> no...the asking the mormons about magnets.
[05:50:13] <Lambda_Aurigae> they have answers for everything.
[05:50:33] <Thrashbarg> 2 + 2 = Jesus
[05:51:12] <Lambda_Aurigae> gotta find some n48 supermagnets.
[05:51:33] <theBear> oooooh, they're not just baffled by magnets, they're christian-science-deny'ers about 'em
[06:13:05] <Lambda_Aurigae> Be wary of strong drink. It can make you shoot at tax collectors - and miss.
[06:13:58] <Thrashbarg> A penny saved is a penny taxed.
[06:14:11] <Lambda_Aurigae> After eating an entire bull, a mountain lion felt so good he started roaring. He kept it up until a hunter came along and shot him. The moral: When you're full of bull, keep your mouth shut.
[06:16:24] <theBear> heh
[10:01:43] <federico3> hi
[10:17:08] <federico3> I'm getting "ser_recv(): programmer is not responding" from avrdude from 2 different hosts when trying to upload
[10:23:56] <Lambda_Aurigae> something's broke
[10:24:37] <theBear> mmm, sounds like the programmer isn't responding, possibly cos it never getting the request, possibly cos it broken, possibly it ain't plugged in
[10:25:11] <Lambda_Aurigae> reverse the polarity of the neutron flow.
[10:25:13] <federico3> the "arduino" GUI can upload successfully. Only avrdude does not work
[10:25:32] <federico3> the device is a Nano 2 with atmega168
[10:25:57] <Lambda_Aurigae> first problem....arduino...just sayin'
[10:26:35] <Lambda_Aurigae> without some additional information, like, avrdude command line, arduino virtual serial port number, etc,,,we can't be much help.
[10:27:05] <Lambda_Aurigae> are you using an external programmer or trying to go through the arduino bootloader?
[10:28:05] <theBear> sidenote: arduino gui USES avrdude
[10:29:58] <federico3> theBear: good - any way to get arduino to print out the commands it's running?
[10:30:46] <federico3> found it
[10:31:26] <federico3> and it worked
[10:32:33] <theBear> probably, and good - cos i dunno where, and hooray
[10:32:35] <theBear> what worked ?
[10:32:43] <theBear> ya know, so we can tell the next guy
[10:33:06] <Lambda_Aurigae> why? we can just badger the next one the s!ame way
[10:33:38] <theBear> heh, and it's never gonna be a problem that affects me, good call <grin>
[10:33:52] <federico3> I've been googling around for this issue as it seems to happen frequently. Some people hinted at lowering the baudrate from 115k to 57k, yet arduino is using 19k2
[10:35:06] <Lambda_Aurigae> theBear, as we discussed earlier....I'm a total bastard!
[10:35:34] <Lambda_Aurigae> federico3, the usb-cdc on the pc side can likely be from 9600 to 57K
[10:36:04] <Lambda_Aurigae> it will wait for the ttl-serial side running at 19.2K
[10:36:23] <Lambda_Aurigae> there's some soft-flow control implemented in the CDC side.
[10:37:21] <federico3> interestingly, out of 9600, 19200, 38400, 57600 and 115200 only 19200 works
[10:37:44] <Lambda_Aurigae> then they have hardcoded it to 19.2
[10:37:44] <theBear> federico3, usually a given (serial) bootloader will expect a definate/specific speed
[10:38:09] <federico3> theBear: thanks again for this tips
[10:38:19] <Lambda_Aurigae> the bootloader on the atmega chip there is definitely hard coded.
[10:38:29] <Lambda_Aurigae> the usb-interface chip can be set several ways.
[10:38:38] <theBear> even if you did implement auto-sense on the speed, it'd be "risky" trusting data from such a thing for flashing anyway
[10:38:42] <Lambda_Aurigae> depending on how they are doing the usb interface.
[10:38:59] <Lambda_Aurigae> theBear, parity check!
[10:39:05] <theBear> you really don't want any extra chance of effed up data than you absolutely need in these cases
[10:40:13] <federico3> being a serial over USB, I wonder if there's some out-of-band signalling to advertise the baudrate
[10:40:28] <theBear> Lambda_Aurigae, like i said to someone yesterday, guaranteed 0 or >= 1 bit errors, with likely guaranteed integrity except in the case of various multi-bit errors where parity tells you the opposite of what it there to do, pfft, this is the future, we can do better than that :)
[10:41:05] <Lambda_Aurigae> CRC then
[10:41:13] <Lambda_Aurigae> or, parity check every bit!
[10:41:31] <Lambda_Aurigae> I like the original lego mindstorms IR comms protocol.
[10:41:37] <Lambda_Aurigae> would send everything forward, then backward
[10:41:50] <theBear> federico3, there ain't, and nano sounds like an italian thing, which means there's a slightly >50% chance it got a ft232/similar chip rather than usb native avr, which means such out of band anything would be quite impossible, as only 2 serial lines for in and out are physically wired up
[10:42:06] <Lambda_Aurigae> theBear, atmega168..no hardwaer usb.
[10:42:17] <theBear> Lambda_Aurigae, howbout just set the right baud if you wanna have a serial bootloader ? <grin>
[10:42:53] <theBear> Lambda_Aurigae, *BAM* thankyou for proving my suspicions and confirming my formerly possiblly relevant point :)
[10:43:35] <Lambda_Aurigae> hehe.
[10:43:46] <Lambda_Aurigae> dang but I can't type today.
[10:43:52] <Lambda_Aurigae> also doing 3 things at once.
[10:44:06] <Lambda_Aurigae> there are some ardweenies that are using v-usb too.
[10:44:36] <Lambda_Aurigae> the nano has an ftdi chip on the underside.
[10:44:38] <Levitator> Arduino's library is truly weak.
[10:45:00] <Lambda_Aurigae> ft232rl
[10:45:11] <Lambda_Aurigae> Levitator, bloated as hell.
[10:45:50] <Levitator> There are some Chinese nano clones with some sort of knockoff USB chip, and they work great.
[10:46:52] <Levitator> I don't know whether it's bloated. I mean, I don't think it even uses malloc. Think it's like >300 bytes to link digital I/O.
[10:47:00] <Levitator> Trouble is, its feature set is shite.
[10:48:23] <Lambda_Aurigae> hehe.
[10:48:38] <Lambda_Aurigae> have you seen the comparison of digital_write and simply setting the pin?
[10:48:48] <Levitator> I'm working on a universal async I/O stream library in C++
[10:49:06] <antto> don't tell Lambda_Aurigae
[10:49:06] <Levitator> Worried it's going to be big, but at least it would be good instead of useless.
[10:49:09] <antto> oh wait
[10:49:35] <Lambda_Aurigae> http://pastebin.com/dSRSxgax
[10:49:48] <Levitator> <300 bytes, I meant.
[10:49:55] <Lambda_Aurigae> if you can do it, good for you.
[10:50:50] <antto> why is it so much asm?! does it use a LUT to find which port/pin it has to poke?
[10:51:06] <Levitator> I'm going to abstract all of the AVR devices as streams of IO consisting of pipelines of configurably chained operations.
[10:52:07] <antto> in my own xmega "library" i have a input and an output pin object, which contain a pointer to the PORT_t struct, and a mask for the actual pin
[10:52:48] <Levitator> The idea behind this is that you can use all of the devices on the MCU at the same time.
[10:53:05] <antto> i haven't looked at the asm, but the on() method really should just do *p_port.set = mask;
[10:54:08] <Lambda_Aurigae> Levitator, in theory you can without such...but you do have to watch for interrupt overload.
[10:54:39] <Lambda_Aurigae> without dma on the chip you get some interrupt interferences sometimes.
[10:55:35] <Levitator> Sure. You can bang out all of your applications in machine code using a morse key if you really want to.
[10:56:36] <Lambda_Aurigae> I've seen such implementations before and watched them bloat horribly.
[10:56:56] <Lambda_Aurigae> but, if you can do it, GREAT!
[10:57:28] <Levitator> Working on the first device "driver" for it now. So far it's 126 bytes to initialize the USART.
[10:57:54] <Lambda_Aurigae> I've done a full state machine implementation to drive peripherals on the avr too...but that was waaay too much
[10:58:58] <Levitator> This is based on an idea of reusable chainable modules, so that should minimize code generation.
[11:00:40] <Levitator> So, for example, it will use the same DIO-sampling module regardless of whether the device is polled or interrupt-driven bitbang.
[11:15:45] <Levitator> Hrm. Trying to decide how to fix the asymmetry with how you would flush an output pipeline at the near end, but the input at the far end.
[11:16:14] <Levitator> Think it makes sense to wrap each linked list of pipe objects in a pipeline container. That should be neat and clean.
[11:17:48] <carabia_> still unsure of the practicality of this, but to each their own...
[11:18:20] <Levitator> Will probably have a significant fixed overhead, but the user code should be tiny.
[11:19:29] <Levitator> Besides, a lot of C++ optimizes out to nothing.
[11:20:15] <carabia> C++ mentioned, you just confirmed my suspicions
[11:21:45] <Levitator> Beats writing code that is 50% string expansions in the preprocessor.
[11:23:22] <Levitator> Ah. Better solution than the Pipeline object. Seems kind of useless. Just flush on the output, and let the device on the far call you back with its own flush, and that's a round rip.
[11:23:24] <Levitator> *trip
[11:23:25] <carabia> "okay"
[11:26:09] <carabia> "< Levitator> Sure. You can bang out all of your applications in machine code using a morse key if you really want to." ...But after this line, it all makes sense to me, haha
[11:26:57] <Levitator> C relies heavily on dirty hacks implemented with the preprocessor, which is in itself the dirtiest of hacks.
[11:27:34] <carabia> agreed on the preprocessor
[11:27:37] <Levitator> C++ replaces almost all of that nonsense with compiler functionality.
[11:28:18] <carabia> but this isn't just about the preprocessor
[11:28:37] <Levitator> C++ isn't inherently more bloated. You can use it like C if you want to.
[11:29:23] <carabia> surely you can, something tells me you aren't, though
[11:29:32] <carabia> of course you can use it like C is you write C with it
[11:30:06] <Levitator> They aren't identical, but C++ idioms are for the most part a superset.
[11:30:43] <Emil> Someone bashing C here? Off with their heads!
[11:31:43] <carabia> that's also true and besides the point
[11:32:12] <Emil> Anycase, 126 bytes to init uart? What?
[11:32:22] <carabia> it's an arduino rewrite \:D/
[11:32:52] <Levitator> If it were Arduino, then it would be totally useless.
[11:32:59] <Emil> So bloat build on top pf bloat? I see
[11:33:01] <Levitator> Arduino is a glorified demo library.
[11:33:30] <Emil> Nah, the general idea of Arduino code is okay, depending on how you view hal
[11:33:50] <Emil> What I hate about it is the inefficient implementation
[11:33:51] <carabia> Emil: except that HAL is almost useless on a platform like avr haha
[11:34:00] <Levitator> Arduino is terrible. It is all blocking calls.
[11:34:05] <Emil> carabia: depends
[11:34:11] <carabia> Emil: doesn't even
[11:34:17] <Emil> It does.
[11:34:31] <carabia> for what do you need hal?
[11:34:43] <Levitator> HAL does seem rather useless in Arduino land.
[11:35:00] <Emil> But I sort of agree that the AVR is small enough that you dont really need one per se
[11:35:12] <carabia> well that's kind of my point
[11:35:31] <Emil> But not useless
[11:35:49] <theBear> hal is the only way arduino can get as lower performance outta thigns as simple as writing a byte to a port as they apparently insist upon
[11:35:49] <Emil> Its all about what you are willing to code about
[11:36:03] <carabia> well, okay, not completely, but the way i see it it's overkill
[11:36:23] <Levitator> Well, the thing about HAL is that here you are talking about an environment where the user is going to be soldering something together to use your application.
[11:36:25] <Emil> theBear: I'm not talking about the shit on arduino, Im talking about the general idea of hal, which is good
[11:36:28] <theBear> and if you think about it, in many ways just the basic avr-libc defines and register names and that kinda stuff already do a limited amount of abstracting
[11:36:44] <Levitator> And he has to think consciously about which pin is which anyway, so it doesn't save very much.
[11:36:57] <theBear> Emil, yeah i know <grin> but if i see an opening, i gotta get out one insult when arduino or ms or someone like that is involved
[11:37:09] <Emil> ;)
[11:37:12] <carabia> theBear: well yes, it would be hal on-top of pseudo-hal
[11:37:29] <Emil> We can all agree that Arduino code is shit
[11:37:34] <Emil> that's for sure
[11:37:48] <Levitator> I don't know about the code itself. The API design is shit for certain.
[11:38:02] <Emil> The code, also
[11:38:15] <Levitator> Makes almost zero use of the device's extensive capabilities.
[11:38:33] <Emil> But then you'd have to think!
[11:38:43] <Emil> And we can't have that
[11:38:55] <Emil> Anycase,` 126 bytes to config uart, what?
[11:38:59] <carabia> Emil: you're missing the point, he's writing his own fast-proto set
[11:39:20] <Levitator> What's wrong with that?
[11:39:26] <Emil> carabia: I'm not missing anything
[11:39:36] <carabia> and while I do agree for someone new to writing for avrs, it could be alright. Personally I just have a distaste for it because avr's are simple enough so I can demo anything in an afternoon
[11:39:36] <Levitator> It's dynamic config, not statically compiled. You can specify any port configuration at runtime.
[11:41:05] <Emil> So?
[11:41:22] <Levitator> So, is 126 bytes large?
[11:41:27] <Emil> Yeah
[11:41:46] <Levitator> Well, it's also setting up the chained IO framework.
[11:42:08] <Emil> Thats better
[11:42:17] <Emil> And what bytes are you exactly meaning?
[11:42:24] <carabia> anyway, i'm going back to writing a simplified hal for an arm chip. ":D" But anyway I have enough memory, mips and bandwidth I can piss any of them at will.
[11:42:40] <Levitator> Bytes as displayed by avr-size
[11:42:42] <Emil> carabia: do you use a cli setup?
[11:42:44] <Emil> Please share
[11:42:59] <carabia> Emil: well, I could
[11:43:07] <Emil> Levitator: so you mean flash size?
[11:43:16] <Levitator> Binary size, yeah.
[11:43:31] <Levitator> Well, binary size delta.
[11:43:31] <theBear> final word, as far as my standard anti-uino rants/thoughts go, the port remapping is one of the ones that still shocks me the most... renaming/juggling ports/bits to the point where what is one byte/port on the physical board/naming/layout is made up of more than one real port (same bits, just from weird avr-pins) !! apart from just being mental, this leads to ugly inefficient bit juggling everytime the raw/uncompiled code tries to change a pin/port
[11:43:31] <carabia> but I'm using eclipse because it integrates gdb+openocd
[11:43:32] <theBear> output
[11:43:36] <theBear> iinputs too i spose
[11:43:51] <Emil> carabia: faith, ripped out
[11:44:23] <Emil> theBear: completely agree
[11:44:26] <carabia> Emil: it's not that bad, it's not any different than using whatever to write the code and then compiling with arm-gcc and uploading with openocd
[11:44:29] <Levitator> How does ARM development compare to the AVR?
[11:44:45] <carabia> though eclipse is /bloated/, but integrated debugging is a boon.
[11:44:48] <Emil> Levitator: "depends"
[11:45:00] <theBear> one is done at the gym, and another at a desk :)
[11:45:07] <carabia> Levitator: I'm actually quite sure you'd find yourself at home with ARMs
[11:45:08] <Levitator> Eclipse is a mess. I used it for a long time. It is super slow and full of bugs. Netbeans is better.
[11:45:18] <Levitator> Visual Studio is also massively bloated, but it does work properly.
[11:45:22] <Emil> >Netbeans
[11:45:59] <Emil> It is cli or nothing! Occasinal gedit for some random things
[11:46:06] <carabia> I'm not sure which version of eclipse sw4stm is running, but it's not all that slow. Could be because this computer is quite a beast also.
[11:46:27] <theBear> seriously tho, i'd tend to inquire if you mean teeny arm like the new thumb /tiny micro dev things the last few years, or if you mean big arm like your mobile or a fancy laptop perhaps, cos the only common things are the name and (some at least of) the instruction set
[11:46:41] <Emil> Levitator: ARMs have no "sane defaults". Nothing works out of the box
[11:46:51] <carabia> Emil: but you don't really need any setting up for cli, openocd has all the necessary scripts needed for uploading
[11:47:10] <Levitator> Incidentally, VS is a feat of code bloat worthy of Microsoft. How they write something in C++ that chugs like a comparable Java application is impressive.
[11:47:29] <Emil> Levitator: you have to tell it to set up a lot of shit befoee you can actually use the device
[11:47:39] <carabia> and i'm using a graphical tool for clock generation PLLs. Though I've been doing this for long enough so that I know the registers by heart by now
[11:47:46] <carabia> Emil: not really
[11:48:06] <carabia> Emil: you enable the clock for the peripheral, configure it, enable it
[11:48:21] <Emil> carabia: exactly.
[11:48:37] <Emil> carabia: I know what you need to do to get it working but the defaults arent sane
[11:48:42] <carabia> really the only difference to avrs is that you manually enable clocks separately for every device
[11:48:57] <carabia> and you specify a clock source
[11:49:02] <Emil> It would be way better if you disabled stufd on startup
[11:49:11] <carabia> meh
[11:49:27] <Levitator> I've been out of a job for a long time, and playing with Arduinos is gradually motivating me to consider opening a hardware business.
[11:49:56] <Levitator> When I look at the prices of consumer products, and then I add up the cost of the parts to build it by hand, it starts to open up some opportunities.
[11:50:25] <Emil> Levitator: you arent taking into account cost of design, marketing, manufacturing and so on
[11:50:47] <Emil> I used to think the same way too: "This really should just cost pennies!"
[11:50:58] <Levitator> Well, I do all the design and manufacturing myself.
[11:50:58] <Emil> Then I started in uni and grew up a bit
[11:51:02] <carabia> well if you're a one-man-show, design costs are negligible
[11:51:23] <Emil> carabia: still not neglible if you want to run an effective business
[11:51:24] <carabia> then you just go all-in and order a huge batch of pcbs off china!!
[11:51:30] <Emil> time is money
[11:51:34] <Levitator> The part I have questions about are what sort of endorsements or licenses you need to deliver a physical product.
[11:51:49] <Levitator> And what sort of cost is involved in molding or 3D printing the enclosures.
[11:51:50] <carabia> Emil: well, no, basically it boils down to whether you want to eat or not.
[11:52:04] <carabia> And who likes eating anyway, just makes you fat.
[11:52:28] <Emil> Levitator: 3d printing only for low volume and you need to do surface finishing
[11:52:30] <carabia> Levitator: well, depends on the batch size
[11:52:42] <Emil> injection molding starting from 5k for a small mold
[11:52:49] <Emil> usually around 10k
[11:53:01] <Levitator> So, like, do you have to get your product verified by the FCC, or some other government agency? Do you have to subcontract to an electrician to do installs? Etc.
[11:53:21] <carabia> for 'murica, I guess
[11:53:22] <Emil> If you want wireless then buy modules, at least in the beginning
[11:53:28] <carabia> for any RF-stuff yeah
[11:53:38] <Emil> carabia: if you buy modules then no
[11:53:46] <carabia> because they're already approved...
[11:53:50] <Emil> yeah
[11:54:30] <Levitator> Anyway, I've seen stuff for sale, and run a quick tally of the parts involved, and feel like I could build the same thing by hand on a workbench for about 25% of retail.
[11:55:44] <Emil> So 25% cheaper?
[11:55:56] <Levitator> No, 75% cheaper.
[11:56:01] <Emil> Lolno
[11:56:28] <Emil> Yeah, for yourself you probably could
[11:56:29] <Levitator> Well, I mean, considering have an $8 computer with WiFi on my desk in front of me.
[11:56:33] <Emil> but not for sale
[11:56:44] <Levitator> The difference between the two is the cost of the enclosure.
[11:57:16] <Emil> By all means, try
[11:57:30] <Emil> if you succeed we'll have a lot cheaper products
[11:57:34] <Levitator> I need to look into 3D printing and get an idea of cost.
[11:57:40] <carabia> Levitator: there's a lot of /everything/ for sale, due to pcbs being so cheap (for simple applications, 2 layer, with nice trace widths and clearances), and everyone and their dog having a diy-reflow oven
[11:58:18] <carabia> even 4-layers aren't all that bad anymore.
[12:00:35] <carabia> for a small batch test run i'd probably design the product and functionality around an already-existing case, however
[12:06:13] <Tom_itx> hammond mfg has some nice cases
[12:06:35] <Levitator> I went to some asshole's computer shop to pick up scrap equipment and he gave me a power supply unit that was nearly blown apart by a lightning strike.
[12:06:45] <Levitator> Then, he kicked me out for questioning his pro-Trump schpiel.
[12:07:18] <Tom_itx> the cost of doing business for free
[12:07:29] <Levitator> Was indignant that I would insult his generosity by maligning his idiotic candidate.
[12:07:38] <Levitator> Need to take up dumpster-diving.
[12:08:06] <Tom_itx> yeah nobody will bother you there unless you fall in on another diver
[12:09:15] <Levitator> Damn thing practically has a hole burned through the board, and they are made of fiberglass.
[12:09:39] <Levitator> There is a greasy stain where a group of components arced to the case.
[12:10:18] <carabia> Tom_itx: hammond cases are just quite expensive, last I checked
[12:10:55] <carabia> though i've never bought them in big batches, maybe they're okay then
[12:17:42] <Tom_itx> i used them on my programmer
[12:18:16] <Tom_itx> http://tom-itx.no-ip.biz:81/~webpage/boards/USBTiny_Mkii/USBTinyMkII_Blue.jpg
[12:18:31] <carabia> How much were they per unit?
[12:18:36] <Tom_itx> ask dk
[12:18:39] <Tom_itx> they stock em
[12:18:41] <Tom_itx> i forget
[12:18:49] <Tom_itx> maybe a buck or two ea
[12:18:50] <carabia> how big a batch?
[12:18:53] <Emil> dk?
[12:18:57] <Tom_itx> digikey
[12:18:57] <carabia> digikey
[12:19:00] <Emil> ah
[12:19:05] <Tom_itx> i'll get you a pn..
[12:19:33] <carabia> nah it's fine, i'm not in the market for cases. And yeah, that's a reasonable price though the case is quite small.
[12:19:40] <Tom_itx> 1551FTBU
[12:19:51] <carabia> obviously amazingly suited for the application. I like it.
[12:20:11] <Tom_itx> http://www.digikey.com/product-search/en?keywords=1551FTBU
[12:20:19] <Tom_itx> i milled the cutouts myself
[12:20:47] <Tom_itx> i usually got 25-50 at a time
[12:20:48] <carabia> did you just template + spray the markings? ;D
[12:20:58] <Tom_itx> depending on where the price break was
[12:21:04] <Tom_itx> hang on...
[12:21:32] <Tom_itx> http://tom-itx.no-ip.biz:81/~webpage/boards/USBTiny_Mkii/silkscreen/
[12:21:42] <Tom_itx> i made a silkscreen
[12:21:54] <Tom_itx> used UV paint so it would stick to the plastic
[12:22:23] <carabia> Yea alright. Kind of elaborate!
[12:22:34] <Tom_itx> i made it all
[12:22:56] <carabia> I like the "speedball" branding on the sensitizer.
[12:23:08] <Tom_itx> http://tom-itx.no-ip.biz:81/~webpage/boards/USBTiny_Mkii/silkscreen/lid_cutout2.jpg
[12:23:18] <Tom_itx> to position the lids in the silkscren
[12:24:01] <Tom_itx> you use very very little of that
[12:24:30] <Levitator> Heh. I could cater to steampunk types and build them electronics in woodgrain.
[12:24:37] <Tom_itx> the print came from an eagle layer
[12:25:05] <Tom_itx> printed on clear sheet and exposed on the silk
[12:25:13] <carabia> Levitator: nono, for steampunk make the cases from old pcbs or something
[12:25:23] <Tom_itx> http://tom-itx.no-ip.biz:81/~webpage/boards/USBTiny_Mkii/silkscreen/exposure.jpg
[12:25:29] <Levitator> Reminds of the 80s, when they sold the Atari 2600 with faux wood veneer.
[12:25:51] <Tom_itx> you can barely see but the clear is on top of the mask with a piece of glass to hold it flat
[12:26:06] <carabia> Tom_itx: I would have seriously just not bothered, haha
[12:26:16] <Tom_itx> makes it look better
[12:26:18] <Levitator> Because people were still accustomed to consumer electronics resembling furniture, and it would be weird to make something all-plastic.
[12:26:22] <Levitator> Ah, the old days.
[12:26:23] <carabia> As the markings are a nice touch, surely, but not very informative
[12:27:01] <theBear> if it could assemble the furniture too, ikea would suddenly be accessible to those without any practical skill too :)
[12:27:18] <carabia> Levitator: i have always liked sony's style the best, really
[12:27:34] <Levitator> Also, I bet there is money to be made creating semi-functional reproductions of sci-fi artifacts.
[12:27:51] <Levitator> I don't think I would stoop to such gimmickry, unless it was actually useful.
[12:28:17] <Levitator> Maybe post it to Etsy.
[12:28:42] <carabia> a tricorder
[12:29:14] <Levitator> Yeah, that crossed my mind, because there's all manner of random sensors you could pack into a device, and that would actually be useful.
[12:29:33] <Tom_itx> hell, make em functional...
[12:29:42] <carabia> and buzzers and lights.
[12:29:53] <Levitator> Heartbeat sensor, ultrasonic/laser rangefinder, IR thermometer, night-vision?
[12:30:20] <carabia> a radar with a backpack-battery
[12:30:38] <Levitator> And just make a thing that's basically a swiss-army knife for nerds, and model it after the tricorder just for kicks.
[12:30:51] <carabia> motorized 360 that extends the dish with a boom from the backpack
[12:31:14] <Levitator> I'm not sure how many hardcore Inspector Gadget fans there are out there.
[12:31:31] <carabia> all of this connected to the internet for another useless interweb of thongs(tm) -application
[12:31:38] <Levitator> But anyway, I can envision a multi-function nerdcore device which I would actually use, now that I think about it.
[12:31:49] <carabia> Inspector gadget not so much, the chick it the movie, would bang.
[12:32:52] <Levitator> Maybe someday I'll sit down and work out what shape/size cost you could manage to cram the most exotic tools into.
[12:34:11] <Levitator> Cell phones already essentially serve the same purpose for non-technical applications. They do everything.
[12:34:40] <Levitator> It's a phone, it's a camera, it's Internet access.
[12:37:03] <Levitator> So, this sort of niche thing is where you have an opportunity, because you aren't going to get high volume, but you might get people who are highly interested and willing to pay for it.
[12:43:08] <carabia> yeah, hackaday serves as an amazing marketing platform as every article nowadays - as bss pointed out - is something along the lines of "arduino/esp strapped to x" and the masses LOVE IT!
[12:45:04] <Levitator> Well, I want to make something actually useful.
[12:45:35] <Levitator> I mean, if people are knitting random trinkets together and making money off of that, then it stands to reason that you should be able to sell hand-made niche tools and electronics, too.
[13:16:39] <Lambda_Aurigae> Levitator, I would love to see what you are making when it's done or functional even. Along with use cases.
[13:18:40] <Levitator> I haven't decided what to make yet. It's just the only plausible business idea I feel motivated about right now.
[13:19:32] <Lambda_Aurigae> aahh...I'm interested in the actual use cases really, to see what you are working towards and compare to other methods of doing the same thing.
[13:19:57] <Lambda_Aurigae> so far what you've proposed sounds good but I don't see a real use for it in anything that I would be doing but I don't think I fully understand it either.
[13:21:18] <theBear> that's what she said
[13:21:34] <Levitator> But, I mean, there are various tools someone could want integrated for hobbyist purposes. If you are into model aircraft, for example, you might want a multimeter, a thermal imager, a weight scale, a servo debugger.
[13:23:28] <Levitator> I would use something with thermal sensing constantly for cooking, checking computer hardware, and checking HVAC efficiency.
[13:24:00] <Levitator> I wouldn't mind having a rangefinder for measureing rooms for wiring, especially if I still did LAN parties.
[13:25:20] <carabia> I'm not sure whether he was referring to your avr framework
[13:25:33] <Lambda_Aurigae> yup.
[13:26:01] <Levitator> Oh. Yeah. I'll find out shortly whether it's going to be big or not.
[13:26:03] <Lambda_Aurigae> by use case, I mean a real world application built upon avr, or, as I am understanding it, multiple microcontrollers.
[13:26:31] <Levitator> It's for AVR. I haven't messed with ARM yet.
[13:26:49] <Lambda_Aurigae> those aren't the only two microcontrollers out there though.
[13:27:18] <Levitator> They are the two I have in mind atm, anyway.
[13:27:45] <carabia> Indeed. If you don't want to venture into arm, I think pic32s would be more beneficial of that abstraction
[13:30:17] <Levitator> If nothing else, get an idea of the implications of working in an AVR environment with 2kB of RAM.
[13:30:24] <Levitator> I will, I mean.
[13:30:53] <Lambda_Aurigae> try one with 512 bytes of ram.
[13:30:57] <Emil> But
[13:30:58] <Emil> why
[13:31:07] <Levitator> I read that some had 128 bytes of RAM.
[13:31:08] <Levitator> Why what?
[13:31:17] <Lambda_Aurigae> some have 32 bytes...just the 32 registers.
[13:31:38] <Levitator> Ew.
[13:31:39] <Emil> But hey, at least everything is blazing fast ;)
[13:32:01] <Lambda_Aurigae> hell, go back to the commodore vic-20 and code for that.
[13:32:19] <Lambda_Aurigae> 5K of ram for both data and program.
[13:33:23] <Lambda_Aurigae> them was the days.
[13:33:43] <Lambda_Aurigae> homemade 8K expansion board and all.
[13:36:28] <carabia> Oh haha, commo reminds me.
[13:36:47] <Lambda_Aurigae> my vic-20, c-64, and c-128 all still work.
[13:37:09] <carabia> I stripped a 6581 off one of my old 64s
[13:37:22] <carabia> "the beast"
[13:37:24] <Levitator> I lost all of my vintage computing stuff because I left it at home when I moved out, and of course, nobody took care of it.
[13:37:54] <carabia> taking care is not a synonym for "not throwing out"
[13:38:16] <Lambda_Aurigae> the sid chip rocked.
[13:38:37] <Levitator> My Apple IIGs got chewed up by the dog, and who knows what happened to the rest.
[13:38:56] <carabia> bad doggy.
[13:39:22] <carabia> but it's apple, clearly, commo 64/128/v20 > apple II
[13:39:29] <Levitator> Well, I mean, nobody taught him not to chew up cables, which is a shame, because I had him fetching in a few minutes. He wasn't untrainable.
[13:40:01] <carabia> well, you can still repair it. I'm sure the dog didn't chew up the pcbs, yes?
[13:40:20] <Levitator> It got thrown out, no questions asked (of me).
[13:40:47] <carabia> Well, it was an apple II after all. Useless.
[13:41:19] <carabia> Reminds me, I gotta replace the microswitches on the joystick (yet again)
[13:41:28] <carabia> too much daily thompson...
[13:41:28] <Levitator> Eh. First computer had sentimental value. I patched a IIGS emulator recently for my purposes.
[13:41:43] <carabia> daley, even*
[13:41:47] <Levitator> I don't know why I develop patches for people so that they can completely ignore them.
[13:42:20] <carabia> didn't you say you patched it for yourself?
[13:42:42] <carabia> there should be some intrinsic value to that, by itself, for you
[13:42:51] <Levitator> Specifically, I patched it so that it wouldn't scramble the buttons on a modern game controller with >2 buttons.
[13:43:05] <Levitator> Forwarded it to the dev, never heard back.
[13:43:17] <Levitator> Patched in a pixel multiplier so that it wouldn't be tiny on my screen.
[13:43:46] <carabia> projects like these can often fall into oblivion
[13:44:04] <carabia> as can everything with software.
[13:44:14] <carabia> it's either a) forgotten b) rewritten
[13:46:26] <Lambda_Aurigae> hmmm.
[13:46:42] <Lambda_Aurigae> an AVR could easily simulate a 6502...it's been done several times.
[13:46:54] <Lambda_Aurigae> even running faster than the original 1MHz 6502.
[13:47:17] <Levitator> Probably. Don't know how much memory was typical on those.
[13:47:31] <Levitator> I think the IIgs had like 768kB, or something like that.
[13:47:35] <Lambda_Aurigae> oh, run external memory.
[13:47:41] <Levitator> It's been quite a long time.
[13:47:51] <Lambda_Aurigae> the C64 had 32K of rom and 64K ram...bank switched.
[13:47:58] <Lambda_Aurigae> or was it 48K rom?
[13:48:02] <Lambda_Aurigae> something like that.
[13:48:03] <Levitator> Does the 328p have an external memory interface?
[13:48:07] <Lambda_Aurigae> no.
[13:48:10] <Levitator> Blah.
[13:48:30] <Lambda_Aurigae> but for this purpose, an external memory interface would be easily done through the spi port with a 1Mbit serial sram.
[13:48:58] <Lambda_Aurigae> just wondering if I could duplicate the vic-II and sid-II chips.
[13:49:07] <Lambda_Aurigae> hmmmmmm.
[13:49:26] <Levitator> This place is full of nerds.
[13:49:34] <Lambda_Aurigae> of course.
[13:49:38] <Lambda_Aurigae> and damned proud of it!
[13:49:57] <Lambda_Aurigae> I was a nerd probably before you were born.
[13:50:47] <carabia> the pincount with 328p, could not possibly facilitate an external memory interface
[13:50:54] <Lambda_Aurigae> I still have my original vic-20...and it still works.
[13:50:55] <carabia> well, it could, but it would be quite ridiculous
[13:51:11] <Lambda_Aurigae> carabia, yeah..lots of interleave there.
[13:51:14] <Levitator> carabia: He said serial memory.
[13:51:17] <cehteh> there are onewire memories :D
[13:51:28] <carabia> well "serial memory" interface is pretty much... SPI
[13:51:37] <Levitator> For emulating an ancient processor, the latency is not unacceptable.
[13:51:56] <carabia> mcp has them, up to 1 Mbit, and they run up to what? 50MHz? I can't remember
[13:52:07] <Lambda_Aurigae> cehteh, yeah, but the 23LC1024 is spi...with sdi/sqi capabilities.
[13:52:10] <carabia> no wait, 20 I think
[13:52:18] <Lambda_Aurigae> rated at 20MHz
[13:52:22] <carabia> Yeah
[13:52:26] <Lambda_Aurigae> I've clocked them to 50MHz with little problem.
[13:52:38] <carabia> external memory interface != serial memory
[13:52:45] <Lambda_Aurigae> the logic-shrimp clocks them to 50MHz.
[13:53:03] <cehteh> but ,, when a single avr isnt suitable for the task then why do you go lengths to make it fit
[13:53:06] <Levitator> Is that anything like a babel fish?
[13:53:08] <carabia> yeah I somehow got it from that
[13:53:16] <cehteh> you can get a 32bit mpu even cheaper now
[13:53:17] <Lambda_Aurigae> but I wouldn't be using that external ram as real memory for the AVR, but as simulated ram for the virtualized c-64.
[13:53:35] <Lambda_Aurigae> cehteh, for the fun of it.
[13:53:37] <carabia> Lambda_Aurigae: obviously
[13:53:58] <cehteh> well there was this linux on avr simulator
[13:54:03] <carabia> simavr?
[13:54:09] <carabia> or simulavr, smth
[13:54:16] <cehteh> next logical step would be to emulate a host processor, x-series
[13:54:17] <carabia> Oh wait
[13:54:17] <Lambda_Aurigae> logic shrimp should have been logic pirate
[13:54:21] <carabia> That's the other way round, hahaha
[13:54:23] <Lambda_Aurigae> http://dangerousprototypes.com/docs/Logic_Pirate
[13:54:25] <cehteh> nothing else is fun
[13:54:30] <cehteh> carabia: yes :)
[13:54:32] <Lambda_Aurigae> there was linux on avr, yes
[13:54:37] <carabia> I think there's a js interpreter on avr...
[13:54:41] <Lambda_Aurigae> linux running on an arm emulator running on an avr.
[13:54:44] <carabia> I think that's the next big thing.
[13:54:48] <Lambda_Aurigae> takes about 4 hours to boot to command prompt
[13:54:48] <cehteh> http://dmitry.gr/index.php?r=05.Projects&proj=07.%20Linux%20on%208bit
[13:55:22] <Lambda_Aurigae> I have one here cehteh
[13:55:24] <carabia> cehteh: by any chance would you be russian?
[13:55:31] <cehteh> no
[13:55:35] <Lambda_Aurigae> it is,,,,functional.
[13:55:36] <carabia> but close?
[13:55:36] <Lambda_Aurigae> barely
[13:55:40] <Lambda_Aurigae> but it does run.
[13:55:42] <cehteh> why?
[13:55:50] <carabia> out of curiosity, is it a problem?
[13:56:13] <cehteh> thats not my site :D
[13:56:19] <carabia> i know
[13:56:24] <cehteh> german here
[13:56:37] <carabia> alright, it was for reasons unrelated to the site
[13:58:14] <cehteh> heh
[13:58:17] <cehteh> what reasons?
[13:58:31] <carabia> secret.
[13:58:43] <carabia> you don't have the required security clearance
[13:59:24] <Lambda_Aurigae> we could tell you
[13:59:29] <Lambda_Aurigae> but we would have to kill you first.
[13:59:31] <carabia> but we'd have to kill you
[13:59:33] <carabia> meh
[14:02:43] <cehteh> :D
[14:16:59] <Levitator> Ok, so I'm playing with the idea of overloading the pipe operator so that if you apply it to two things which are a stream, you get a pipeline which is also a stream.
[14:17:14] <Levitator> And although it sounds elegant, I wonder whether it actually accomplishes anything.
[14:17:51] <cehteh> i just dont use C++ at all
[14:18:05] <Levitator> Currently, it just attaches two pipes together as a linked list. The thing that I can imagine gaining from tracking things as pipelines, is that you know both endpoints of the linked list without linking it doubly.
[14:18:22] <cehteh> also whatever you do in C++ it most likely bloats
[14:18:56] <Levitator> Depends on the programmer, not the language.
[14:20:29] <cehteh> well if you dont use any C++ features, then C++ wont bloat that much :)
[14:27:34] <Emil_> This right here
[14:28:28] <Emil_> With C you can be pretty confident in knowning how your code translates
[14:28:28] <Emil_> to assembly
[14:28:28] <Emil_> and direct instructions
[14:28:40] <cehteh> not really true with gcc .. but still
[14:28:47] <cehteh> much better than C++
[14:28:48] <Emil_> Pretty true
[14:29:05] <Emil_> But yeah, you can never be completely sure
[14:29:12] <cehteh> gcc does a lot surprising optimizations ..
[14:29:36] <Emil_> Can you tell about them?
[14:29:58] <Emil_> Would be interesting to learn what you have figured out
[14:32:23] <cehteh> just read the gcc docs about what optimizations it can do
[14:32:58] <Levitator> The hell is wrong with Freenode today?
[14:33:50] <cehteh> on x86 https://blog.fefe.de/?ts=a922623d (use a translator of your choice) for example
[14:34:06] <cehteh> but thats a moderate trivial optimization
[14:34:08] <Emil_> Levitator: ddos
[14:34:12] <Emil_> Over now
[14:34:46] <Levitator> Why the hell do people even do that?
[14:34:51] <cehteh> register/code rescheduling and lto will put your code into a blender and result into something very different
[14:35:19] <cehteh> i have good results with lto, but the resulting assembly looks really strange
[14:35:58] <cehteh> fragments of my code all reorders and littered over the place, jumps, shortcuts, etc
[14:36:53] <Emil_> A macro compiler project would be cool to do
[14:37:04] <Emil_> or even just a very basic compilee
[14:37:17] <Emil_> -e+r
[14:37:54] <Emil_> Levitator: because shitnode politics is cancer
[14:38:02] <cehteh> looks at tinycc
[14:38:20] <cehteh> but its close to impossible to port to 8bit
[14:38:54] <Emil_> Why's that?
[14:42:42] <carabia> Fucking freenode.
[14:43:17] <cehteh> the codebase was not written with portability in mind, its very much 32 bit (or 64bit) code
[14:43:43] <cehteh> it originates from a obfuscated c code contest ... :)
[14:44:07] <carabia> what is?
[14:44:13] <Chillum> :)
[14:50:37] <Emil_> carabia: ioccc is awesome
[14:52:28] <carabia> Good. I hope freenode gets its shit together now
[14:53:55] <Emil_> As id
[14:53:58] <Emil_> As if*
[17:56:08] <daey> im searching for a single chip boost converter that does 3 ish V->5V. current demand is in the 10mA range. Does something like that exist?
[17:57:36] <LeoNerd> By "single chip", do you mean one with onboard inductor inside the chip?
[17:57:41] <daey> yes
[17:57:43] <LeoNerd> I'm not sure you'll find one of those.
[17:58:02] <LeoNerd> I've never seen a silicon-scale inductor for power use
[17:58:56] <theBear> silicon doesn't grow scales like fish do
[17:59:39] <daey> guess ive to use an external one then. apparently lcds tend to need 5V to be driven :/
[17:59:42] <theBear> you do know there are inductors roughly the size of a 1/2w resistor we assumes ? look like little tiny medicine-capsules
[18:00:13] <daey> theBear: its for a watch project along the line of https://github.com/carrotIndustries/pluto
[18:00:14] <theBear> probly most lcds or ones copying a certain veyr popular ancient controller
[18:00:54] <daey> he uses a uC that has an integrated lcd driver which apparently boosts to 5V internally. i want to do it via gpios and an avr
[18:01:40] <cehteh> mhm charge pump, but 10mA might be already bit much
[18:02:07] <LeoNerd> If you're just driving an LCD, that won't need anywhere like 10mA
[18:02:13] <daey> 10mA was just a guess. ive no clue what the watch draws. it would essentially be the uC and the display + backlight
[18:02:32] <LeoNerd> Capacitive charge pump for LCD is fine because they're in the microamps range, if that. LCD passes almost no current
[18:03:01] <cehteh> if it is just the lcd then a charge pump could be approbiate
[18:03:06] <cehteh> yes
[18:03:21] <cehteh> and less loss than a boost converter
[18:03:40] <daey> well its the inputs of the lcd. i doubt i have the room for individual charge pump circuitry
[18:03:52] <Lambda_Aurigae> max233
[18:04:02] <Lambda_Aurigae> not exactly built for that, but,,,
[18:04:08] <theBear> daey, i've heard people "steal" power from max232 or similar chips, basically using their integrated switched cap volt pumps for something else too
[18:04:16] <Lambda_Aurigae> probably a bit large.
[18:04:22] <Lambda_Aurigae> yup.
[18:04:24] <theBear> Lambda_Aurigae, dammit, think i got a useful idea we all forgot and *BAM* there you are
[18:04:31] <cehteh> discrete charge pump doesnt need much components
[18:04:32] <Lambda_Aurigae> I used a max233 to generate a high voltage for a pic programmer once.
[18:04:40] <cehteh> i bet there are IC's for that
[18:04:45] <Lambda_Aurigae> max233 has built in caps.
[18:04:55] <cehteh> but itself its huge :D
[18:04:59] <Lambda_Aurigae> yup.
[18:05:07] <Lambda_Aurigae> as I said, a bit large.
[18:05:25] <cehteh> first of all, make really sure how much current you need
[18:05:29] <daey> cehteh: there are the problem is the displays have random amount of Common connectors. the lcd interfaces are build for a specific amount from what i was able to grasp. so you are kinda screwed
[18:05:49] <cehteh> huh?
[18:05:55] <cehteh> i dont understand
[18:06:25] <daey> cehteh: of the former two, it turned out that people at Casio /really/ like building LCDs with 5 backplanes. For some reason, the LCD panels need also to be driven with around 5 Vpp to achieve decent contrast.
[18:06:39] <daey> To my best knowledge, there's no (non-unobtainium) MCU on the market, that can boost a 3 V supply to 5 V for driving and LCDs /and/ can do 5-mux.
[18:06:45] <daey> from the project site
[18:07:21] <cehteh> you mean boost gpio's to 5v?
[18:07:37] <daey> yes
[18:07:49] <cehteh> some mcu's can be driven as open collector
[18:07:50] <daey> or well. let everything run at 5V which was my first idea
[18:08:36] <cehteh> avr's iirc cant, or at least the nasty protection diodes come into your way
[18:08:53] <cehteh> what do you mean by watch .. wristwatch?
[18:09:01] <daey> yes
[18:09:22] <daey> the project i linked is about custom hardware for a casio f91 (the terrorist watch)
[18:09:33] <cehteh> you should be able to use a small fet array for driving the lcd
[18:10:23] <daey> well then i would have a charge pump circuit and a fet array on top
[18:11:15] <cehteh> charge pumps can be pretty small
[18:11:25] <Lambda_Aurigae> just had a horrid horrid idea....a way to get higher voltage but super inefficient.
[18:11:43] <Lambda_Aurigae> a whole bunch of teeny tiny solar cells in series fed by an LED.
[18:11:46] <cehteh> super inefficient for wristwatch is prolly a bad idea :)
[18:11:52] <daey> lol
[18:11:58] <Lambda_Aurigae> I did say it was a horrid horrid idea.
[18:12:02] <daey> active backlight to activate screen :P
[18:12:05] <Lambda_Aurigae> but,,,it would work, kinda.
[18:12:23] <Lambda_Aurigae> or, rather, could be made to work, kinda.
[18:12:27] <cehteh> lol
[18:12:48] <cehteh> daey: are you tied to avr's?
[18:13:18] <cehteh> otherwise i'd look into other chips .. w/ open collector .. maybe you even find cheaper smaller arm cores
[18:13:26] <daey> well i dont have any programming hardware for other chips
[18:14:06] <daey> and i want something that is well supported by gcc etc.
[18:14:12] <cehteh> http://cache.freescale.com/files/microcontrollers/doc/fact_sheet/KINETISKL02CSPFS.pdf?fpsp=1&WT_TYPE=Fact%20Sheets&WT_VENDOR=FREESCALE&WT_FILE_FORMAT=pdf&WT_ASSET=Documentation
[18:14:14] <Lambda_Aurigae> make the tools to make the tools
[18:14:15] <cehteh> :D
[18:14:22] <cehteh> some arm core would be fine
[18:14:24] <daey> i could go arm but im not sure thats the right step :P
[18:14:33] <Lambda_Aurigae> msp430?
[18:14:35] <cehteh> 1.9x2mm ..
[18:14:50] <cehteh> and you get tons of features :D
[18:14:58] <cehteh> but eh assembling will be extra fun
[18:15:53] <daey> the pcb alone is already a pita. 0.6mm castellated vias ...
[18:16:11] <Lambda_Aurigae> that big?
[18:16:26] <daey> 'big'
[18:16:29] <cehteh> hehe
[18:16:35] <cehteh> yes for a wristwatch
[18:17:21] <daey> i have to check what the pcb costs. probably a fortune T_T
[18:19:04] <cehteh> what shall that become, some simple smartwatch?
[18:19:41] <daey> well i would rather call it a feature watch :P i doubt its going to be very smart
[18:20:43] <cehteh> thats my thinking, with avrs you cant do that much
[18:21:28] <cehteh> software is already finished?
[18:21:34] <cehteh> pluti-fw?
[18:22:14] <daey> the project i linked is done. mine hasnt even started. wanting to try to resolve the glaring issues first
[18:24:32] <cehteh> me thinks it would be more fun to have more capabilities on the mpu :D
[18:25:28] <cehteh> which avr where you thinking about 328p?
[18:27:24] <daey> i dont really care. it only needs to fullfill a few criterias. small and enough pins
[18:27:54] <daey> everything with 8kb should be sufficient
[18:28:41] <daey> atmel even has a app note on driving LCDs with gpios :D http://www.atmel.com/Images/doc2569.pdf
[18:29:36] <cehteh> i like fig.1 :D
[18:31:06] <cehteh> where do they say anything about voltages?
[18:31:25] <daey> its on the project site. it claims that many displays need 5V
[18:31:33] <daey> otherwise the contrast is terrible
[18:31:52] <cehteh> i meant in the atmel appnote
[18:32:20] <cehteh> they just switch the outputs .. fine .. but at 3.3v or whatever you do (did you consider to go for even less?)
[18:34:45] <cehteh> at 2.7v you can still go for 8mhz ... but i think avrs on a wristwatch stretch the limit a bit, you need to be very power conservative
[18:34:57] <cehteh> you plan for a external RTC chip?
[18:35:19] <daey> cehteh: no a quarz, just like the project i linked
[18:35:52] <cehteh> keeping avr's oscillators running will draw much more juice from the battery on a avr than on a rtc chip
[18:36:07] <daey> it needs to run anyways because it has to drive the lcd constantly
[18:36:23] <cehteh> that would be horrible
[18:36:33] <daey> thats what the casio clocks do as well
[18:36:54] <daey> the display needs a 30ish hz ac input
[18:37:02] <cehteh> yes but they have special circruits designed for extremely low power
[18:37:10] <cehteh> AC or pulse?
[18:37:24] <daey> well pulses
[18:37:35] <cehteh> either way you dont want to generate that with the mpu
[18:37:47] <cehteh> that would be way to wasteful
[18:37:50] <daey> but the display sees Com input vs another pin input which will result in ac
[18:38:18] <cehteh> keep the mpu sleeping most of the time
[18:38:24] <cehteh> look at the datasheets
[18:38:31] <cehteh> see how much power you have in your battery
[18:38:40] <daey> how do i keep the input going when its sleeping?
[18:38:41] <Lambda_Aurigae> how does a real watch do it?
[18:39:11] <daey> they probably have a tiny circuit that gets an input and keeps hammering what it got last time
[18:39:18] <daey> the rest sleeps
[18:39:22] <cehteh> first of all they have a 32k quarz
[18:39:28] <daey> so do i
[18:39:48] <cehteh> ok
[18:40:12] <Lambda_Aurigae> running the chip at 32KHz or just using it for the rtc input?
[18:40:41] <cehteh> i suppose they have distinguished functional blocks for things like LCD referesh, RTC calculations and the rest (alarm, display/font mapping etc)
[18:40:46] <daey> i havent found a reason not running it at 32kHz
[18:41:10] <daey> i dont really need anything fast. the inputs need to do 30ish hz pulsing
[18:41:12] <cehteh> "run to idle"
[18:41:26] <daey> outputs*
[18:41:40] <cehteh> aka the faster you can do something the earlier you can go to sleep
[18:41:47] <daey> the project uses an msp430 he gets a battery lifetime of roughly a year
[18:42:03] <daey> ah you mean sleep between the driving?
[18:42:07] <cehteh> with what battery? AA?
[18:42:10] <cehteh> yes
[18:42:11] <daey> aa lol
[18:43:17] <cehteh> you want to sleep as much as possible and clock as fast as you can w/o drawing too much current and wake as little as possible
[18:44:02] <cehteh> AVR's are not that bad when programmed carefully, but running from a button cell for a year could become a challenge on a watch
[18:44:25] <cehteh> (some people make devices which wake far less and dont drive displays, there it works)
[18:44:46] <cehteh> cr2032 cell has plenty juice if used carefully
[18:45:22] <daey> its crazy how long these casio watches run
[18:45:32] <cehteh> but keep the mpu running or even waking at 30Hz .. i have a gut feeling thats a nogo
[18:45:33] <daey> i think i had one that lasted >10years
[18:46:16] <Lambda_Aurigae> casio watches have custom built asic devices.
[18:46:35] <daey> i know. im not trying to get anywhere near that value
[18:46:47] <cehteh> and there is a big difference on AVR's between power-down current and idle current
[18:47:29] <cehteh> so you want a RTC chip, which keeps the time and sends a INT0 to the AVR to wake it whenever anything is to be done
[18:47:41] <cehteh> (once a second to flip the LCD display)
[18:48:01] <cehteh> the LCD itself should not need the AVR to be running
[18:48:13] <daey> again thats not enough. the uC is the LCD driver, not just a data to display driver thing
[18:48:28] <cehteh> yes that idea is flawed
[18:48:37] <cehteh> it may work with a AA cell
[18:48:47] <cehteh> but not on a small button cell
[18:48:48] <daey> ? it works fine
[18:48:54] <daey> https://github.com/carrotIndustries/pluto/raw/master/photos/pcb.jpg
[18:48:58] <daey> see?
[18:49:24] <cehteh> you saied thats a msp430 .. i dunno that
[18:49:45] <daey> i dont see the difference.
[18:49:50] <Lambda_Aurigae> looks like one.
[18:49:52] <daey> the fundamental problem seems to be the same
[18:50:05] <cehteh> well msp where made for extremely low current
[18:50:10] <Lambda_Aurigae> msp430 is made to run much lower current consumption than an avr
[18:50:15] <daey> ok
[18:50:35] <cehteh> i never used one, but thats pretty much i know about it
[18:50:55] <Lambda_Aurigae> they aren't bad chips overall.
[18:51:04] <Lambda_Aurigae> bit more difficult to use than an avr.
[18:51:21] <Lambda_Aurigae> fewer features overall from my experience.
[18:51:28] * cehteh just reads wikipedia .. they have buildin RTC
[18:51:35] <Lambda_Aurigae> some do, yes.
[18:51:51] <cehteh> there is a gcc
[18:52:02] <Lambda_Aurigae> I think I saw a pic that had built in rtc too.
[18:52:26] <Lambda_Aurigae> different chips made for different things.
[18:52:31] <cehteh> yeah
[18:52:34] <daey> his has rtc
[18:52:44] <daey> its essentially the perfect chip for the task
[18:52:51] <cehteh> i think AVRs are pretty nice, but they wont excel at wristwatch
[18:52:59] <Lambda_Aurigae> and that rtc runs very low current and async to the cpu.
[18:53:23] <Lambda_Aurigae> with an avr, you have to have the cpu handle the time tick updates.
[18:53:25] <cehteh> and much much more precise
[18:53:52] <cehteh> just look at the datasheet of this common DS rtc what was its name?
[18:54:05] <cehteh> things you can never do with AVR's
[18:54:15] <Lambda_Aurigae> ds1306 is what I use
[18:54:22] <cehteh> i tihnk that was it
[18:54:25] <cehteh> ultra low power
[18:54:43] <cehteh> temperature and aging compensation for the crystal
[18:54:54] <daey> hm the msp430 does look nice indeed. ti even has official gcc support.
[18:55:02] <cehteh> yes
[18:55:06] <Lambda_Aurigae> yup.
[18:55:18] <Lambda_Aurigae> it didn't exist when I first tried msp430
[18:55:32] <cehteh> like i saied, if you want to build a clock with a AA cell for your desk, AVR's are ok
[18:55:35] <theBear> heh, "i'm slowing down a lot now i'm in my sunset years..." -- (uncompensated) aged crystal
[18:55:46] <cehteh> but wristwatch, i think that stretches the limit
[18:56:10] <daey> i wonder how avr argues the extra power consumption
[18:56:25] <Lambda_Aurigae> they aren't made for such.
[18:56:26] <daey> to me they look the same feature wise on the first look.
[18:56:27] <cehteh> different market
[18:56:35] <theBear> sorry delay, but for ref. the inductors i described are a bit smaller than the crystal in that watch
[18:56:38] <daey> well. i doubt the other market wouldnt like less power consumption
[18:56:49] <cehteh> when you look at your pc's motherboard you pretty sure find some AVR's for example
[18:57:05] <Lambda_Aurigae> I've seen them in xerox copiers.
[18:57:13] <cehteh> and the few mA the AVR#s suck there doesnt matter
[18:57:17] <cehteh> yes
[18:57:28] <Lambda_Aurigae> ripped a couple out of them in fact.
[18:57:56] <cehteh> most clocks are ASIC's .. cheap mass production makes it possible
[18:58:17] <cehteh> you dont need a custom programmable mpu with tons of features there
[18:59:02] <cehteh> they certainly have some mpu core, but only the feature groups it really needs
[18:59:07] <daey> ofc they are asics. casio probably sells more watches then atmel atmegas :P
[18:59:18] <cehteh> and i wont surpused if some are 8051 :D
[18:59:28] <Lambda_Aurigae> very possibly.
[18:59:33] <Lambda_Aurigae> or even something closer to a 4004
[18:59:46] <daey> someone said they are 4bit masked roms
[18:59:59] <Lambda_Aurigae> could be a simple finite state machine.
[19:00:05] <cehteh> or .. casio ... had some 4bit stuff ... BCD calculation ftw
[19:00:06] <daey> yeah thats what i think
[19:00:47] <daey> mine has an interesting bug though
[19:01:08] <daey> the backlight is on for 5s. if you hit the menu button afterwards its instantly off :P
[19:01:12] <cehteh> anyway if i would build a watch i would like it to be somewhat more capable .. going for 32bit arm would be cool
[19:01:39] <daey> at which point you run into the hardware/packaging issue
[19:01:47] <cehteh> not really
[19:01:52] <daey> where do you put it in?
[19:02:00] <cehteh> you seen the kinetis i posted earlier
[19:02:04] <Lambda_Aurigae> low end arm is smaller than that msp430
[19:02:07] <cehteh> thats much smaller than AVR's
[19:02:16] <cehteh> 2x2mm
[19:02:24] <cehteh> with 20 pins (err balls)
[19:02:31] <Lambda_Aurigae> hell, mid range arm can be smaller than that msp430...and often come in a PGA package.
[19:02:43] <cehteh> yes
[19:02:56] <cehteh> and sometimes they are even cheaper :/
[19:03:00] <Lambda_Aurigae> lots more pins in a smaller package.
[19:03:07] <daey> cehteh: size is only part of it. if you have a 32bit arm, you got processing power. what do you do with it? you need a proper display.
[19:04:02] <daey> at that point you would have to buy a smartwatch to have a proper platform
[19:04:04] <cehteh> the only thing what bugs hobbyists are that they are hard to obtain, imposible to solder and sometimes have some strange electric things like dual voltage (1.8v/3.3v) weak drive strength, not 5v tolerant etc
[19:04:19] <cehteh> yes
[19:04:29] <cehteh> well its not my project :)
[19:04:44] <cehteh> what cpu is used by pebble?
[19:05:02] <daey> but the smartwatch has immense power drain due to the display. and natively supports java. so you would be gimping yourself with rebuilding one
[19:05:42] <cehteh> daey: pebble use eink
[19:06:02] <cehteh> i am not speaking about these apple or samsung crap
[19:06:22] <daey> y eink is the only proper way imo
[19:06:31] <cehteh> charge your watch once a day .. hey what .. wait wtf
[19:06:40] <daey> :D
[19:06:41] <cehteh> lcd could be fine to
[19:06:56] <daey> e-ink looks nicer
[19:07:06] <cehteh> yes
[19:13:24] <carabia> or you can do what everyone does and go with the sharp pseudo-eink
[19:13:51] <daey> i dont see how its feasable to build a case
[19:14:02] <daey> not waterproof terrible looking etc.
[19:14:30] <daey> so using a ready to buy watch as the platform is mandatory imho
[19:14:44] <carabia> it's not
[19:14:54] <daey> ?
[19:15:19] <carabia> there was just recently a watch that was really small in hackaday
[19:15:26] <cehteh> building a case would be simple for me :D
[19:15:37] <daey> if you have a machineshop yeah :P
[19:15:39] <cehteh> hey carbon fiber case :)
[19:15:48] <cehteh> even handworking
[19:15:48] <carabia> daey: it was 3d printed afaik
[19:16:01] <cehteh> 3dprint may work to
[19:16:42] <cehteh> https://twitter.com/CehtehCt << lots of manual labor to build copter frame (and propellers below)
[19:16:52] <carabia> you could probably mill a frame, too
[19:16:55] <cehteh> http://www.pipapo.org/cfprops/cfprops.html
[19:17:00] <carabia> well, i'm sure you could
[19:17:02] <cehteh> i wnat a mill yes
[19:17:26] <daey> yeah if you had a mill/lathe you could do some neat things
[19:17:32] <carabia> meh, copters are so fucking boring
[19:17:34] <cehteh> but with some patience and experience one can do a lot things with handwork
[19:18:00] <cehteh> daey: i dont have either, just a big wood router
[19:18:07] <carabia> daey: i'm sure hackerspaces can let you use their tools for a fee or so
[19:18:28] <cehteh> that too
[19:18:33] <carabia> as i think this wasn't for production
[19:18:58] <carabia> "hackerspace" generally means pull out your copters and arduinos
[19:19:02] <daey> xD
[19:19:06] <carabia> jesus i fucking hate both.
[19:19:20] <cehteh> lol i never visited the local hackerspace here
[19:19:43] * daey lets an arduino powered copter fly infront of carabia window
[19:19:52] <cehteh> copters are fun when you build them yourself, not just screwing premade things together
[19:20:37] <daey> yes the kalman filter looks interesting
[19:20:38] <carabia> daey: i'd greet it with a shotgun
[19:20:43] <daey> D:
[19:20:49] <carabia> copters are not fucking fun in any way shape or form
[19:20:58] <daey> i agree
[19:20:59] <cehteh> well thats your opinion
[19:21:01] <carabia> and it's not even flying
[19:21:24] <carabia> every man and their dog has a copter and a camera strapped to it, and youtube is so god damn full of the footage
[19:21:35] <cehteh> yes that sux
[19:21:48] <daey> and all you hear in those videos is: BSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
[19:21:52] <carabia> with some god damn downtempo song to it and pretending it's somehow different from the rest of it
[19:22:48] <carabia> i need to build myself a super high powered transmitter i'll put on the roof to transmit garbage on the frequencies they use for video
[19:23:05] <daey> with garbage you mean porn
[19:23:09] <carabia> there aren't too many channels as they all use pretty much the same modules
[19:23:29] <daey> tbh. ive never seen a copter irl :D
[19:23:41] <carabia> probably, hardcore sweaty truckdriver-on-truckdriver gay action.
[19:24:33] <cehteh> not?
[19:24:35] <carabia> i have, quite a few, too
[19:24:43] <carabia> copters, seen, irl ... that is.
[19:24:55] <carabia> but i fly /real/ stuff, so perhaps that's why.
[19:25:12] <daey> yeah im sure if i go to our local rc flying club they have lots of those things
[19:25:22] <cehteh> where are you?
[19:25:27] <daey> but noone here seems to be in RC stuff outside of the club area
[19:25:30] <daey> germany
[19:25:35] <daey> NW
[19:25:43] <cehteh> karlsruhe here :D
[19:25:50] <daey> muenster :)
[19:26:07] <cehteh> https://www.youtube.com/watch?v=DjSRDeXeGvI << very good german pilot :D
[19:27:10] <daey> i love this one https://www.youtube.com/watch?v=_oBuqhEg0xU
[19:27:10] <carabia> oh please not even gonna bother
[19:27:20] <daey> carabia: mine is non quadcopter :P
[19:27:38] <carabia> i really don't give two fucks, if it has a pid controller, it's not real flying.
[19:27:48] <daey> it doenst
[19:27:49] <cehteh> whatever
[19:28:07] <carabia> daey: oh yes? what is it
[19:28:09] <daey> it doesnt even have a motor :P
[19:28:09] <cehteh> a friend of mine flies big shiut ..
[19:28:24] <carabia> oh. a glider?
[19:28:26] <daey> if it has a motor its not real flying :D
[19:28:34] <cehteh> his saying is that he only flies models he can trow himself .. thats close to 25kg :D
[19:29:08] <carabia> daey: well, no. DLGs are pretty cool though, i've got a cf one buried somewhere in the garage
[19:29:32] <daey> ive never flown any RC model
[19:29:50] <carabia> oh. in that case your opinions are ignored hereafter
[19:29:58] <daey> ?
[19:30:14] <daey> i ahvent written any opinions
[19:30:21] <carabia> yes you have
[19:30:24] <daey> no
[19:30:46] <daey> at least non that require you to fly something
[19:30:49] <carabia> cehteh: so, he has glow or gas?
[19:30:57] <cehteh> glider
[19:31:22] <carabia> 25 kg hand-launch? ok then
[19:31:28] <carabia> ...
[19:32:16] <daey> well the pebble is completely useless...its not water resistant :D
[19:32:36] <cehteh> i am just searching his vids
[19:32:36] <carabia> cehteh: I somehow doubt that, but w/e
[19:33:22] <daey> 25 does sound kinda beasty :P
[19:33:47] <carabia> i'm quite sure it's not 25 and nowhere near it.
[19:34:21] <carabia> you'd basically be handlaunching something heavier than a bench press bar.
[19:34:34] <carabia> and the wingspan on a 25kg glider would be something... phenomenal
[19:34:36] <daey> from what i can tell. the bigger motorized ones are self launching. the others use catapults even for smaller stuff
[19:34:52] <cehteh> 6m or so
[19:35:08] <daey> thats almost real size wing span o0
[19:35:23] <cehteh> he told he he has a for for a 9m discus
[19:35:32] <cehteh> well he is not small :)
[19:35:33] <carabia> you either have a motor, a catapult, a tow (tug?) or a DLG
[19:35:49] <daey> discus as in discus 2 clone?
[19:37:23] <cehteh> https://vimeo.com/50906495
[19:37:30] <carabia> well, even something like a ~4m wingspan is waaay under 3 kg.
[19:37:59] <carabia> well yeah, but that's not DLG.
[19:38:03] <daey> nice view
[19:38:04] <carabia> i'd like to see him dlg'ing that.
[19:38:17] <cehteh> i didnt say DLG
[19:38:26] <carabia> he told he he has a for for a 9m discus
[19:38:27] <carabia> video?
[19:38:44] <carabia> and yes, that's "hand-launch"
[19:38:53] <cehteh> form .. not yet build, mold for GFK building
[19:39:00] <cehteh> https://vimeo.com/73089906
[19:39:12] <cehteh> and yes that are smaller ones .. but thats his channel
[19:39:27] <cehteh> just look if you find him a launching a bigger ones
[19:39:33] <cehteh> with some comments about weight
[19:39:44] <carabia> I get it, yes, you can push a 25kg glider into a strong headwind
[19:40:01] <carabia> if it's anything a sane person would do? no
[19:40:17] <daey> well given the right spot they kinda launch themselves :P
[19:40:31] <daey> like the first one. or as shown in my video
[19:40:31] <carabia> you just give it a lil' push
[19:40:48] <cehteh> https://vimeo.com/27965588
[19:40:52] <carabia> and you need a specific place to fly those in any case.
[19:41:06] <cehteh> of course .. you need excelent wind to launch such big planes from hand
[19:41:09] <daey> not a single hill in sight here T_T
[19:41:26] <cehteh> the later is my favorite :D
[19:41:27] <daey> only thing you could try is thermic gliding
[19:41:47] <carabia> well you need a catapult or a tug
[19:41:56] <daey> yea
[19:41:59] <carabia> well you could do dlg too
[19:42:01] <daey> and hope for good weather
[19:42:12] <carabia> you can have dlgs up to very long wingspans
[19:42:31] <cehteh> not that much
[19:42:33] <carabia> and they're featherweight, doesn't need a downhill such as those
[19:42:47] <carabia> I've seen ~4m dlgs
[19:42:48] <cehteh> anyway DLG's are cool too
[19:42:50] <carabia> which are huge.
[19:42:53] <cehteh> yes
[19:42:56] <daey> i hope the fpv cams get better. i went glider flying once and it was frickn awesome. but its too much effort in a club
[19:42:59] <carabia> given you /handlaunch/ them
[19:43:15] <cehteh> and catch on landing
[19:43:30] <carabia> well, if you're good ey
[19:43:53] <cehteh> bad landing breaks the rudder on a dlg
[19:44:17] <carabia> depends on the dlg, and generally bad landing breaks something with everything
[19:44:41] <daey> theres this official rc glider class. iirc. 2m span width. and no ruder. that looks pretty interesting for my region
[19:44:42] <cehteh> high performance DLG's are like eggshells ..
[19:44:47] <daey> forgot the name of the class thoughh
[19:45:00] <cehteh> wont break in the air, but one wrong touch or handling
[19:45:04] <carabia> of course, but high performance dlgs are also something not a whole lot of people can afford
[19:45:13] <cehteh> RED
[19:45:14] <carabia> well, or, not willing to spend.
[19:45:15] <cehteh> RES
[19:45:19] <daey> yeah
[19:45:27] <daey> im not a fan of these real world plane replicas
[19:45:37] <carabia> and by the time you've got kilos to drop on dlgs you're not a rookie.
[19:45:38] <cehteh> 2m is pretty soon out of sight on a glider :D
[19:46:16] <carabia> daey: huh? scale flying?
[19:46:24] <cehteh> http://www.ig-nurfluegel.ch/IMG_6149.JPG .. thats awesome
[19:46:33] <daey> https://www.youtube.com/watch?v=AZSEqUdv_IU this one looks so gud
[19:47:06] <daey> cehteh: yeah thats something i like :) models should push boundaries, not copy flaws
[19:47:45] <carabia> daey: not really
[19:47:56] <cehteh> reports are it flied bad :/ .. but horten flying wings have shitload of complex aerodynamics
[19:48:04] <daey> :(
[19:48:08] <carabia> "flaws" is a loaded term, and also some people do enjoy scale-flying.
[19:48:10] <cehteh> but it looks awesome
[19:48:29] <daey> carabia: sure depends on your goals. if you are into real model stuff i can understand
[19:48:51] <cehteh> well .. i am off
[19:48:52] <daey> i just find it odd that people actually put figures into the cockpit ^_^
[19:49:19] <daey> good night
[19:49:23] <carabia> also some designs are just purely amazing. Such as the sbach, which is actually german.
[19:49:30] <carabia> at least the nazis did something right
[19:49:37] <carabia> even if they didn't quite finish with the jews.
[19:49:44] <carabia> well, post-nazis.
[19:50:11] <daey> its just a model doesnt benefit from a cabine.
[19:50:36] <carabia> actually, in some cases it does.
[19:51:03] <carabia> such as flying knife-edge, it provides a (albeit small) stabilizing effect
[19:51:17] <daey> well sure if you just remove it on a normal modell it would break the aerodynamic 
[19:51:41] <carabia> not to consider that, too
[19:52:28] <daey> my point is just, a cabine is there because a planes main purpose is to carry a person. its whole design is wrapped around that concept. this is not true for a model.
[19:52:53] <carabia> sure, but there are still scenarios where it helps.
[19:53:41] <carabia> arguably you could just put a low-profile stabilizer there, but they really are the two same things. And in my opinion an airplane should look like an airplane!
[19:54:04] <daey> thats what i mean. some people like those reallife models, i dont really
[19:54:42] <daey> thats why i find the RES concept more appealing than these DLG's.
[19:55:34] <carabia> dlgs do not have any resemblance to any full-size aircraft.
[19:56:06] <daey> oh. my bad
[19:56:20] <daey> i hadnt heard that term before, and i assumed its a term for these discus clones
[19:56:24] <carabia> apart from the fact that they have a wing, stabilizers and spoilers
[19:56:43] <carabia> dlg, discus-launch glider
[19:56:50] <daey> :P
[19:56:53] <carabia> do you actually know how athletes throw discuses?
[19:56:55] <daey> i see
[19:56:56] <daey> yeah
[19:56:59] <carabia> dlgs are thrown the same way.
[19:57:06] <daey> i thought discus as in Discus the plane
[19:57:12] <daey> yeah i know how they do it
[19:57:22] <carabia> there's a grip in the wingtip, you spin and release.
[19:57:38] <carabia> that's a dlg.
[19:57:43] <daey> Discus is also the name of a real glider :)
[19:57:55] <carabia> probably is.
[19:58:14] <daey> a mate crashed on :P
[19:58:26] <carabia> http://rcexplorer.se/wp-content/uploads/2012/09/f3k2-2.jpg
[19:58:28] <carabia> this is a dlg.
[19:58:48] <daey> yeah those are nice as well
[19:58:55] <carabia> more precisely it's a f3k-class dlg
[20:00:33] <daey> is it doable to strap an fpv cam to those?
[20:00:37] <daey> a*
[20:00:49] <carabia> to a dlg? no
[20:01:05] <carabia> well, if you go big enough, you can
[20:01:19] <daey> well i dobut the launch type makes a difference here? rather the plane size
[20:01:45] <carabia> well there's a limitation how heavy of a plane you actually can discus...
[20:02:08] <daey> ive no idea how much weight is added by a fpv setup, but it shouldnt be too much?
[20:02:09] <carabia> but you need to go bigger than an f3k, because it's not just about strapping the camera, you need a tx aswell.
[20:02:19] <daey> the cam is tiny. the battery can be small as well.
[20:03:02] <daey> yeah i figured it wouldnt be so simple :(
[20:03:05] <carabia> depends on the tx power, for a glider you can get away with something low power though, cause you'd mostly be flying on an open area
[20:03:42] <daey> yes that was my thought. you wont fly far away. and the flights are short so swapping the battery a few time isnt much of an issue
[20:04:17] <carabia> maybe you could actually strap an ultra light fpv setup to an f3k...
[20:04:37] <carabia> if you can just find a spot for the video tx, as you probably don't want it next to the servos
[20:04:52] <carabia> actually yeah, it's doable.
[20:05:11] <carabia> usually you put some ballast in the gliders to set the CG anyway
[20:05:33] <carabia> and you can get as light as 12g 25mW txs
[20:06:00] <carabia> so ye, doable. you probably want someone else launching it for you though :D
[20:06:20] <daey> yeah it will certainly fly worse. :P
[20:07:01] <carabia> not necessarily
[20:07:14] <carabia> as i said gliders usually have ballast in them to set the CG properly
[20:07:24] <carabia> or cog, center-of-gravity
[20:07:53] <daey> i assumed that you would still end up with more than you would usually have. plus the camera aerodynamic problem, you probably cant fully integrate it into the main body etc.
[20:08:00] <carabia> usually they are a bit too back-weighted so you put the ballast in front, and i guess you can just carve out a hole for the camera and shove the tx in there somewhere...
[20:08:30] <carabia> i can't imagine why you couldn't, just drill a hole and glue it in place
[20:08:41] <carabia> you can get super tiny cmos cameras.
[20:09:13] <carabia> obviously you can't do a hard nose-landing or camera goes R.I.P, but so would your plane probably.
[20:10:42] <carabia> you wouldn't probably have toooooo much flight time as the batteries are miniscule
[20:11:42] <carabia> you'd probably need a boost reg in there to provide the tx with 5 volts
[20:11:55] <carabia> but 25 mW at 5 volts is ridiculously low current
[20:12:03] <daey> 2x18650 under the wings :P
[20:12:50] <carabia> but i don't have an f3k, i'm not sure whether they fly 1s or 2s
[20:12:50] <daey> or 4 detachable when empty :D
[20:13:51] <daey> main reason i didn't dive into the hobby is missing space. building them is part of the hobby, and i dont have the space for a workshop atm
[20:13:53] <carabia> but you'd probably want a buddy to throw the plane as I think you'd get a bit disoriented spinning around with the plane having the glasses on
[20:14:58] <carabia> luckily i've crashed enough planes that i don't have too many to store. :)
[20:15:04] <daey> lol
[20:15:26] <carabia> one epp-3d, one balsa-3d, and the glider
[20:17:13] <daey> the motorized stuff doesnt really interest me. except maybe for the turbine ones (which i def. will never get into due to the crazy prices)
[20:17:26] <daey> the pulse jet ones are hilarious as well
[20:17:40] <carabia> jesus fuck
[20:17:50] <carabia> pulse jets make a lot of sound for nothing.
[20:17:58] <daey> yeah the infernal noise
[20:17:59] <daey> awesome :D
[20:18:06] <carabia> no it sure isn't
[20:18:08] <carabia> ew
[20:18:40] <daey> :D
[20:18:54] <carabia> and the fact that a valve pulse jet can run for so long before the valve goes bust.
[20:19:03] <carabia> +only somewhere in there
[20:19:22] <daey> https://www.youtube.com/watch?v=r50DRou0LsM its like a motorcycle with wings *.*
[20:20:04] <daey> oh and these highspeed helis are interesting as well
[20:20:08] <daey> but money :D
[20:20:25] <daey> https://www.youtube.com/watch?v=myxp7e1J-1o these ones
[20:20:27] <carabia> pulsejets are not expensive
[20:20:32] <carabia> but they suck.
[20:20:39] <daey> i know. but no thanks :D
[20:20:54] <daey> i dont think you are welcome anywhere with those things
[20:21:02] <carabia> Pylons go way faster than that.
[20:21:13] <carabia> well, pylon racers
[20:21:25] <daey> faster than what?
[20:21:33] <carabia> than that heli.
[20:21:38] <daey> ah yeah obviously
[20:21:40] <carabia> they warm up the batteries ;D
[20:21:48] <daey> but its not a helicopter
[20:22:02] <carabia> sure isn't, but it's fast
[20:22:38] <daey> you cant surpass the soundbarrier with the blades, so helis cant go all that fast
[20:22:48] <carabia> it's kinda like when referring to speed you can make a tank that goes 40mph or you can make a formula that does 200
[20:23:28] <theBear> hmmm, never considered that, but i spose a prop ain't much use in a land beyond sonic barriers and largely devoid of nice firm air to push on :)
[20:23:51] <carabia> you think? :D
[20:24:49] <theBear> i pretty sure i do :-) either way, at least i'm trying to think <grin>
[20:25:34] <theBear> on the other hand, imagine how fat and unwieldly the trailing edge of the blades could afford to be at that speed :)
[20:26:49] <carabia> turbines sure are interesting, but there's so much fiddling around with them.
[20:27:20] <carabia> you could go for an edf which is basically a turbofan, just without the turbine :)
[20:27:31] <daey> yeah its something one likes to watch but doesnt want to own
[20:27:45] <carabia> but i guess you want a turbojet rather.
[20:28:00] <carabia> which is wholly inefficient and difficult.
[20:28:29] <theBear> heh
[20:28:44] <carabia> never been drawn into turbines that much, the way i see models is that they shouldn't be overly complicated or big.
[20:29:06] <carabia> nevertheless, for real-size applications they are quite interesting.
[20:30:29] <theBear> they sure look simple, on paper... i never checked out how accurate that appearance is by building one in the kitchen yet :)
[20:30:33] <carabia> that said, now i'm actually intrigued if anyone's ever built an rc turbojet with an afterburner.
[20:30:47] <daey> afaik there are no working ones
[20:30:50] <daey> only fake fire
[20:31:01] <daey> there was an explanation why it isnt feasable, but i forgot it
[20:31:20] <theBear> aren't we just talking about jet/turbine engines generally the last few lines ?
[20:31:35] <daey> anyhow im off as well
[20:31:37] <daey> gn8
[20:31:41] <theBear> or have i ignored a word cos i don't know what it means in context, and am very far from what you guys thinking :)
[20:31:42] <carabia> nighty night
[20:31:44] <theBear> cya daey
[20:31:50] <carabia> theBear: huh?
[20:31:59] <theBear> heh, that's a fun name to try pronouncing
[20:32:06] <carabia> oh, no, we were talking about rc models actually
[20:33:06] <theBear> oh, he was saying no working ones to you (turbojet with afterburner ...)
[20:33:38] * theBear checks a diagram before he says something about a completely not-called-turbojet engine ....
[20:35:05] <carabia> it does not necessarily have an afterburner.
[20:35:06] <theBear> mmm, that's the kinda jet i thought it was at a glance... i'm FAR from a rocket scientist, but i can't think of some reason a tiny model-sized one couldn't theoretically and even practically be afterburned
[20:35:47] <theBear> indeed, i didn't think anything did ALWAYS have a afterburner, thus it getting a name for itself instead of just being part of the jet insides
[20:36:03] <theBear> maybe that and being optional
[20:36:12] <carabia> i'm lost, but okay
[20:37:26] <carabia> turbojet is what people refer to as "jet engines", usually.
[20:37:43] <theBear> heh don't feel bad, i'm not sure i'm not-lost, but i'm saying a lot on the assumption that i'm not <grin> short/100%non-fiction version: i wonder why scale versions of such a jet would not scale/work in the afterburner dept
[20:38:10] <theBear> yeah, it's what i always thought of as a "jet engine" cos when i first looked 'em up it was in relation to "jet planes" :)
[20:38:10] <carabia> I'm not sure, but an afterburner is superduperhyper inefficient anyway
[20:38:20] <carabia> as you're just pissing off fuel
[20:39:29] <theBear> hmmm, i thought they were supposed to burn unburnt fuel already in the air/jet flow from earlier in the engine, get more than nothing outta the already spent fuel in exhchange just for someone lighting them to get em burning
[20:39:47] <theBear> like i say, not a rocket scientist, and that applies whether i thought right or not :)
[20:41:20] <carabia> i'm not an afterburner expert, but I was under the impression that extra fuel is pumped to the exhaust.
[20:41:56] <carabia> because i'm not completely sure what you mean "earlier in the engine"?
[20:42:17] <carabia> the fuel is burned in the combustion chamber, which is by definition at the back of the engine.
[20:43:13] <theBear> interesting... at least one of us is bound to be right, you'd think <grin>
[20:44:42] <carabia> the "earlier" part is pretty much the compressor.
[20:44:47] <theBear> aww, i misplaced the diagram i was using for memory/ref. that i was thinkin the right jet type, meh
[20:45:36] <carabia> and there's no flueflow there.
[20:46:44] <theBear> flue in this context = air passing from front end to business end ?
[20:46:49] <theBear> flueflow
[20:46:54] <carabia> i'm not familiar with a jet engine that recycles fuel twice ever, but maybe it's an opportunity for me to learn
[20:47:28] <carabia> fuel flow as in with a turbojet, fuel gets injected into the combustion chamber, it burns there, and exhausts.
[20:47:39] <carabia> afterburner = extra fuel is pumped into the exhaust
[20:50:27] <theBear> hmmm, either lack of understanding needed background info LONG ago made me misundserstand how afterburners work, and/or memory switched unburned/used oxygen from atmosphere/intake air with unburned fuel in my head, and finally (oh, you were very right re: lots of extra fuel inefficiently burned for a little boost in power btw) wow ! the afterburner thing is kinda sneaky with the gas vapor pressures and
[20:50:27] <theBear> temps and stuff being heavily involved, and it not just being a bit of a "rocket" added on the back of the engine
[20:51:17] <theBear> carabia, few minutes back i was asking clumsily "what does flueflow mean in this context ?"
[20:51:36] <carabia> well i meant wherever the fuel goes.
[20:53:18] <theBear> gotcha, meh, i'm somewhere between satisfied-curiosity/bored with jets now/and just plain unable to even comment any more due to needing a full detailed theory refresher to go any further, and it time for ya know, gettin the house a big 1st-world grade again :)
[20:54:13] <theBear> thanks fer provoking me to at least have a mini-refresh on afterburners and how they do tho, certainly can't be a bad thing to be less-uninformed about the whole situation
[20:54:29] <carabia> well that's why they're called afterburners...
[20:57:30] <carabia> but even then, most jet engines that are used are not turbojets, they're turbofans rather
[20:58:11] <carabia> which is basically a turbine driving a fan
[20:58:43] <carabia> you can have an afterburner in a turbofan too, of course
[20:59:53] <theBear> yeah, i caught a pic of a old turbofan earlier while scanning for a diagram/reminder and all that came back to me ... plus all the old non-turbo stuff like why you want/(kinda)"need" something like the old soup-er-charger cos of air is different way up high etc etc
[21:00:36] <carabia> well even then, an afterburner for a turbofan is no different.
[21:00:44] <theBear> plus i always kinda liked the inside-out motor approach of traditional prop engines, kinda like the combustion version of those modern 3phase rc "outrunner" insideout motors, for the same rteasons i guess
[21:00:56] <theBear> oh yeah, not suggesting it is, just idly commenting :)
[21:03:19] <carabia> but "trad. prop engines" are piston...
[21:03:29] <carabia> unless turboprops
[21:05:33] <theBear> oh yeah, i mean piston... like i said, idly commenting
[21:05:59] <theBear> modern rc 3 phase brushless aren't even combustion engines, but i slipped them in there too :)
[21:06:14] <carabia> well obviously they are electric \:D/
[21:06:39] <carabia> anyway i gotta run, was good talkin to ya!
[21:08:10] <theBear> likewise, enjoy yer day !
[22:44:23] <Casper> hey rue_house/rue_shop4/rue_bed2: I beleive you said you made some gearbox in the past?
[22:45:24] <theBear> he's got a webpage about how to diy from a pile of old printer even !
[22:45:53] <theBear> and he's "around" a few mins ago, be virtuous
[22:46:39] <rue_house> yup
[22:46:57] <rue_house> http://ruemohr.org/~ircjunk/tutorials/mech/gearbox/gearbox.html
[22:47:17] <Casper> ok... so you have no good source of gears...
[22:47:26] * rue_house shudders to think how much space that box of gears would take up if he dumped it today
[22:47:38] <rue_house> I have about 3x as many as I did
[22:47:45] <rue_house> those are good gears
[22:47:56] <rue_house> just not consistant
[22:49:18] <Casper> rue_house: here's what I need to do: make some vent 'valve', I'm making an air exchanger and need to valve the air... so I have a few idea: a folding design 'V' and a sliding door... Both would require some gearing of some sort...
[22:49:39] <Casper> but... kinda a pain to figure out where to get the stuff :(
[22:49:43] <theBear> hey, i like to think that like myself making something perfectly available pre/pro made, rue doesn't do such things because he NEEDS to, financially or any other reasons, but cos he can <grin> so ya know, no jumping or concluding too close to one another, take a break between so they don't mix
[22:49:56] <Casper> do you have a better idea or suggestion?
[22:50:58] <theBear> Casper, the umm, is a single oval/one-piece one-pivot butterfly like you see on silly car engine super blower intakes not simpler and equally effective in that kinda situation ]
[22:51:30] <Casper> 4 out of 7 need to be able to go from 0 to 100%
[22:51:35] <rue_house> in commercial stuff they use a heavily geared motor with a return spring
[22:51:42] <Casper> the 3 others can be 10-100%
[22:51:57] <rue_house> they can use pwm (slow) to hold the gates partially open
[22:52:01] <Casper> and need to actuate in about 10 seconds
[22:52:14] <rue_house> prettymuch timer motors
[22:52:29] <rue_house> single direction
[22:54:32] <Casper> yeah that would have been somewhat ideal, but pricey
[22:55:46] <rue_house> you could do it with a small stepper
[22:55:59] <rue_house> have the control line operate a forward/reverse line
[22:56:15] <rue_house> step slow, and make sure the stepper cn nicely quietly slip when it hits the limit
[22:56:26] <Casper> I calculated that I'ld need like 16ozin of torque...
[22:56:33] <theBear> isn't a single vane (vein?) butterfly like i described able to do 0-100%, or at least very close to 0 to 100 ?
[22:56:34] <Casper> for a 4"
[22:56:48] <Casper> close, but not 0%
[22:58:34] <theBear> fair deuce... interest passing, but before i forget alltogether, watcha trying to let/un-let flow with such absolute end-of-range kinda demands maaan ?
[22:58:46] <theBear> and what does a air-exchanger do for you
[22:58:50] <Casper> I should take a look at those actually, and guestimate the leakage
[22:58:50] <theBear> generally speaking :)
[22:59:18] <Casper> will change the air in my work area
[22:59:25] <Casper> and recover most of the heat
[22:59:38] <Casper> so heat exchanger inside
[22:59:47] <Casper> thing is... FUCKING WINTER!!!!!
[23:00:00] <Casper> I mean.
[23:00:16] <Casper> you know... winter... freeze...<
[23:00:29] <Casper> meaning that the unit WILL freeze due to the condensation
[23:00:43] <theBear> wow.... does heat exchanger there mean the same as "heat-pump" does here, which is basically reversible compressor air-con
[23:00:53] <Casper> need to block all incomming air and let a small stream of inside warm air flow throught to unfreeze it
[23:01:00] <Casper> passive
[23:02:56] <Casper> a commercial unit usually use cross flow... basically it's a cube made out of many layers of corrugated plastic, each layer rotated 90 degree
[23:03:23] <Casper> this mean that one layer air can pass, but the second can't... on the same side...
[23:03:31] <Casper> the second side the order is reversed...
[23:03:50] <Casper> so air kinda do a X in it...
[23:03:52] <Casper> anyway
[23:04:28] <Casper> short story short: outside air go in, get heated up by the inside air that goes in, outside go now inside and inside go outside
[23:04:41] <Casper> commercial unit do 50-60% recovery
[23:05:00] <theBear> hmmm... i get half the practicals, that's enough for now :)
[23:05:04] <Casper> the unit I plan to build is 84%+ efficient
[23:06:36] <Casper> a commercial unit will recirculate the inside air while defrosting, and cut the outside air...
[23:06:49] <Casper> since mine is for my work area... must be able to work all the time
[23:15:10] <rue_house> how big is the pipe?
[23:15:15] <rue_house> round or rectangular?
[23:17:44] <Casper> pipe will be 4", going to a like.. 11x16" intake
[23:18:32] <rue_house> round or rectangular?
[23:18:49] <rue_house> GOD DAMNIT THIS TEA CUP IS EMPTY AGAIN
[23:19:35] <Casper> pipe is round, target is rectangular, plan is to put it in the 4" round, most likelly easier
[23:20:19] <rue_house> yea
[23:20:28] <rue_house> just a single O disk on a stick
[23:21:46] <Casper> that's one of the plan, and if I'm right, there should be basically zero torque there, so might be able to power down the stepper too, but geared might be better as it would increase the resolution and prevent it from moving by itself
[23:25:25] <Casper> rue_house: do you think car hvac blower do 300+cfm?
[23:26:29] <rue_house> no idea of scale on cfm
[23:26:37] <rue_house> whats a high power server fan?
[23:27:26] <carabia> Jesus. Debugging this fucking touchscreen controller is a pain in the ass.
[23:29:22] <Casper> server fan, around 100-125cfm
[23:29:28] <Casper> and 30-50$ each
[23:29:33] <carabia> Y-coordinate goes to nigh-zero when doing a quick vertical swipe, and it happens nowhere near where the coordinate would suggest, well, top.
[23:29:47] <carabia> It's a newhaven touchscreen, anyone had similar experiences?
[23:30:14] <carabia> well, s/touchscreen/controller
[23:30:43] <carabia> http://www.newhavendisplay.com/appnotes/datasheets/touchpanel/FT5336.pdf