#avr | Logs for 2012-09-13

Back
[01:25:02] <jadew> do this tools have any other name than "nibbler": http://www.radioshack.com/product/index.jsp?productId=12345949
[01:25:12] <en0> Hey, has anyone used these smoke sensors before (sen117a2b) Are they supposed to get hot?
[01:25:15] <jadew> I'm trying to find one of those at stores in europe, but I've had no luck so far
[03:22:43] <Roklobsta> hey any Altium PCB layout contractors in Melb. lurking here?
[03:23:29] <Richard_Cavell> You mean Melbourne Florida or Melbourne Australia?
[03:24:52] <Roklobsta> the ORIGINAL Melboutne, .au
[03:38:56] <megal0maniac> !seen fuzzybugkilla
[03:38:56] <tobbor> fuzzybugkilla was last seen in #avr on Sep 12 14:24 2012
[03:39:16] <megal0maniac> !resistors 2
[04:08:24] <megal0maniac> Can anyone tell me what the deal is with blue resistors? I can't seem to find a reliable answer...
[04:11:19] <jacekowski> blue are metal film
[04:11:29] <jacekowski> orange-ish are carbon film
[04:11:43] <jacekowski> it's easier to make precise metal film resistors
[04:11:48] <jacekowski> that's all
[04:13:33] <Richard_Cavell> And the metal film resistors in the shop where I buy them from, the legs are somehow easier to bend
[04:13:42] <Richard_Cavell> like, they'll keep a local bend better than the carbon film resistor legs
[04:26:04] <megal0maniac> I have an issue reading the colour codes, though. Are they the same?
[04:31:00] <Richard_Cavell> you mean are the colour codes on a metal film resistor the same as on a carbon resistor?
[04:31:01] <Richard_Cavell> yes
[04:34:00] <megal0maniac> Okay, thanks. I'll take another look :/
[04:34:34] <jacekowski> megal0maniac: only thing is that there are different types of colour codes
[04:34:42] <jacekowski> megal0maniac: as in, 4 band up to 7 band
[04:34:59] <jacekowski> megal0maniac: but idea is the same
[04:35:58] <megal0maniac> That's where the confusion comes in... I've got one which is Red Red Black Black Red
[04:36:03] <megal0maniac> Or the other way around
[04:36:29] <megal0maniac> How many for value, and how many for precision?
[04:39:04] <Richard_Cavell> That should be 22k ohms
[04:40:09] <megal0maniac> It's 1K... Within 2% of 1K 983 ohms.
[04:43:53] <Richard_Cavell> Are you colour blind?
[04:43:54] <Richard_Cavell> I am
[05:37:03] <OndraSter> http://sphotos-h.ak.fbcdn.net/hphotos-ak-ash3/564055_10151241043629973_779171217_n.jpg
[05:37:07] <OndraSter> Apollo's RAM
[05:37:11] <OndraSter> 2.5MB of LOL memory
[05:37:40] <OndraSter> whop
[05:37:44] <OndraSter> ROM*
[05:37:53] <nomis> *MB*?
[05:37:58] <nomis> I somewhat doubt this.
[05:39:10] <OndraSter> http://en.wikipedia.org/wiki/Core_rope_memory
[05:39:16] <OndraSter> this was just small test sample
[05:39:22] <nomis> seems 2.5MB is the data density. 2.5MB per cubic metre.
[05:39:22] <OndraSter> 2.5MB per cubic meter
[05:39:26] <OndraSter> yes
[05:40:40] <nomis> "The Block II Apollo Guidance Computer (AGC) used 36,864 sixteen-bit words of core rope memory (placed within one cubic foot) and 4,096 words of magnetic core memory (within two cubic feet)."
[05:41:09] <nomis> that'd be 80kB in total.
[05:47:36] <OndraSter> sorry. I read that on opera's facebook page and before I wiki'd it I already posted it here.
[05:57:22] <Richard_Cavell> what is core rope memory?
[05:58:01] <Richard_Cavell> I just read it
[05:58:04] <Richard_Cavell> that is bloody unbelievable
[05:58:10] <Richard_Cavell> that someone would go to the moon on something powered by that
[05:58:30] <Roklobsta> it was robust
[05:58:43] <Roklobsta> no cosmic rays are going to screw with that ram
[05:59:04] <Richard_Cavell> Well using that logic you should use pen and paper
[05:59:04] <Roklobsta> anyway, it did the job
[05:59:28] <Roklobsta> it was the cutting edge of embedded systems research too. i think a lot of embedded stuff was conceived on the apollo program
[05:59:57] <Roklobsta> like a supervisor program doing task and context switching
[06:01:06] <Roklobsta> the ROM was woven like in that picture. if the code changed the weaving ladies had to restitch wirings in the rom.
[06:01:22] <Roklobsta> no avrdude for them
[06:08:16] <Richard_Cavell> that's obscene
[06:15:07] <Roklobsta> it was robust
[06:15:21] <Roklobsta> andit worked
[07:17:05] <OndraSter> oh blimey
[07:17:06] <OndraSter> 14:04:21: [ERROR] USB driver attach timed out, ModuleName: TCF (TCF command: Tool:connect failed.)
[07:17:15] <OndraSter> Dragon..
[07:17:26] <OndraSter> I don't feel like reboot
[07:17:48] <OndraSter> hah
[07:17:50] <OndraSter> restarted AS6
[07:19:32] <OndraSter> YEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEWSSSSSSSSSSSSSSSSSSSSSSSSSS
[07:19:33] <OndraSter> http://clip2net.com/s/2iqhy
[07:20:02] <OndraSter> now
[07:20:04] <OndraSter> where to get VID & PID
[07:20:09] <OndraSter> rikus
[07:20:12] <OndraSter> where is h
[07:20:12] <OndraSter> e
[07:27:29] <karlp> $1000, or talk to your usb silicon vendor and ask for one.
[07:27:35] <karlp> also try openmoko.org
[07:29:01] <OndraSter> $2000 actually it used to be for one VID block
[07:31:04] <OndraSter> I tried asking them why do not they sublicense VID blocks
[07:31:05] <OndraSter> I got no reply
[07:31:12] <OndraSter> I got reply that they could hook me up with "hobbyists VID"
[07:31:14] <OndraSter> and PIC
[07:31:15] <OndraSter> PID
[07:31:18] <OndraSter> so I asked them to send it to me
[07:31:20] <OndraSter> again, no reply :D
[07:31:38] <OndraSter> MAC addresses for ethernets and BTs and such are sublicensed by different blocks
[07:31:44] <OndraSter> starting from $200 I think
[07:31:49] <OndraSter> for 4096 MACs I think
[08:07:03] <OndraSter> hmm
[08:07:07] <OndraSter> how can endpoint tell me
[08:07:10] <OndraSter> in wLength
[08:07:17] <OndraSter> that the data stage is 2036 ?!
[08:07:19] <OndraSter> 0x7f4
[08:07:27] <OndraSter> it is bloody full speed USB
[08:07:29] <OndraSter> 2.0
[08:07:35] <OndraSter> on control endpoint
[08:07:37] <OndraSter> 64 bytes max?!
[08:12:12] <jadew> WOOT! my open bench logic sniffer just arrived :D
[08:13:22] <OndraSter> cg
[08:13:37] <jadew> thanks
[08:44:29] <OndraSter> one can clearly see the progress of writing comments in my code
[08:44:39] <OndraSter> the stuff that I wrote on beginning = comments exist
[08:44:50] <OndraSter> and the more I progressed the more comments do not exist
[08:44:57] <OndraSter> basically, first hour = comments, anything after that = none
[08:45:06] <OndraSter> http://pastebin.com/5sfrFv3R
[08:51:34] <jadew> you'll experience the regret of the missing comments soon enough
[08:52:30] <jadew> about the open bench logic sniffer... I'm not particularly impressed with the software
[08:53:01] <jadew> it's buggy and it's missing important features, like turning off channels
[08:53:13] <jadew> and it can't properly decode 1-wire
[09:03:39] <rue_house> sounds like its made of java
[09:03:53] <jadew> heh, yeah
[09:04:15] <jadew> might roll my own if I run out of projects
[09:04:35] <jadew> seems like a simple project to take on, yet it seems all the available solutions are full of fail
[09:05:40] <rue_house> Tom_itx, funny funny, I'm to make a data logger
[09:16:09] <RikusW> jadew: you mean dallas 1-wire ?
[09:16:23] <jadew> RikusW, yeah
[09:16:41] <jadew> it's screwing up the reset
[09:16:56] * RikusW heard of it, read a little about it, and haven't had the motivation to really study it...
[09:17:21] <RikusW> though how hard can it be ?
[09:17:24] <jadew> it's easy
[09:17:35] <jadew> and quite useful really
[09:17:39] <RikusW> haven't had any use for it yet
[09:17:43] <jadew> 1 wire and lots of devices
[09:17:56] <jadew> no conflicts
[09:18:12] <jadew> address conflicts that is
[09:18:13] <RikusW> will probably put off learning it until I actually need it
[09:18:59] <jadew> get a 1-wire based chip :P
[09:19:12] * RikusW had a quick peek at the 1-wire packet format a while ago
[09:19:15] <jadew> I needed it for the DS18S20
[09:19:29] <jadew> it's a temperature sensor
[09:19:36] <RikusW> yeah, without a chip there is no easy way to try it out...
[09:23:03] <OndraSter> RikusW, don't ever say "how hard can it be"
[09:24:09] <jadew> nah, it's an easy protocol to bitbang
[09:24:23] <RikusW> how hard can USB programming be ? :-P :-D
[09:25:27] <jadew> the hardest protocol I bitbanged so far was usart
[09:25:42] <jadew> because of the crazy timing and the fact that it's asynchronous
[09:27:42] <karlp_> rx.
[09:27:46] <karlp_> tx bit banging is easy :)
[09:28:07] <jadew> it's not easy if rx starts happening in the middle of tx :)
[09:30:14] <jadew> that's a case I couldn't handle
[09:30:32] <jadew> tx during rx is fine, cuz you start sending on the rx timing
[09:30:42] <RikusW> mmh then try doing an autobauding uart...
[09:30:47] <jadew> but receiving on tx timing is not that reliable
[09:31:29] <RikusW> better to have a hard uart
[09:32:13] <jadew> yeah
[09:32:26] <jadew> soft uart works fine in half duplex tho
[09:32:54] <karlp_> I've got a couple of devices at home that are tx only.
[09:32:59] <jadew> my soft implementation works in full duplex too, but only at low speeds and not for a very long time
[09:33:11] <jadew> since it can't sync on the start bit during full duplex
[10:38:01] <OndraSter> The Communications Class interface requires one endpoint, a management element. It optionally can have an additional endpoint, the notification element. The management element uses the default endpoint for all standard and Communications Class-specific requests. The notification element normally uses an interrupt endpoint.
[10:38:04] <OndraSter> does that mean that:
[10:38:08] <OndraSter> a) control + bulk
[10:38:10] <OndraSter> b) control
[10:38:11] <OndraSter> ..?
[10:55:03] <amee2woof> are screen names starting with an "x" overused? o.O
[10:56:36] <Landon> <xXcloud4lyfeXx> no
[13:27:38] <OndraSter> RikusW, how did you solve that VID&PID combinationissue?
[13:27:41] <OndraSter> combination issue*
[13:27:46] <OndraSter> aka... not licensing whole VID block
[13:48:45] <RikusW> OndraSter: I used the atmel demo ones
[13:49:15] <RikusW> afaik you can't (don't want to)
[13:49:24] <RikusW> its expensive to get a VID
[13:49:38] <RikusW> there is some companies selling PIDs
[13:50:29] * RikusW is a little distracted, just started SG1 S9, finished S8 a few minutes ago :)
[13:52:44] <specing> Haven't they reserved a VID for internal use?
[13:53:18] <specing> imho 65536 PIDs is a little too much even for the largest companies
[13:53:38] <specing> they should have a B class for small companies
[13:53:51] <specing> and/or a /28
[13:54:00] <specing> 16 products...
[13:54:04] <specing> more than enaugh.
[14:01:17] <OndraSter> yep, if they sold 8 PIDs for $10
[14:01:25] <OndraSter> they would make 80 grand US bucks
[14:02:49] <OndraSter> MAC addresses have different blocks that are being sold
[14:02:56] <specing> $5368709120 for the whole range
[14:03:08] <specing> No, MACs are /24 blocks
[14:03:34] <specing> but that is reasonable since MACs are supposed to be unique
[14:03:46] <OndraSter> no
[14:03:51] <OndraSter> MACs are sold not just in /24 blocks
[14:06:06] <specing> /16 too?
[14:06:10] <OndraSter> don't remember
[14:06:13] <OndraSter> but there were two or three options
[14:07:59] <OndraSter> I think there was 4096 option as well
[14:08:02] <OndraSter> hmm
[14:08:05] <OndraSter> it is quite some time..
[14:08:53] <specing> /36 ?
[14:09:11] <specing> that seems a bit ... little
[14:17:57] <OndraSter> if USB was selling per 4096 PIDs
[14:17:59] <OndraSter> wouldn't you take it?
[14:18:03] <OndraSter> even for $100
[14:18:17] <OndraSter> more like $200
[14:21:46] <specing> no
[14:22:16] <specing> What would I do with 4096 PIDs?
[14:27:09] <Tom_itx> sell them
[14:31:50] <OndraSter> specing, better than trying to get rid of 65535 PIDs
[14:37:38] <tmpvar> OndraSter, do you need PIDs for generic devices?
[14:38:21] <Kevin`> tmpvar: technically, yes
[14:43:33] <tmpvar> bah, that is slightly annoying!
[14:54:20] <Kevin`> most people making such devices don't bother getting usb certification at all
[14:55:13] <Kevin`> either just use the chip's vid/pid if it has one, or copy another devices
[14:55:23] <Kevin`> all's fair in china
[15:14:43] <karlp_> macs are sold individually too.
[15:14:51] <karlp_> yu just buy the eui48 eeprom from digikey, and off you go.
[15:28:31] <specing> karlp_: :D
[15:30:39] <RikusW> Or just build an informal settlement in PID land :-P
[15:46:08] <tmpvar> so, what are the consequences for using a random VID? Will you unleash the hordes of USB lawyer robots?
[15:46:54] <specing> No, numbers cannot be trademarked/patented
[15:47:58] <tmpvar> what is the worst case?
[15:51:12] <tmpvar> if you happen to choose a vid that conflicts with another vendor, they happen to be using the same pid, and your user plugs in both devices at the same time.. you will get a conflict or some such? that seems pretty unlikely.
[15:52:03] <specing> and because drivers are userspace...
[15:52:35] <tmpvar> right, so the devices would also have to be different classes
[15:53:06] <tmpvar> has anyone here paid the $2k ?
[15:55:37] <RikusW> Windows and Linux will detect the device incorrectly....
[15:56:02] <RikusW> unless you're building a clone
[15:56:46] <RikusW> tmpvar: there is a list of registered VIDs
[15:57:10] <RikusW> I'll go and take an unused one and grab a random PID
[15:57:48] <RikusW> if the device is for internal or non widespread use it shouldn't cause too much if any trouble
[16:12:44] <OndraSter> hmm
[16:12:45] <OndraSter> for xboard
[16:12:51] <OndraSter> so yes, for now "non widespread" works lol
[16:17:31] <tmpvar> haha
[16:25:26] <OndraSter> how did arduino solve VID issue?
[16:25:27] <OndraSter> hmm
[16:26:11] <tmpvar> i believe they use the FTDI's VID
[16:26:16] <OndraSter> I mean on Leonardo
[16:26:20] <OndraSter> it has mega32u4
[16:26:20] <tmpvar> on non-leonardo..
[16:26:22] <tmpvar> right
[16:26:31] <OndraSter> obviously... why would I ask for FTDI one :P
[16:27:45] <tmpvar> oh, they may just use the atmel vid?
[16:28:06] <tmpvar> (thats what lufa does anyhow)
[16:28:16] <OndraSter> hmm
[16:28:17] <OndraSter> could be
[16:28:23] <OndraSter> but I thought that Atmel doesn't give out PIDs
[16:29:17] <tmpvar> looks like they are using 9025
[16:29:21] <tmpvar> which i dont see in the list
[16:29:38] <OndraSter> hmm
[16:29:54] <OndraSter> I am still taking ownership of my ex-desktop folder form previous windows installation
[16:30:21] <OndraSter> hmm! really no 9025
[16:30:23] <OndraSter> http://www.linux-usb.org/usb.ids
[16:30:25] <OndraSter> I am checking here
[16:31:10] <tmpvar> i was looking through http://www.usb.org/developers/tools/comp_dump
[16:31:26] <tmpvar> which seems less "complete"
[16:31:42] <tmpvar> oh i gave you their VID in decimal
[16:32:14] <tmpvar> 2341 Arduino SA
[16:32:23] <OndraSter> damn
[16:32:28] <OndraSter> they have their own VID
[16:33:25] <tmpvar> we should all go in on a "Super Awesome Devices" VID.. heh
[16:34:52] <OndraSter> buy one all of us together? :D
[16:34:53] <OndraSter> I am in
[16:35:16] <OndraSter> it is like $.03 per PID. I suggest everybody will grab 32 PIDs for $1 :)
[16:35:34] <OndraSter> hey, that is not a bad idea at all!
[16:36:03] <OndraSter> isn't it actually illegal to re-sell or sublicense PIDs?
[16:36:29] <tmpvar> lol, no clue
[16:36:43] <OndraSter> I think it is
[16:36:47] <OndraSter> sell for sure, give out not sure
[16:36:52] <OndraSter> oh well
[16:36:56] <OndraSter> I'd be giving them out for a small fee :)
[16:36:58] <LoRez> illegal, probably not. violation of your contract with the issuer, probably.
[16:37:03] <OndraSter> yes, that
[16:38:01] <OndraSter> seriously, 64k of numbers is too much for one company to accomodate... $2k for running company is not much, but for smaller project it is
[16:38:10] <tmpvar> agreed
[16:38:34] <OndraSter> I'd have to sell ... about 300 XBoard cocos to pay for one VID block :(
[16:38:57] <OndraSter> a bit more actually
[16:39:19] <OndraSter> when one counts all the stuff other than parts in it as well (electricity to power reflow oven, solder paste, time, ..)
[16:39:33] <tmpvar> from wikipedia: "Some microcontroller manufacturers offer a free or low cost sublicense of their vendor ID for development/testing and limited production (generally under 10,000 units)." -- http://en.wikipedia.org/wiki/USB_Implementers_Forum#Obtaining_a_vendor_ID
[16:40:00] <OndraSter> let's contact atmel
[16:40:05] <OndraSter> they have been very helpful on everything so far
[16:40:12] <OndraSter> they have even issued some datasheet fixes from my reports
[16:40:16] <OndraSter> just as AS6 fixes
[16:40:19] <tmpvar> nice
[16:40:32] <OndraSter> like the EBI stuff... that talks about PortL
[16:40:39] <OndraSter> which lacks every single xmega :D
[16:41:12] <tmpvar> I'd love to hear what they say
[16:41:58] <OndraSter> "yada yada, your description is the way it should be, yes yada yada
[16:41:59] <OndraSter> I fully understand that this documentation is difficult to find, and will report a bug on
[16:41:59] <OndraSter> the manual so that it will be improved.
[16:42:01] <OndraSter> yada yada"
[16:44:33] <OndraSter> http://clip2net.com/s/2ixxR
[16:44:38] <OndraSter> but what about my own PID for atmel's VID? :P
[16:49:32] <OndraSter> annnnd request sent
[16:49:46] <tmpvar> awesome
[16:51:32] <tmpvar> ok, so since I can't seem to get an atmega board of my own design to work im going to steal a barebones example and try that instead
[16:52:24] <Tom_itx> which atmega?
[16:52:30] <tmpvar> 32u4
[16:53:55] <OndraSter> tmpvar, eh
[16:54:08] <OndraSter> define not working
[16:54:24] <OndraSter> wait, weren't you the guy that had those LED stuff?
[16:54:38] <tmpvar> well, at this exact moment, it only enumerates when ISP (avrispmkii) is connected
[16:54:47] <Tom_itx> tmpvar, you mean like this one? http://tom-itx.dyndns.org:81/~webpage/boards/atmega32u4/atmega32u4_1.jpg
[16:54:48] <OndraSter> show us the design
[16:55:25] <Tom_itx> i bet you overwrote the bootloader if you used ISP on it
[16:55:36] <tmpvar> OndraSter, sec
[16:55:48] <tmpvar> Tom_itx, yeah, that was on purpose ;)
[16:55:52] <OndraSter> also, why people use mega32u4? It costs two times xmega32a4u :(
[16:55:58] <OndraSter> well, nearly two times
[16:56:04] <Tom_itx> i had these
[16:56:19] <tmpvar> OndraSter, im a noob
[16:56:30] <OndraSter> xmega just expands mega :)
[16:56:33] <OndraSter> the core is still the same
[16:56:37] <OndraSter> and the philosophy :)
[16:56:47] <tmpvar> and its faster, no?
[16:56:48] <OndraSter> I want to bring xmega into broad public with my XBoard
[16:56:50] <OndraSter> yes
[16:56:55] <OndraSter> and has many times more possibilities
[16:57:20] <tmpvar> yeah, i got some xmega's in today
[17:03:23] <tmpvar> OndraSter, we've done this dance before .. here's the brd and schematic http://tmpvar.com/tmp/atmega32u4-debug.tgz
[17:03:36] <OndraSter> so it is you, the "LED" guy :)
[17:04:38] <tmpvar> lol
[17:06:08] <OndraSter> ow that grounding is nasty :D
[17:06:25] <OndraSter> just as vcc :o
[17:06:56] <OndraSter> let me fix that for you :)
[17:07:05] <tmpvar> yeah, you yell at me every time about something new.. lololol
[17:07:12] <OndraSter> :D
[17:09:10] <OndraSter> well
[17:09:17] <OndraSter> you forgot to connect UVCC to vcc
[17:09:31] <OndraSter> no wait
[17:09:35] <OndraSter> it just isn't named :)
[17:09:36] <OndraSter> sorry
[17:09:44] <OndraSter> actually
[17:09:48] <OndraSter> eagle says those are different lines
[17:12:18] <tmpvar> doh.
[17:12:33] <OndraSter> I will then re-check your brd
[17:15:10] <tmpvar> oh man, i feel like an idiot now
[17:15:14] <OndraSter> lol
[17:15:23] <tmpvar> still curious how it works with ISP connected though
[17:21:22] <OndraSter> once I will re-check your brd.. :)
[17:23:56] <OndraSter> http://tempah.myxboard.net/mega32u4.zip
[17:24:01] <OndraSter> that is new brd :)
[17:24:05] <OndraSter> hopefuly fixed
[17:25:42] <tmpvar> haha, no vias. nice
[17:26:52] <OndraSter> sure no vias
[17:27:00] <OndraSter> such board doesn't need them :)
[17:27:06] <OndraSter> you'd be surprised what can be made on one layer
[17:33:29] <tmpvar> ok, this looks good. thank you OndraSter
[17:34:02] <Tom_itx> he's our new resident guru
[17:34:16] <OndraSter> tmpvar, well
[17:34:24] <OndraSter> the issue you have is simple :)
[17:34:26] <OndraSter> let me draw it all
[17:35:16] <OndraSter> http://img.ondraster.cz/up/errord56.png
[17:35:16] <OndraSter> here
[17:35:35] <OndraSter> also it isn't actually UVCC but UCAP I think
[17:35:44] <OndraSter> which is actually good on that schematics, sorry for that
[17:35:46] <OndraSter> yellow = ignore :)
[17:35:50] <OndraSter> but the green one is the culprit
[17:36:18] <OndraSter> also you could do the design I sent you much, much smaller
[17:40:09] <OndraSter> when you connected your programmer it was most likely supplying power
[17:40:18] <OndraSter> either through the Vtg pin directly or through data pins :)
[17:40:29] <Vutral> eh
[17:40:35] <Vutral> apropos
[17:40:37] <OndraSter> it is regular stuff really
[17:40:50] <OndraSter> you can power atmega through data pins
[17:40:52] <OndraSter> it makes sense as well :)
[17:40:57] <Vutral> is it possible to drive fullduplex rs485 with only rx + tx ?
[17:41:06] <Vutral> leaving the the enable pins away ?
[17:41:54] <Vutral> mhm well i could just use 2 attiny ^^ maybe
[17:42:00] <Vutral> and use i2c between them mhm
[17:42:06] <OndraSter> heh
[17:42:11] <OndraSter> or bigger tiny/mega?
[17:42:18] <Vutral> well i got some atmega328 here
[17:42:22] <Vutral> but it got too many pins
[17:42:25] <Vutral> i dont got matching sockets
[17:42:32] <Vutral> i would need another bigger tiny
[17:42:35] <OndraSter> ah
[17:42:44] <OndraSter> MULTIPROCESSING WITH ATTINY! lol
[17:42:50] <Vutral> not really
[17:42:54] <Vutral> its getting complicated
[17:43:02] <Vutral> i just try to convert rs485 to spi
[17:43:11] <Vutral> so i got a long range link to a sigma delta adc
[17:43:42] <Vutral> so i need a small automaton in the attiny which buffers the spi
[17:43:53] <Vutral> so it works no matter which propagation delay on the link
[17:43:59] <Vutral> but usually i would need more signals *g*
[17:44:06] <Vutral> 4 wires for spi and 1 for dataready
[17:44:16] <OndraSter> hmm packet system? :)
[17:44:25] <Vutral> yeah i build some kind of routing
[17:44:27] <OndraSter> always ... 64? bytes long packets
[17:44:31] <OndraSter> and always send back CRC
[17:44:31] <Vutral> converting all the stuff to ethernet in the end
[17:44:40] <OndraSter> once the data is sent set SPI to slave
[17:44:49] <Vutral> yeah
[17:45:25] <Vutral> but well the attiny needs to write and read
[17:45:29] <Vutral> mhm
[17:45:38] <Vutral> and i am a bit unsure how i signal back data ready
[17:45:47] <Vutral> as i got a full implementation for the adc
[17:45:53] <Vutral> but it wont ever fit in the tiny *g*
[17:45:55] <OndraSter> ... I'd suggest USB way :)
[17:46:01] <Vutral> usb sucks
[17:46:02] <OndraSter> keep just one tiny as master
[17:46:03] <OndraSter> yes
[17:46:04] <OndraSter> it does
[17:46:05] <Vutral> i need galvanic isolation
[17:46:06] <OndraSter> really heavily
[17:46:11] <OndraSter> mm galvanic
[17:46:16] <OndraSter> optoisolators?
[17:46:21] <OndraSter> what about optical SPI link? :o
[17:46:22] <Vutral> they need power
[17:46:23] <Vutral> :P
[17:46:25] <OndraSter> so?
[17:46:30] <Vutral> i got optoisolators my friend
[17:46:33] <Vutral> 6n137 even
[17:46:38] <Vutral> which work for rs485
[17:46:48] <Vutral> but i need a dc/dc converter for the master interface
[17:46:49] <Vutral> at least
[17:46:51] <OndraSter> I was more leaning towards rxtx diodes + cable :P
[17:46:55] <Vutral> else i have a gigantic ground loop
[17:47:01] <Vutral> rxtx diodes ?
[17:47:06] <OndraSter> you know, optical link
[17:47:10] <OndraSter> optical wire
[17:47:15] <OndraSter> damn, I can't find the proper word
[17:47:18] <Vutral> as long as it snot for audio sure
[17:47:25] <Vutral> but anyway i got everything for this way yet
[17:47:32] <Vutral> now switching is a bit weird
[17:47:40] <OndraSter> ah
[17:47:53] <Vutral> and anyway if i implement full rs485 i can exploit existing infrastructure
[17:47:53] <OndraSter> one of my friends is building desktop meter with few modes
[17:47:57] <OndraSter> and made it modular
[17:48:02] <OndraSter> with infrared data transfers :)
[17:48:03] <Vutral> yeah welll
[17:48:14] <Vutral> i build fully isolated sigma delta 24bit adc measurement dataq module
[17:48:28] <Vutral> guess i want to add timestamps
[17:48:56] <Vutral> can use modbus or so over rs485
[17:49:56] <Vutral> well for prototyping some ready2use stuff would be nice
[17:49:59] <Vutral> but anyway
[17:50:19] <Vutral> i only got stuff here for soldering it :P
[17:51:16] <OndraSter> I need to find out if I still have got here somewhere the mega168 I want to use
[17:51:26] <OndraSter> if not I will probably remake it for xmega32d4
[17:51:28] <OndraSter> much cheaper
[17:51:29] <OndraSter> more flash
[17:51:32] <Vutral> well
[17:51:33] <OndraSter> more pins that I need :)
[17:51:38] <Vutral> lpc isnt bad
[17:51:49] <OndraSter> I need it for oven controller
[17:51:56] <Vutral> i build something similar
[17:51:58] <Vutral> temperature control
[17:52:03] <OndraSter> lpc = nxp's lpc or low pin count? :D
[17:52:07] <Vutral> low pin count
[17:52:09] <Vutral> we are in avr
[17:52:14] <OndraSter> oh :D
[17:52:15] <Vutral> i thought its obvious i dont mean nxps stuff
[17:52:21] <OndraSter> I first thought about those NXPs
[17:52:24] <OndraSter> because they are dirt cheap
[17:52:30] <Vutral> well i use pt1000 thermocouples
[17:52:42] <Vutral> i could probably just implement something more analog
[17:52:54] <Vutral> but its nice if its programmable
[17:52:57] <OndraSter> I am using thermocouple + max3185 or w/e is the number
[17:52:58] <Vutral> and i dont got zero drift opamps here
[17:53:06] <OndraSter> thermocouple in, SPI out :)
[17:53:12] <Vutral> i couldnt find really small affordable dataaq systems
[17:53:33] <Vutral> oh well i do not only want to measure temperature so well
[17:53:43] <Vutral> max3185 let me check
[17:53:49] <OndraSter> I am not sure about the number
[17:53:51] <OndraSter> from top of my head
[17:53:59] <OndraSter> it is the only thermocouple controller I found
[17:54:11] <Vutral> well i got 4 wires
[17:54:20] <Vutral> i think you only will find 3 wire maximum
[17:54:22] <Vutral> because of the cost
[17:54:28] <Vutral> this is laboratory grade thermocouple :P
[17:54:30] <OndraSter> oh noes, maxim has new website
[17:54:38] <Vutral> i know their new website yet
[17:54:45] <Vutral> i use max3089 rs485 chips
[17:55:01] <OndraSter> MAX31855
[17:55:03] <OndraSter> I used that one
[17:55:15] <Vutral> well i can order that in 1 month or so
[17:55:16] <Vutral> a bit more
[17:55:18] <Vutral> not earlier
[17:55:25] <OndraSter> wait between samples? :D
[17:55:26] <Vutral> ordering only every 2 month lol
[17:55:27] <Vutral> *g*
[17:55:47] <Vutral> well the new one is at least not more expensive
[17:55:49] <Vutral> but i wonder what the differences are
[17:55:50] <OndraSter> the thing is - the more pins the better for me. I am controlling LED display from old microwave :)
[17:56:04] <Vutral> lol
[17:56:05] <OndraSter> I had to go only for one button operation :D
[17:56:17] <OndraSter> i like it more than $2.5 HD44780 16x2 display :D
[17:56:44] <Vutral> sure
[17:56:55] <Vutral> hey OndraSter
[17:57:03] <Vutral> do you think something like a approximating slope adc makes sense
[17:57:11] <OndraSter> huh?
[17:57:17] <Vutral> adding a dac driven opamp infront of a approximating adc
[17:57:19] <Vutral> :P
[17:57:36] <OndraSter> sounds liek a lot of work :)
[17:57:40] <Vutral> i mean i could balance for middle with a comparator
[17:57:45] <Vutral> well there are very fast dacs
[17:57:49] <Vutral> a sigma delta isnt so difficult
[17:58:25] <Vutral> oh well it would make sense if you could autorange :P
[17:58:50] <Vutral> but the nonlinearities are bad :P
[18:01:15] <OndraSter> atmega168l-8AU
[18:01:17] <OndraSter> AAAARGH
[18:01:21] <OndraSter> why!
[18:01:43] <OndraSter> I need 20MHz for refreshing the display...
[18:01:48] <Vutral> lolololol
[18:01:55] <Vutral> know what sucks
[18:01:59] <Vutral> if you need to convert level
[18:01:59] <OndraSter> no?
[18:02:01] <Vutral> and got no fets
[18:02:02] <Vutral> :P
[18:02:04] <OndraSter> heh
[18:02:24] <Vutral> well guess i could use some bjt
[18:02:40] <OndraSter> relays yo :)
[18:02:44] <OndraSter> isolated even
[18:02:47] <Vutral> relays ?
[18:02:53] <Vutral> yeah drive a relay with 30 mhz
[18:02:55] <OndraSter> yes!
[18:02:58] <OndraSter> lol
[18:02:59] <Vutral> :P
[18:03:05] <Vutral> sounds clever
[18:03:09] <Vutral> isolated high speed mechanical spi
[18:03:29] <OndraSter> haha
[18:03:31] <Vutral> no well but anyway
[18:03:42] <Vutral> capacitance is evil
[18:04:14] <OndraSter> mmm I would sometimes like to try out capacitive qtouch buttons :)
[18:04:37] <Vutral> mhm
[18:04:41] <Vutral> why not
[18:05:24] <Vutral> you can build capacitive stuff yourself
[18:05:27] <Vutral> ^^
[18:05:32] <OndraSter> ye
[18:05:39] <Vutral> two foils with traces got variable capacitance :P
[18:05:39] <OndraSter> but not sure I want to for this oven controller :)
[18:05:48] <Vutral> *g*
[18:05:49] <OndraSter> hmm atxmega16d4 cheaper than atmega168
[18:06:08] <Vutral> lol
[18:06:09] <Vutral> xmega
[18:06:10] <OndraSter> 1.48€ vs 1.65€
[18:06:13] <Vutral> totempole rules
[18:06:18] <OndraSter> totempole?
[18:06:27] <Vutral> yeah
[18:06:34] <Vutral> you know multimaster spi ?
[18:06:43] <OndraSter> I know multiprocessor UART
[18:06:52] <Vutral> multiprocessor uart ?
[18:06:58] <OndraSter> yes
[18:07:03] <OndraSter> it is probably atmel-only thing though
[18:07:09] <Vutral> lol
[18:07:13] <Vutral> how does it work
[18:07:18] <Vutral> interrupt routing table ?
[18:07:21] <OndraSter> special transfers with 9th bit
[18:07:27] <OndraSter> signalize that this is special transfer
[18:07:33] <Vutral> special ?
[18:07:37] <OndraSter> yes
[18:07:42] <OndraSter> and when you have enabled MPM it waits for this transfer
[18:07:53] <OndraSter> and then it interrupts when it comes and you can eg check address or anything
[18:07:53] <Vutral> sounds senseless
[18:07:59] <Vutral> mhm
[18:08:10] <OndraSter> if you do not want to become active - you just put the MPM bit back on
[18:08:14] <OndraSter> and wait for new one to come
[18:08:27] <Vutral> strange
[18:08:34] <Vutral> whats the application of this ?
[18:08:37] <Vutral> :P
[18:08:50] <Vutral> i mean why would i want to stop all cores
[18:08:54] <OndraSter> no idea
[18:08:57] <OndraSter> multiprocessor UART :)
[18:09:05] <OndraSter> multiple devices on one UART
[18:09:12] <OndraSter> don't ask me
[18:09:16] <OndraSter> I just read it in the datasheet
[18:09:19] <Vutral> so some kind of internal routing capability
[18:09:28] <Vutral> allowing you to implement it in software
[18:09:43] <Vutral> the datasheet
[18:09:48] <Vutral> i would prefer a application note of it
[18:10:04] <OndraSter> ask atmel's website :)
[18:10:06] <OndraSter> they might have
[18:10:57] <Vutral> is there a possibility to switch chipselects circular ?
[18:11:08] <Vutral> lol
[18:11:29] <Vutral> eh then i could just pulse it *g*
[18:11:31] <OndraSter> Tom_itx, your oven controller is mega168, right?
[18:11:48] <OndraSter> why am I asking, 16kB for code is more than enough :D
[18:12:12] <Vutral> 16kb are enough probably
[18:12:19] <Vutral> it matters waht you do with it
[18:12:20] <OndraSter> more than enough
[18:12:22] <Vutral> if you store big tables or so inside
[18:12:26] <OndraSter> big? no
[18:12:30] <Vutral> or pgmspace for chars
[18:12:31] <Vutral> :P
[18:12:45] <OndraSter> it has got 2kB RAM
[18:12:48] <OndraSter> compare that to mega168
[18:12:50] <Vutral> and well i guess for some applications you exploit flash writing for a log
[18:12:54] <OndraSter> with 1k
[18:12:59] <OndraSter> no
[18:13:01] <OndraSter> I don't need log :D
[18:13:08] <Vutral> well some people might want to log into flash
[18:13:12] <Vutral> *g*
[18:13:13] <OndraSter> I might put onto some microSD connector
[18:13:17] <OndraSter> since I will have spare pins
[18:13:19] <Vutral> without microsd
[18:13:22] <OndraSter> heh
[18:13:25] <Vutral> for errors
[18:13:25] <Vutral> you know
[18:13:27] <Vutral> ^^
[18:13:27] <OndraSter> ye
[18:13:35] <tmpvar> OndraSter, i was worried about the thin traces, but they seem to have toner transferred quite well (http://i.imgur.com/DI8OP.jpg)
[18:13:36] <Vutral> anyway
[18:13:52] <OndraSter> tmpvar, I told you to optimalize it :)
[18:13:54] <Vutral> eepram would be nicer ^^
[18:13:58] <OndraSter> you could have made the space half
[18:14:02] <OndraSter> Vutral, and dump it to PC how? :P
[18:14:19] <Vutral> space half ?
[18:14:26] <Vutral> what do you mean
[18:14:26] <OndraSter> cut the required space in half*
[18:14:31] <OndraSter> that is on tmpvar
[18:14:32] <tmpvar> OndraSter, haha, well I don't want to touch it. If it works then I'll feel more comfortable tweaking
[18:14:36] <OndraSter> :D
[18:14:40] <Vutral> i dont get you
[18:14:48] <OndraSter> I did not check datasheet, so if there are more errors I haven't seen them
[18:14:52] <OndraSter> but the board layout should be k
[18:14:53] <OndraSter> ok*
[18:15:00] <Vutral> gna
[18:15:02] <Vutral> so guys
[18:15:04] <OndraSter> gn
[18:15:14] <Vutral> i need something for prototyping
[18:15:14] <OndraSter> gna != gn?
[18:15:28] <Vutral> gna <-- sound
[18:15:36] <Vutral> a expression of frustration
[18:15:37] <OndraSter> oh lol
[18:15:43] <OndraSter> the same as Cartman does?
[18:15:50] <Vutral> no idea
[18:15:57] <Vutral> i stole it from disney
[18:16:01] <OndraSter> :D
[18:16:04] <Vutral> long ago
[18:22:16] <karlp_> vutral, why do you need to separate the s/d adc from the processor?
[18:22:36] <karlp_> why not put the processor at the far side, and let rs485 do what it was meant to do.
[18:22:37] <Vutral> i just isolate the interface
[18:22:47] <Vutral> the processor is at the far side
[18:22:56] <Vutral> but i need a processor to translate from spi and rs485
[18:25:40] <karlp_> I think I'm missing something sorry :)
[18:26:07] <Vutral> karlp_, ?
[18:26:16] <Vutral> that you cant attach spi to rs485 directly ?
[18:26:21] <karlp_> I have a power meter on my desk at work which is atmega running as a modbus slave over rs485, talking spi to the fancy sigma /delta power metering adc,
[18:26:46] <karlp_> you seem to want to put the chip handling the readings at the far end of the rs485 link and do some sort of spi over rs485?
[18:26:57] <Vutral> yeah
[18:27:17] <karlp_> where does the data eventually need to be?
[18:27:24] <Vutral> i need to process multiple spi commands at once
[18:27:35] <Vutral> and the sigma delta adc needs to tell me when a convesion is ready
[18:27:44] <Vutral> i would need to pass all this data to the far end ^^
[18:28:09] <Vutral> so with a higher conversion speed
[18:28:19] <Vutral> i would generate invalid commands because the time window is missed
[18:28:44] <Vutral> if its on your desk
[18:28:48] <Vutral> it shouldnt be that critical sure
[18:28:50] <Vutral> ^^
[18:29:52] <karlp_> yeah, so, I would put the thing doing the collection right beside the adc,
[18:30:03] <karlp_> and let it communicate the resulting processed data out the rs485 port.
[18:30:13] <Vutral> sure
[18:30:23] <OSterver> hmm I have dilemma... either use the display or use HD44780 which can actually show more than just 7segment numbers and some signs :(
[18:32:33] <OSterver> I would save up on transistors and such as well
[18:33:00] <karlp_> Vutral: so, what's stopping you from doing that? why the effort to run spi over rs485 from a distance?
[18:33:07] <karlp_> what's the motivation to do it that way that I'm missing?
[18:33:34] <Vutral> well i doing modbus
[18:33:47] <Vutral> just trying to choose a uc to sit beside the s/d adc
[18:33:57] <Vutral> :P
[18:34:24] <OSterver> hmm
[18:34:30] <karlp_> the cheapest thing with 1 spi port, 1 uart, and 1 spare gpio to run the tx enable line on the 485 transceiver?
[18:34:32] <OSterver> does the hd44780 work off 3.3v? :P
[18:34:56] <Vutral> karlp_, yeah and one line for dataready of the adc
[18:35:07] <Vutral> so the adc needs 4 + 1 wire
[18:35:18] <karlp_> whatever it needs, spi things normally need irq/ready/blah lines
[18:35:19] <Vutral> and rs485 at least 3 wires
[18:35:23] <Vutral> with shutdown it needs 4 wires
[18:35:42] <Vutral> yeah MISO MOSI SCK CS DRDY
[18:35:44] <karlp_> yeah, I've never used shutdown stuff.
[18:36:16] <Vutral> rs485 needs RX TX DE/RE
[18:36:19] <karlp_> so, there's about 100000 parts that fit that bill.
[18:36:27] <karlp_> pick the one you've most wanted to play with.
[18:36:37] <Vutral> sure
[18:47:30] <OSterver> nope hd44780 doesn't work at 3v3
[18:47:38] <OSterver> level translator time!
[18:47:47] <OSterver> well, maybe not
[18:47:54] <OSterver> maybe it will take 3v3 logic!
[18:47:58] <Vutral> mhm
[18:48:02] <Vutral> level translator ?
[18:48:08] <OSterver> yes
[18:48:18] <OSterver> either 74hc244 or similar
[18:48:21] <OSterver> or transistors
[18:48:26] <Vutral> i got logic gates well
[18:48:38] <Vutral> i saw ti sells a universal one
[18:48:45] <Vutral> where you can set the levels
[18:49:00] <Vutral> often enough you dont need to convert level mhm
[18:49:23] <OSterver> yes, that is what I am about to test out
[18:49:29] <Vutral> bidirectional conversion with a bc548 but i dont remeber it how it worked
[18:49:43] <Vutral> i got a 74LS circuit on my wall
[18:49:48] <Vutral> squaring circuit etc :P
[18:49:50] <OSterver> I don't need bidirectional
[18:50:11] <Vutral> have you checked the levels it takes ?
[18:50:27] <Vutral> i got some cmosses here which work fine with bipolar levels
[18:50:39] <OSterver> I am just about to
[18:51:19] <Vutral> i would be more intrested in low speed lvds quad ^^
[18:51:40] <OSterver> this means that xboard can not directly interface hd44780 :(
[18:51:46] <OSterver> without another 5v power supply
[18:51:51] <OSterver> will be fixed in revB
[18:51:57] <Vutral> but i got tired checking all the possibilities
[18:51:59] <OSterver> with VUSB being broken out
[18:52:13] <Vutral> :(
[18:52:17] <Vutral> what about a step up converter
[18:52:40] <OSterver> lol
[18:52:52] <OSterver> so I can power it FROM USB, step it down to 3v3
[18:52:55] <OSterver> and then back up to 5v?
[18:53:04] <Vutral> lol
[18:53:29] <Vutral> maybe if you power it from 3v3 volt it amkes sense
[18:53:34] <Vutral> i dont know your psu possibilities
[18:53:41] <Vutral> but if i run from a 3v7 liion
[18:53:46] <Vutral> i would like to step up maybe
[18:53:48] <OSterver> well then
[18:53:49] <OSterver> yes
[18:54:03] <OSterver> you need to do that with arduino as well though
[18:54:17] <Vutral> there are universal linear regulators
[18:54:24] <OSterver> step up - yes
[18:54:26] <Vutral> buck boost
[18:54:33] <OSterver> yes
[18:54:38] <OSterver> but why would I use those on my board?
[18:54:40] <Vutral> with powersaving if you got the right voltage already
[18:54:41] <OSterver> I need 3v3 for the MCU
[18:54:44] <Vutral> opening a direct path
[18:54:50] <Vutral> no idea
[18:54:56] <Vutral> you probably want a dual regulator
[18:54:59] <Vutral> which can output both
[18:55:16] <OSterver> or I don't want anything
[18:55:18] <OSterver> just 3v3 ldo
[18:55:22] <Vutral> sure
[18:55:28] <Vutral> ^^
[18:55:33] <OSterver> and let the user decide
[18:55:40] <Vutral> sounds annoying
[18:55:43] <OSterver> heh
[18:55:47] <OSterver> he has to do the same thing on arduino
[18:55:58] <Vutral> better buy no arduino then :P
[18:55:59] <OSterver> even worse, he can't use single 3v7 battery to power his arduino
[18:56:04] <OSterver> he can power my xboard just fine
[18:56:04] <Vutral> arduino got 3v3 and 5v out in rev3
[18:56:06] <Vutral> on uno
[18:56:07] <Vutral> :P
[18:56:08] <OSterver> yes
[18:56:16] <OSterver> but does it run on 3v3?
[18:56:24] <Vutral> no idea
[18:56:29] <Vutral> it got a 3v3 volt pin
[18:56:38] <OSterver> it does not run on 3v3 for sure
[18:56:39] <Vutral> but its 5 volt tolerant mhm
[18:56:43] <OSterver> yes
[18:56:46] <OSterver> so it must be @ 5v
[18:56:49] <Vutral> anyway there is 3v3 volt output pin
[18:56:52] <OSterver> but 5v should die
[18:56:52] <Vutral> no idea what they use it for
[18:56:59] <OSterver> hopefuly soon
[18:57:06] <Vutral> better get one of these new avr
[18:57:10] <Vutral> they run at 0,7 volt
[18:57:13] <Vutral> 3,3 volt should die :P
[18:57:22] <OSterver> lol
[18:57:34] <Vutral> really nice integrated boost converter :P
[19:00:46] <Vutral> i want samples :P
[19:17:35] <OSterver> nice
[19:17:41] <OSterver> it works with 3v3 logic
[19:17:59] <OSterver> took me only 15 minutes to hotwire two power supplies and the display on breadboard
[19:18:03] <OSterver> and manually enter commands :P
[19:33:44] <OSterver> dear datasheet makers
[19:33:45] <OSterver> http://clip2net.com/s/2iy78
[19:33:47] <OSterver> THIS IS NOT READABLE
[19:35:12] <Tom_itx> sry i don't think they speak english
[19:35:32] <OSterver> heh
[19:35:33] <OSterver> ALPS
[19:35:36] <OSterver> is the author of this mess
[19:35:55] <Tom_itx> oh hell that one is easy
[19:36:07] <OSterver> too much useless numbers
[19:36:18] <OSterver> I want pads area
[19:36:22] <OSterver> where should be pads to solder.
[19:36:23] <Tom_itx> well they didn't leave anything out
[19:36:27] <OSterver> they did not
[19:36:34] <OSterver> but they left in too much for me.
[19:37:39] <Tom_itx> it's all referenced from the x y plane which is rather unusual
[19:38:21] <Tom_itx> generally you gotta do alot of calculation to get some of the dimensions
[19:38:26] <CapnKernel> It's pretty feral with the X dimensions in the lower left :-)
[19:39:32] <OSterver> would be cool if there was footprint for it in eagle
[19:39:36] <OSterver> I don't feel like making it
[19:40:13] <Tom_itx> awww
[19:40:21] <CapnKernel> I don't know why programs such as Eagle and KiCad don't let you import a pic, then you plonk down pads, etc, then you can take the pic away
[19:40:40] <CapnKernel> Doesn't have to be micrometer accurate, just as long as the pads line up
[19:40:58] <CapnKernel> (Caveat: Drawing has to be to scale)
[19:41:17] <OSterver> write ulp for it :)
[19:41:23] <OSterver> there is already bitmap->silkscreen converter
[19:41:24] <Tom_itx> what are the dark gray pads opposed to the lighter ones?
[19:41:25] <jadew> ok, I had time to play with the open bench logic sniffer, the recommended software is garbage
[19:41:27] <OSterver> I could do it by that
[19:41:31] <OSterver> and just change it to write to TOP layer instead
[19:41:41] <Tom_itx> jadew all you had to do is ask
[19:41:51] <OSterver> but the fact that it is not layout thing but it contains junk I can not do it that way
[19:42:01] * CapnKernel uses KiCad, so no ULP here
[19:42:02] <jadew> Tom_itx, heh, yeah, I thought it would actually work
[19:42:04] <Tom_itx> jadew one reason i got the saleae
[19:42:12] <CapnKernel> Coming soon in KiCad: Python :-)
[19:42:24] <OSterver> I will go for this lol
[19:42:25] <OSterver> http://cz.mouser.com/ProductDetail/Yamaichi-Electronics/PJS008U-3000-0/?qs=sGAEpiMZZMtaouNQoacPIDi8JGMbrqPVWB4XP%252brLbLM%3d
[19:42:31] <OSterver> much easier
[19:42:37] <jadew> I hit analyze, it screws up SPI bytes, but it gets them
[19:42:47] <jadew> second time I hit analyze it only sees CS up / CS down
[19:43:23] <jadew> the device seems fine tho
[19:43:25] <Tom_itx> a tester shouldn't introduce error
[19:43:46] <jadew> that's very true
[19:44:01] <jadew> kinda dissapointed in the whole deal
[19:45:13] <OSterver> heh I have just noticed
[19:45:14] <OSterver> http://clip2net.com/s/2iy8t
[19:45:18] <OSterver> when I was testing out strings
[19:45:30] <OSterver> for some reason it was crapping up the string :/
[19:45:49] <jadew> maybe you forgot a 0 between the chars
[19:45:52] <OSterver> nope
[19:45:54] <OSterver> it was all unicode
[19:45:55] <jadew> it's UTF16 IIRC
[19:46:05] <OSterver> dunno, it works with L"somestring"
[19:46:27] <jadew> that should do it
[19:46:39] <OSterver> oh well, I just shortened it :)
[19:47:04] <OSterver> deleted revision and XBoot version from it
[19:47:04] <jadew> did you specify the right size?
[19:47:07] <OSterver> yes
[19:47:11] <OSterver> dunno, I don't care now
[19:47:18] <OSterver> I am glad it work
[19:47:18] <OSterver> s
[19:47:25] <OSterver> upcoming month = CDC implementation.
[19:47:26] <OSterver> lol
[19:48:09] <jadew> I think I'm gonna write a logic analyzer software for this open bench sniffer
[19:48:40] <jadew> this experience reinforces my belief that opensource is generally garbage, written by noobs with nothing better to do
[19:49:47] <jadew> I'm amazed at the ammount of bugs I found just by trying to do the most basic thing it has to do
[19:50:19] <jadew> how the hell do you screw up so bad? and how can you release something with out catching such huge bugs?
[19:51:58] <jadew> not to mention that implementing the software part for a logic analyzer is freaking straight forward
[20:04:58] <OSterver> bloody hell, does any SD card socket on mouser fit the footprint from sparkfun?
[20:05:16] <Tom_itx> no
[20:05:23] <Tom_itx> i've made 3 or 4 myself
[20:06:09] <OSterver> hmm could you please share? :)
[20:08:04] <w|zzy> jadew. what sort of bugs? I get the occasional crash cause it doesn't handle the serial lines well.
[20:08:05] <OndraSter> hmm I found some libs on cadsoftusa
[20:08:06] <OndraSter> let's see
[20:09:02] <jadew> w|zzy, not decoding properly, randomly refusing to decode at all until restart
[20:09:12] <jadew> decoding the same thing twice and getting different data
[20:10:09] <jadew> w|zzy, what version are you using?
[20:10:54] <jadew> we are talking about OLS, right? not the original SUMP client (which doesn't work on my pc)
[20:17:47] <OndraSter> :( the readable datasheet is for 1.9€ part
[20:17:48] <OndraSter> oh well
[20:21:35] <OndraSter> scrap the idea, I will not have SD card logging
[20:21:37] <OndraSter> I don't need it
[20:21:50] <OndraSter> that's 1 hour wasted ..
[20:22:21] <OndraSter> KISS will be better
[20:24:17] <w|zzy> ill have to check at home. I'm using the old unofficial client. it was hard to find
[20:29:09] <w|zzy> jadew http://www.lxtreme.nl/ols/ ?
[20:29:28] <jadew> yeah, that's the one I'm using as well
[20:29:32] <OndraSter> lol I have just noticed... my thermocouple controller is 0 - 4V maximum
[20:29:34] <OndraSter> aka 3v3 operation
[20:29:37] <OndraSter> I'd have burned it
[20:29:47] <jadew> w|zzy, 0.9.5
[20:30:07] <w|zzy> Yup. its better than the original if that matters
[20:31:19] <jadew> yeah, I figured
[21:50:22] <rue_house> anyone have a bit bang library for either i2c or 3wire I can use?
[21:51:52] <OndraSter> HOLY CRAP IT IS 0439!
[21:52:47] <jadew> rue_house, I can give you my sniffer code, which you can adapt for receiving
[21:52:55] <jadew> actually wait
[21:53:00] <jadew> that won't work
[21:53:04] <jadew> you need master code
[21:53:19] <jadew> it should be easy to implement
[21:56:16] <w|zzy> jadew what are you going to write it in?
[21:56:45] <jadew> w|zzy, the analyzer?
[22:02:57] <w|zzy> yes
[22:03:03] <jadew> c++
[22:03:13] <w|zzy> Cross platform?
[22:03:20] <jadew> already started writting a wrapper for the sniffer
[22:03:41] <jadew> not sure, at least not at first
[22:04:06] <w|zzy> Cant use a cross platform framework, aka Qt?
[22:04:29] <w|zzy> Though it doesnt have serial capabilities
[22:04:36] <jadew> I'm not familiar with it, I only used it for a bit in the past
[22:05:24] <w|zzy> Worth a look.. The only two things i can see that could be a problem are there isnt a serial or graphing library included..
[22:05:24] <w|zzy> there are quite a few qt based graphing libraries though
[22:05:45] <jadew> I already rolled my own serial handler
[22:05:57] <jadew> cross platform, based on boost::asio
[22:06:11] <jadew> so that's not an issue
[22:06:17] <w|zzy> :)
[22:06:30] <jadew> the graphics shouldn't be pre-generated anyway, it needs to be owner drawn
[22:06:39] <jadew> otherwise it will eat tons of memory and it will be slowish
[22:06:57] <w|zzy> yup
[22:07:48] <jadew> oh, I remember why I stopped working with Qt
[22:08:11] <jadew> doesn't the free version forces you to opensource your code?
[22:17:09] <w|zzy> No.
[22:17:15] <w|zzy> Only if you staticly compile it
[22:17:34] <w|zzy> Would that matter for this project?
[22:17:43] <jadew> not really
[22:18:14] <w|zzy> It is lgpl
[22:18:18] <jadew> I'll decide which route I'll take after I finish the wrapper for the hardware
[22:18:56] <w|zzy> Fair.
[22:19:12] <w|zzy> If you need someone to test it, let me know.
[22:19:36] <jadew> sure
[22:19:42] <jadew> you're on linux, right?
[22:20:13] <w|zzy> and windows.
[22:20:21] <jadew> cool
[22:20:33] <w|zzy> What are you on?
[22:20:46] <jadew> both, but windows is my main dish
[22:21:26] <w|zzy> Same. Until someone can tell me an adequate replacement for atmel studio's(avr studio) debugging interface it will remain that way
[22:21:43] <w|zzy> Well that and games..
[22:21:58] <OndraSter> hehe
[22:21:59] <jadew> and a decent IDE
[22:22:46] <w|zzy> sublime text is cross platform! Eclipse isnt half bad once you work your way around it.
[22:23:58] <jadew> eclipse makes my pc seem like it's a 486 trying to run windows xp
[22:24:43] <jadew> I just use a text editor (geany) when I'm under linux
[22:25:10] <jadew> for bigger projects I go with code lite
[22:25:37] <OndraSter> <w|zzy> sublime text is cross platform! Eclipse isnt half bad once you work your way around it.
[22:25:44] <OndraSter> except you can never work your way around it
[22:25:49] <jadew> hehe
[22:25:53] <w|zzy> Being a gamer... My computers are pretty well spec'd
[22:25:59] <OndraSter> so is mine
[22:26:04] <OndraSter> and java anything is pure crap :)
[22:26:09] <OndraSter> I am off to "sleep"
[22:26:11] <OndraSter> it is 0513
[22:26:12] <OndraSter> lol
[22:26:12] <jadew> my pc is pure awesomeness
[22:26:13] <w|zzy> Night..
[22:26:16] <OndraSter> so is mine
[22:26:18] <jadew> night
[22:26:20] <w|zzy> Maybe you'll be more rational when you have slept :P
[22:26:22] <OndraSter> 4.5GHz quad Intel, 16GB RAM
[22:26:23] <OndraSter> :D
[22:26:38] <jadew> nice
[22:26:53] <jadew> mine is 3.3Ghz, 6 core AMD, 16GB
[22:27:01] <OndraSter> 6core amd = 2core intel :P
[22:27:05] <OndraSter> "gn"
[22:27:05] <jadew> lol
[22:27:08] <jadew> night
[22:27:14] <w|zzy> i7 950 + 12gb here.. Old by comparison but more than good enough
[22:27:38] <jadew> yeah, the thing is there's no excuse for an IDE to be slow on today's PCs
[22:27:43] <w|zzy> Also.. Eclipse loads and runs faster than AS6!
[22:28:14] <w|zzy> Im running ssd system disk + dual WD blacks in raid0 for storage.
[22:28:44] <jadew> I'm not that savvy when it comes to loading speed
[22:28:54] <jadew> I only have a raid 1 for my work
[22:29:00] <jadew> so I have a backup
[22:29:34] <w|zzy> I keep everything backed up elsewhere.
[22:29:52] <jadew> well, I have an automated backup system in place as well
[22:30:09] <jadew> every morning at 6am all the code gets compressed, crypted and uploaded somewhere :P
[22:30:36] <w|zzy> Do you code for a living?
[22:30:39] <jadew> yeah
[22:30:51] <w|zzy> Fair enough
[22:31:11] <jadew> and I'
[22:31:19] <jadew> I've lost many hdds lateley
[22:31:38] <jadew> it's like they're making them crappier each year
[22:32:46] <w|zzy> They are!
[22:33:02] <w|zzy> I have a raid5 setup in the server.
[22:33:33] <w|zzy> Plus everything important is sync'd across my laptop, desktop and server. Especially important things are on a mates server off site!
[22:33:44] <w|zzy> Which isnt much
[22:33:54] <w|zzy> but i'm just a dirty electrician!
[22:34:25] <jadew> heh, you can never have enough redundancy when it comes to your files :)
[22:47:27] <w|zzy> Hmmm I prefer the speed on my desktop
[22:47:44] <w|zzy> YOu can feel the benefits of the wd blacks when running VMs