Back
[00:11:19] <gorroth> well, boot_rww_enable_safe()
[07:04:20] <Lambda_Aurigae> http://hackaday.com/2015/10/28/we-didnt-know-the-sun-could-do-digital/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+hackaday%2FLgoM+%28Hack+a+Day%29
[07:04:27] <Lambda_Aurigae> a solar clock..and they didn't even use an ardweeny!
[07:12:11] <day> Lambda_Aurigae: they deserve a medal for that!
[07:23:56] <Lambda_Aurigae> yeah..it is impressive.
[07:24:10] <Lambda_Aurigae> getting something posted on hackaday without using an ardweeny.
[07:27:01] <Lambda_Aurigae> the sundial is impressive too...some interesting design work.
[08:20:21] <Jartza> https://drive.google.com/file/d/0B2dTzW9TMeBxZjFpZTlKNUVGUmc/view?usp=sharing
[08:20:23] <Jartza> https://drive.google.com/file/d/0B2dTzW9TMeBxb3l2NUNhTWFZSkU/view?usp=sharing
[08:21:25] <Tom_itx> what's it for?
[08:22:31] <Tom_itx> something vga apparently
[08:25:21] <mike_papa> Hello. Any idea why I can't see Cycle Counter (it's grayed) in Atmel Studio 7, using Dragon to debug Atmega2560 via JTAG?
[08:25:36] <mike_papa> Is it available only in simulator?
[08:28:46] <Jartza> Tom_itx: yea. octapenta, like attiny85
[08:28:57] <Jartza> it's a 32x14 character VGA with 8 colors
[08:29:08] <Jartza> 9600bps UART to feed data in, supports subset of ANSI-escapes
[08:30:13] <Jartza> here's previous version assembled
https://drive.google.com/file/d/0B2dTzW9TMeBxaE5oV29UMDhFV1E/view?usp=sharing
[08:30:52] <LeoNerd> Cute... tiiiiiny
[08:31:41] <Jartza> and here's demovideo, data fed from python-script
https://drive.google.com/file/d/0B2dTzW9TMeBxWV9LSjZBRmliVkU/view
[08:33:42] <Jartza> and if BW-picture is enough,
https://drive.google.com/file/d/0B2dTzW9TMeBxaFFxam1uVW05NlE/view
[08:33:55] <Jartza> https://drive.google.com/file/d/0B2dTzW9TMeBxbmJTbnllVU5qb1E/view
[08:34:38] <Jartza> or whatever you wish it to be, black/green, black/red, black/blue, black/amber :)
[08:34:46] <Jartza> that can be adjusted with 3 resistors
[08:37:02] <Jartza> I just wish I had some time to finish my blog-post about this
[08:37:14] <Jartza> then I could send this to hackaday or smthing :)
[14:53:38] <gorroth> Jartza: cool!
[14:56:01] <gorroth> so, my ATmega168 runs at 8 MHz with an external crystal, and its setup with 8N1 at 38.4kbps. i fixed my PC code so that after each byte it sends, it will do a tcdrain(), which waits for the remote (AVR) side to pick it up before sending the next byte, which has fixed any problems transmitting data to the AVR. however, if the AVR sends a byte back and then sends another byte back as soon as UDRE is set, occasionally the PC will receive a corrupted byte.
[14:56:21] <gorroth> the only resolution i've found it to put a small _delay_ms() after each transmission on the AVR
[14:57:05] <gorroth> is this a common problem that has a solution i don't know about? i've done a lot of googling with no success. people's usual problem is that they are using the internal RC oscillator, which i am not
[14:57:15] <LeoNerd> tcdrain() doesn't do that
[14:57:40] <gorroth> tcdrain() waits until all output written to the object referred to by fd has been transmitted. (from manual)
[14:57:43] <gorroth> either way, it makes it okay
[14:57:52] <gorroth> my focus is on the AVR sending data now
[14:57:52] <LeoNerd> "has been transmitted" is an amusing phrase there :)
[14:58:04] <gorroth> yes, it doesn't directly imply what i said
[14:58:18] <gorroth> but it seems to make things fine for that side; i'm more concerned about the opposite comms direction now
[14:58:25] <LeoNerd> Sending bytes shouldn't care about UDRE
[14:58:57] <gorroth> yes, it should. i shouldn't modify UDR until it has sent data. that's per the datasheet
[14:59:04] <LeoNerd> You want to check the transmit status
[14:59:07] <gorroth> well, until it is empty
[14:59:10] <gorroth> i did that too
[14:59:15] <gorroth> after sending the data, i check TXC
[14:59:24] <gorroth> it didn't have an effect though
[14:59:25] <LeoNerd> TXC0
[14:59:32] <gorroth> yes
[14:59:49] <LeoNerd> There's two main approaches; either wait for TXC0 after you send a byte, or before you send one. I actually prefer the latter approach
[15:00:10] <LeoNerd> It makes transmission -slightly- more efficient because you don't have to wait for the final byte to transmit; the USART can get on with that while you do other things
[15:00:18] <gorroth> it is possible my atmega could be broken, since i also had problems with my bootloader until i moved the boot_rww_enable_safe() from the function that does the erase/fill/write to just before i do my ijmp
[15:00:31] <gorroth> hmm
[15:00:36] <gorroth> okay
[15:00:49] <LeoNerd> Also, be aware that an 8MHz crystal isn't going to give you timing-perfect baud rates
[15:00:50] <gorroth> the datasheets said to wait for UDRE and check it before sending the next byte
[15:01:08] <gorroth> yes, i know, but for the baud rate i'm using, it should result in 0.2% error, according to the datasheet
[15:01:10] <LeoNerd> When I'm using serial I usually use a 14.7456 xtal, as that divides perfectly for common rates
[15:02:06] <gorroth> yeah. i didn't actually know about how to use the UART before i bought the 8 MHz crystal, and i bought that crystal because it made timing 1 us very easy, which results in easy 16-bit PWM calculations for servos
[15:02:13] <LeoNerd> Mmm :)
[15:02:27] <LeoNerd> Yeah; there's often competing concerns here, so you have to accept a compromise one side or other
[15:02:48] <gorroth> yeah, but since i bought the 8 MHz crystal, i've been compromosing on the UART side, hehe :)
[15:02:56] <gorroth> compromising*
[15:03:56] <gorroth> i have this feeling that had i known about UART timing concerns, i would've bought a different crystal, and UART would probably/hopefully be better, and the servo signal could accept more slack than the UART maybe
[15:04:16] <gorroth> i might just buy a new crystal anyway and see if the problem goes away
[15:04:26] <gorroth> or try my "backup" atmega168 and see if the problem goes away
[15:04:40] <gorroth> because i might just have a malfunctioning m168
[15:05:10] <LeoNerd> It's always possible :)
[15:05:32] <gorroth> LeoNerd: b/c as long as both sides have same serial settings and the mega has a 0.2% error rate, parity detection should catch it, right? or is it something that gets set in a register that my AVR code should catch so that i can resend that data?
[15:07:02] <LeoNerd> Mm? At the UART layer, transmitters never know if their transmissions were correctly received
[15:07:10] <LeoNerd> That's why higher-level protocols exist, to resend it
[15:07:13] <gorroth> yes, i realized that after i hit enter :-P
[15:07:43] <gorroth> that's true. i wonder if i need to make my bootloader more sophisticated
[15:08:03] <gorroth> so i can have the PC tell the AVR to resend the last byte, in case it saw something funky
[15:08:10] <LeoNerd> Presuming a worst-case error on the other side of 1%, you could have up to 1.2% of a timing skew per bit. 1 start, 8 data bits, no parity, 1 stop gives you 10 bits of skew, so the end of the STOP bit might be up to 12% off
[15:08:14] <LeoNerd> That's 1/8th
[15:08:34] <gorroth> dang
[15:08:48] <LeoNerd> The AVR UART (when not in X2 mode) uses a 0..15 counter, sampling the input on each of 7/8/9th counts
[15:09:05] <LeoNerd> So the AVR receiver would have to be quite a bit more off in timing than that, to miss the data
[15:09:16] <gorroth> do you think i should try to use x2 mode? i think it also has a 0.2% error rate for 38.4 kbps
[15:09:27] <gorroth> i haven't looked closely enough to see how x2 really works
[15:09:37] <LeoNerd> X2 makes it worse
[15:09:39] <gorroth> like if the PC would still expect a 38.4 kbps rate or whatever
[15:09:40] <gorroth> ah
[15:09:46] <LeoNerd> In X2 mode it only counts 0..7, sampling on 3/4/5
[15:09:58] <gorroth> well dang, i don't want that then :)
[15:10:08] <LeoNerd> You generally want to avoid X2 mode if you can
[15:10:18] <LeoNerd> Er.. U2X.. to give its actual name :)
[15:10:59] <LeoNerd> Take a look at the data sheet, in the UART chapter, under the heading "Asynchronous Clock Recovery". It's 20.8.1 in my sheet, though of course your version might differ
[15:11:23] <gorroth> it's in the same location on my sheet
[15:12:53] <gorroth> LeoNerd: i saw that section last night but didn't go into detail reading it, since it looked like it was automatic
[15:13:50] <LeoNerd> In debugging an issue with a piece of hardware, it always helps to have as complete and detailed a mental model of how that hardware works
[15:14:12] <LeoNerd> If you're occasionally suffering byte-receive errors, it's useful to know how the UART receiver recovers the clock
[15:14:33] <gorroth> yeah, i know. i'm just new. it's honestly just taking a while to absorb the info is all
[15:15:00] <gorroth> i was very focused on the bootloader itself failing until late last night (very early this morning)
[15:15:15] <gorroth> so i hadn't started swinging back around to the uart until now
[15:15:27] <gorroth> since i have a hack in place that works for the moment
[15:15:42] <LeoNerd> It can be tricky yes, with lots of moving parts trying to work out which part isn't moving right :)
[15:15:49] <gorroth> oh, yes, i knw
[15:15:52] <gorroth> it's a PITA
[15:16:11] <LeoNerd> I might suggest you write a trivial little echo program. On receipt of a byte on the UART, immediately reflect it back and do nothing else
[15:16:12] <gorroth> it'll be easier going forward, as i build more confidence in my code, to know if it's a hardware or software problem
[15:16:18] <gorroth> i have an echo program
[15:16:19] <LeoNerd> Use that to test if your chip UART and computer are all talking OK
[15:16:20] <gorroth> i use that for testing
[15:16:33] <gorroth> the AVR can lose data if i type a lot of keys too quickly on minicom
[15:16:40] <LeoNerd> That's upsetting :/
[15:16:50] <gorroth> hehe
[15:16:54] <LeoNerd> Do you have a logic analyser to hand? It's at this point I'd definitely be applying mine
[15:17:02] <gorroth> no :(
[15:17:11] <gorroth> and i looked at buying one, and i'm not ready to spend that much just yet
[15:17:15] <LeoNerd> Ah.. Get one :) There's cheap ones for .. hm.. I got mine for £14
[15:17:21] <twnqx> just use 1mbit serial
[15:17:29] <gorroth> but i do have a cheaper test for the moment, which is swapping out the m168 for another m168 i have
[15:17:29] <LeoNerd> It's not *great* but it's certainly good enough to talk 38kBaud serial
[15:18:16] <LeoNerd> (I might get a real Saleae Logic8 one day, but I'm waiting until I actually need it as compared my cheap one, and also hopefully Sigrok will like it by then)
[15:18:54] <gorroth> twnqx: using 1 mbit serial will require a lot of effort on the linux side, which is why i haven't tried anything like that. apparently 250 kbps runs with 0.0% error on the pc, but linux just goes up to B230400 without extra work
[15:19:15] <gorroth> when i said "on the pc" for 250kbps, i meant "on the avr"
[15:19:30] <LeoNerd> Lately, I largely don't bother with using UART for PC <-> device... USB-CDC seems just as good
[15:19:37] <gorroth> LeoNerd: in december is going to be a better time for me to get a logic analyzer
[15:19:42] <LeoNerd> If you're not *actually* UARTing it in the device itself then it's bitwise-perfect
[15:20:35] <gorroth> LeoNerd: let's cross some fingers that tonight, when i'm back at home around chicago, swapping devices results in improvement :)
[15:20:47] <LeoNerd> Mmm
[15:20:49] <LeoNerd> Good luck :)
[15:20:52] <gorroth> but otherwise i'll start doing some of these other things you're talking about, minus the logic analyzer
[15:21:02] <gorroth> that both you and twnqx are talking about, at least
[15:21:10] <LeoNerd> I had a bizarre UART timing bug on a board of mien a while ago, that turned out to be bad soldering on the timing crystal
[15:21:25] <LeoNerd> Just occasionally, the xtal waveform would flip all wrong, and upset the counter
[15:21:31] <gorroth> i could also swap my crystla, as i bought 2 of them
[15:21:35] <twnqx> effort?
[15:21:38] <gorroth> because, you know, for like a few cents, why not
[15:21:52] <LeoNerd> Turned out that the inrush current draw on flipping the TX output pin was the most likely thing to upset it
[15:21:55] <twnqx> the only thing you have to consider is to stop bytewise reading
[15:22:04] <gorroth> twnqx: yeah, to go above the kernel in-built baud rates, i'll have to do much extra work
[15:22:07] <twnqx> because the few µs userspace<->kernel latency make you lose data
[15:22:11] <twnqx> uh
[15:22:15] <twnqx> 2mbit IS kernel built-in
[15:22:20] <LeoNerd> So what would happen is that the timer was nicely stable until you flip the START bit, then the counter gets all upset for about 10 counts, before it recovers... the the end of the START bit into the first data bit upset it again
[15:22:29] <gorroth> twnqx: oh, well, termios doesn't support it
[15:22:31] <LeoNerd> A quick bit of resoldering on the xtal solved that one right up :)
[15:22:43] <LeoNerd> Yeah termios *really* sucks at arbitrary baud rates :(
[15:22:49] <LeoNerd> I really wish it had been an int-valued iotcl
[15:22:50] <twnqx> minicom does work though
[15:22:51] <LeoNerd> ioctl
[15:23:02] <LeoNerd> But no.... it has to be a stupid-ass collection of bitmasks out of a fixed enumeration
[15:23:16] <gorroth> twnqx: hmm, that's true. i could use minicom with the echo code to see if it's good at higher baud
[15:23:17] <LeoNerd> .oO( termios sucks at lots more than this, but that's one on my list )
[15:23:41] <gorroth> LeoNerd, twnqx, what do you guys recommend over termios? that's just all i really knew about for a quick piece of code being put together
[15:23:57] <LeoNerd> I don't. That's the problem
[15:24:02] <gorroth> haha
[15:24:07] <LeoNerd> It sucks but it's the only way to talk to the kernel about this
[15:24:20] <LeoNerd> There just does not exist a nicer system
[15:24:20] <gorroth> i've read about stty to setup a new tty device at the baud rate i likee, but then i still have to do some work in termios to use it
[15:24:32] <gorroth> there are some examples online, but i haven't gone through them yet
[15:24:35] <LeoNerd> stty(1) is just a userland wrapper around the termios(3) kernel ability
[15:24:53] <LeoNerd> Well, rather, a commandline wrapper
[15:25:13] <twnqx> just use B1000000 or B2000000 as parameter, just as you did with B230400?
[15:25:27] <twnqx> of course your serial port chip has to support it :)
[15:25:43] <gorroth> twnqx: my man page only shows up to B230400, which makes me think they don't have a macro higher than that. i can check really quickly though
[15:25:58] <twnqx> i took those from my code :P
[15:26:40] <gorroth> #define B1000000 0010010
[15:26:42] <gorroth> it is there
[15:26:47] <gorroth> the man page just didn't talk about it
[15:26:49] <gorroth> sweeeeet
[15:26:58] <gorroth> thanks for making me check, twnqx :)
[15:27:11] <gorroth> shoot, i'll just try that baud rate out later
[15:27:41] <twnqx> but like i said
[15:27:57] <twnqx> you.. can't really read bytewise
[15:28:13] <gorroth> twnqx: you mean i should use read() to get 1 byte at a time, right?
[15:28:16] <twnqx> took me a while to realize that my lost characters came from kernel-userspace overhead with bytewise reading :D
[15:28:22] <twnqx> no, you can't any more
[15:28:29] <twnqx> you need to implement buffering of blocks
[15:28:40] <twnqx> read 128byte at once, consume what's there
[15:28:47] <gorroth> twnqx: you mean i should have a buffer of, say, 64 bytes, and use read() to get all 64?
[15:28:51] <twnqx> yes
[15:28:53] <gorroth> okay
[15:28:57] <gorroth> i will do something like that
[15:29:07] <gorroth> i just thought my PC being so much faster, it would be fine
[15:29:26] <gorroth> but first i'll upgrade the baud rate and see what happens
[15:29:34] <gorroth> then i'll get smarteer and implement what you said, even if it works
[15:30:15] <gorroth> well, thanks all. this was a very helpful conversation.
[16:08:18] <twnqx> gorroth: i benmarked the kernel<->userspace latency. while it's just a few dozen µS, it's enough to limit bytewise reads so some thousand calls per second
[16:10:06] <jacekowski> kernel<->userspace latency is a lot better than it used to be
[16:10:51] <jacekowski> now that cpu can do quit fast sysenter/sysexit rather than old int 80 method
[16:21:28] <gorroth> twnqx: oh, i'm going nowhere near that fast. writing to the flash really slows things down. however, i like your idea in general and am going to implement it
[16:22:09] <gorroth> jacekowski: what's the sysenter/sysexit method? i didn't realize they'd moved away from int 80
[16:26:58] <twnqx> jacekowski: this is an i7 haswell 4770k
[16:27:22] <twnqx> and i actually ran into problems dumping data from a 500kbit/s CAN bus
[16:27:40] <twnqx> took me LONG to realize where the bytes were lost
[16:28:46] <twnqx> gorroth: amd and intel both created new methods to replace the legacy software interrupt method
[16:29:01] <twnqx> syscall and sysenter/sysexit, respectively
[16:30:10] <gorroth> oh sweet! some basic asm pneumonics i guess
[16:30:12] <gorroth> i'll check it out
[16:30:26] <twnqx> mnemonics, please :P
[16:30:54] <gorroth> well, you know what i mean :)
[17:06:26] <osteri> damn google, i recently discovered that if you block cookies from accounts.google.com you can't even use gmail web interface
[17:07:26] <osteri> that means you are always logged in, in every google's services via web
[17:09:07] <osteri> i just want to log in when i use the service, not being logged 24/7
[19:38:16] <apo_> osteri: 2nd profile for gmail, or just use a proper client
[19:38:31] <apo_> osteri: or private browsing =P
[19:38:49] <apo_> osteri: or another browser!
[19:45:18] <osteri> apo_: 2nd profile; no thanks, proper client; yeah imap maybe, private browsing; makes no sense, another browser; why?
[19:46:03] <osteri> i just block accounts.google.com and start using imap
[19:46:31] <apo_> private browsing deletes cookies when you're done
[19:46:39] <apo_> another browser keeps the cookies away from your normal browsing
[19:46:44] <apo_> 2nd profile does the same
[19:46:59] <apo_> and with a real client you don't get cookies at all =P
[19:47:09] <osteri> oh, i didn't think of that another browser thing
[19:47:46] <osteri> apo_: but you still have to allow cookies to use gmail
[19:48:00] <osteri> best way maybe just use imap
[19:50:03] <osteri> easiest way is to get rid of google's services, but those are temporary solutions until i find new email provider
[19:50:21] <apo_> I do my own email
[19:51:16] <osteri> apo_: i don't want to wrestle with spam filter etc, it's a full day job and when something goes wrong, you loose all your shit
[19:51:46] <osteri> i could pay for good mail service
[19:51:53] <apo_> osteri: spamassassin does a pretty good job with its default config
[19:52:06] <Casper> just use thunderbird and imap google
[19:52:19] <Lambda_Aurigae> you could go to outlook365.....
[19:52:19] <osteri> Casper: yes, i'm planning to
[19:52:36] <apo_> osteri: and unless your server actually rejects the mail, the sender will just retry for a few days
[19:54:14] <osteri> apo_: i know one provider whose mails were rejected by Google, there are many problems with having your own mail server
[19:55:09] <Lambda_Aurigae> osteri, yeah,,,done that, been there, bought the book, read the t-shirt.
[19:55:10] <osteri> or there could be, i don't want to maintain anything more than i'm maintaining now
[19:55:22] <Casper> osteri: are you talking about videotron? :D
[19:55:30] <osteri> Lambda_Aurigae: yup
[19:55:34] <apo_> *shrugs*
[19:55:38] <Lambda_Aurigae> I switched to gmail hosted about 13 years ago.
[19:55:39] <apo_> Not trying to convert you
[19:55:49] <osteri> Casper: no, you don't know, small company
[19:55:51] <apo_> I'm just saying that it's been working fine for me for like six years
[19:56:17] <Casper> osteri: ok, because videotron get regularly blocked by google and many server, because they don't filter anything comming out...
[19:58:31] <osteri> Casper: yeah, IIRC i think Google chose some protocol in their mailing system that small mail providers didn't like
[19:59:06] <osteri> so that the protocol benefits only big mail providers and thinks small providers are sending spam, or something like that
[19:59:13] <Casper> I think it's due to the high volume of mail compared to what they get
[19:59:37] <Casper> like if they get more than x% of mail from a single provider, then they think it's a flood or something like that
[20:01:19] <osteri> yeah, but Google can do whatever they want - choose their own protocols and if you're not with them, you're out of biz
[20:01:27] <Casper> nahhh
[20:01:41] <Casper> it's not a protocol issue, it's a volume issue
[20:02:21] <osteri> ok, i belive you
[20:09:19] <Valen> google hasn't invented any protocols for mail
[20:13:13] <osteri> i couldn't find the news about the case, but now i think about it; it wasn't a protocol, it was something like: they refused to use some spam filter because Google of might block them
[20:14:21] <osteri> so they are receiving more spam, because of Google. therefore Google is indirectly degrading their service
[20:14:32] <ali1234> yeah, no
[20:15:02] <ali1234> google is picky about DKIM and SPF
[20:15:14] <ali1234> that's about it
[20:15:27] <osteri> i think it was about SPF
[20:15:40] <ali1234> google never blocked someone for NOT sending email
[20:16:01] <Casper> osteri: I think you got it wrong, if the compagny don't use spam filter, it mean that their outgoing mail contain lots of spam, which get flagged, and the domain reputation at google drop and google may stop accepting mail from them. THIS IS STANDARD FOR MOST BIG PROVIDER AND ISP.
[20:16:09] <ali1234> and if you don't know how to set up SPF you shouldn't be running an email server
[20:16:32] <ali1234> /rant
[20:16:41] <osteri> Casper: yeah, something like that
[20:17:17] <osteri> but despite anything, google is evil in my eyes
[20:17:27] <Casper> osteri: if this is the case, then the "isp" has been rightfully blocked as it should
[20:18:07] <Casper> every mail server should do some spam filtering... imo, a server that don't do it should be blacklisted
[20:18:14] <osteri> if i can't use your web mail without privacy invading cookies, f u and your service
[20:18:43] <Casper> osteri: then unsubscribe to the internet
[20:19:03] <osteri> Casper: i don't care if some random site uses cookies
[20:20:47] <osteri> so if i go to example.com and it has Google Analytics, i assume they can fully track me with those accounts.google.com cookies
[20:20:58] <osteri> that is the problem in my eyes
[20:21:08] * Casper wonders what is the minimal power needed for a small house that have gas...
[20:21:48] <Casper> yeah that one is a bit invasive, however I prefer that google have that in their hand than doubleclick or worst...
[20:22:36] <osteri> and if i have to allow those cookies to use gmail web client, that means i'm being tracked 90 % of time, sounds scary
[20:23:15] <osteri> but yeah, solution is already there; get rid of Google
[20:23:28] <Casper> osteri: then just use a non-web based client. end of the story. Next topic!
[20:23:40] <osteri> just how, their services are good :)
[20:24:03] <osteri> Casper: yeah, i started using since i understood that
[20:24:17] <Lambda_Aurigae> I use gmail, their web client, their search engine, and their dns servers..track me baby!
[20:24:25] <Valen> I don't do outbound spam filtering, but then the hosts I administer aren't compromised ;-P
[20:24:26] * Casper wonders if 120V 30A is enought...
[20:24:48] <Lambda_Aurigae> Casper, depends on what all you need to power.
[20:25:08] <Valen> that's like one clothes dryer and half a toaster isnt it?
[20:25:14] <Lambda_Aurigae> pretty much.
[20:25:22] <Lambda_Aurigae> or two copiers.
[20:25:22] <Valen> standard in .au is 60A 240v
[20:25:26] <Casper> Valen: there is nat gas dryer
[20:25:43] <Casper> Valen: 240V 30A here
[20:26:10] <Lambda_Aurigae> 240V 100A service to my house...split into two 110V legs.
[20:26:15] <Valen> what are 110v hand held appliences rated at amp wise?
[20:26:23] <Valen> .au has 240v mains
[20:26:25] <Lambda_Aurigae> depends on the appliance...
[20:26:35] <Valen> and things like hair driers are 2300W so ~10A
[20:26:39] <Lambda_Aurigae> hair dryer is usually around 1000W so, 10A
[20:26:45] <Casper> Lambda_Aurigae: 120/240 150A here, panel is full, will replace by a 120/240 200A in a few years
[20:27:11] <Valen> I am glad we settled on 240v lol
[20:27:14] <Casper> Valen: outlets here are 120V 15A, 80% loading = ~1500W max
[20:27:31] <Valen> I reckons eventually the world will move to like 400vdc
[20:27:40] <Lambda_Aurigae> Casper, no 20A outlets? my house is split between 10A, 15A, and 20A outlets.
[20:27:53] <Valen> anything other than a 10A outlet is *very* unusual here
[20:27:58] <Valen> only really see them in industry
[20:28:08] <Valen> anything else is hard wired
[20:28:11] <Lambda_Aurigae> you do have higher overall power though with the 240V feed.
[20:28:26] <Lambda_Aurigae> 240V 10A is like my 120V 20A outlet for overall power.
[20:28:35] <Casper> Lambda_Aurigae: new construction have a 20A outlet on the kitchen counter for no reason imo...
[20:28:50] <Lambda_Aurigae> Casper, toaster and microwave on same circuit..
[20:28:51] <Valen> yeah, its just culture shock really
[20:29:03] <Valen> our toasters are 2000W
[20:29:07] <Valen> kettles are the same
[20:29:21] <Casper> most toasters here are... fucking 700W .....
[20:29:22] <Valen> microwaves are ~1500W from the wall
[20:29:27] <Lambda_Aurigae> Valen, I have a 110V 30A outlet in my workshed...and a 240V 50A for welder.
[20:29:31] <Casper> it's not a toaster! it's a bread dryer
[20:29:52] <Valen> 900W microwave is common
[20:29:56] <Lambda_Aurigae> the 30A outlet is mostly to plug in an RV.
[20:30:13] <Valen> probably they are better now so I'm guessing 110v microwaves are 900W now too?
[20:30:45] <Valen> damn, kettles already take so long to boil, and your electircal heaters must suck lol
[20:30:56] <Lambda_Aurigae> 1100W for my under the counter mount microwave.
[20:31:12] <Valen> most new places are going to 3 phase feeds as well, for things like AC and the like
[20:31:17] <Lambda_Aurigae> under the cabinet, rather.
[20:31:21] <Valen> my house has 2 phases which I find odd and irritating lol
[20:31:29] <Lambda_Aurigae> so does mine.
[20:31:38] <Lambda_Aurigae> my office has 3.
[20:32:00] <Casper> basically, the other day I woke up with an idea (and still stuck on it)... in lake, you sometime see methane bubbles... what if it was huge! like a crack going down to a nat gas reserve... enought to give all the energy required for an house... heating, cooking and refrigerator is now a non-issue... but everything else...
[20:32:23] <Lambda_Aurigae> Casper, more likely from decomposing fish shit in the bottom of the lake.
[20:32:29] <Casper> yup
[20:33:03] <Casper> but what if... it have to have existed many years ago, which probably lead to the discovery of many petrol reserve...
[20:33:09] <Lambda_Aurigae> local hog slaughter house recovers methane from their waste product...the methane reclamation facility is twice the size of the slaughter house.
[20:33:41] <Casper> so... I was like... gas to electricity...
[20:34:02] <Casper> IC engine wouln't last an year and require lots of maintenance...
[20:34:04] <Lambda_Aurigae> I know several people who have gas wells on their property.
[20:34:20] <Lambda_Aurigae> they don't run generators on it though.
[20:34:33] <Casper> gas turbine is too expensive and the maintenance is "undoable", and researches say: 4 years then total rebuild...
[20:34:54] <Lambda_Aurigae> the wells are test drills to see if there is enough gas to be commercially viable..and they weren't...but still give off enough to run heaters and such.
[20:35:12] <Lambda_Aurigae> there are now methane fuel cells.
[20:35:12] <Casper> fuel cell don't run directly on nat gas, require a catalyst to convert to hydrogen... so a consumable...
[20:35:30] <Lambda_Aurigae> catalyst isn't consumed,,by definition.
[20:35:40] <Lambda_Aurigae> so if it's consumed it is not a catalyst but a reactant.
[20:35:45] <Valen> +1
[20:36:04] <Casper> well, the definition don't match the reality, as it wear out slowly, and use expensive metal...
[20:36:07] <Valen> though many times a physical catalyst may become obstructed, coated in crap
[20:36:09] * Lambda_Aurigae took 2 years of chemistry in highschool!
[20:36:24] <Valen> or as Casper says, worn away
[20:36:48] <Casper> so... I think the only viable 24/7 50 years is... steam...
[20:36:50] <Lambda_Aurigae> usually coated more than worn.
[20:36:58] <Casper> yeah..
[20:37:06] <Lambda_Aurigae> steam turbine!
[20:37:18] <Lambda_Aurigae> or sterling engine.
[20:37:34] <Casper> small steam turbine is inneficiant from what I found...
[20:37:44] <Casper> sterling engine, 16ft long for 1000W
[20:37:50] <Lambda_Aurigae> so?
[20:37:57] <Lambda_Aurigae> I got room for that in my back yard!
[20:38:43] <Casper> and considering that it run "dry" metal on metal... I'm worried about the wear...
[20:39:02] <Lambda_Aurigae> ok, then, go with thermocouple.
[20:39:11] <Lambda_Aurigae> use a deep water well as a heatsink.
[20:39:58] <Casper> thermocouple... couln't find much info on it, so musn't be viable as highish power
[20:40:04] <Casper> I found a reversed peltier
[20:40:26] <Lambda_Aurigae> peltier is a semiconductor thermocouple.
[20:40:37] <Lambda_Aurigae> more or less, kindasorta.
[20:40:56] <Lambda_Aurigae> either way you have to have a hot side and a cold side.
[20:41:08] <Lambda_Aurigae> a nearby lake for waste heat would be good.
[20:44:00] <Casper> I wonder what would be the efficiency of that...
[20:44:10] <Valen> steam still has moving parts, it'll still wear out
[20:44:11] <Lambda_Aurigae> big spring return rope...like, a barrel with rope wrapped around it with a coil spring inside...pull the rope out, spring gets wound up...release rope, spring rewinds rope....connect generator to said spinning barrel....
[20:44:17] <Lambda_Aurigae> now a way to pull the rope....
[20:44:22] <Lambda_Aurigae> hot air balloon!
[20:44:28] <Casper> Valen: yes, 2 bearings
[20:44:30] <Valen> thermocouple is stupidly low power and expensive
[20:44:33] <Valen> and all the seals
[20:44:39] <Valen> all the sliding contacts
[20:45:04] <Lambda_Aurigae> hot air balloon to pull rope up and wind spring...let air out when it gets high and zip back down to run generator!
[20:45:16] <Lambda_Aurigae> there...steampunk style generator system.
[20:45:33] <Casper> depend what kind of turbine, "true" turbine only have 2 bearings and a gasket around the 2 half. which might not be that important
[20:45:53] <Lambda_Aurigae> 10 or 20 of those hooked up and you should be able to charge your laptop batteries rather well.
[20:46:43] <Casper> TEG... 5-8% efficient... yuck
[20:47:06] <Lambda_Aurigae> solar panels with bright burning lanterns?
[20:47:19] <Casper> hmmm
[20:47:33] <Casper> actually.... that might not be such a bad idea
[20:47:34] <Lambda_Aurigae> could add TEG to that for extra extraction.
[20:47:57] <Lambda_Aurigae> nice tungsten mantles like a coleman lantern uses.
[20:47:57] <Casper> in the situation that I was thinking, all the unused gas need to be burned
[20:48:19] <Lambda_Aurigae> triple power extraction from lantern.
[20:48:22] <Casper> GENIUS!
[20:48:26] <Lambda_Aurigae> solar power from the light
[20:48:30] <Lambda_Aurigae> teg from the heat.
[20:48:48] <Casper> but really, I don't think teg would be usefull really
[20:48:49] <Lambda_Aurigae> and as the air rises, a set of turbine blades to catch it....doesn't need to be high speed turbine.
[20:48:57] <Lambda_Aurigae> every erg of energy helps.
[20:49:13] <Casper> hmmm
[20:49:16] <Lambda_Aurigae> you could also run a coil of copper tubing through the area and feed water through it.
[20:49:49] <Lambda_Aurigae> bury a 5000 liter insulated to hell water container and just keep water flowing through it...big thermal mass.
[20:50:04] <Lambda_Aurigae> to heat the house or provide hot water for showers and such.
[20:50:26] <Lambda_Aurigae> there's always ways.
[20:50:34] <Lambda_Aurigae> just how much money do you have to spend on it?
[20:51:00] <Casper> little :D
[20:51:21] <Lambda_Aurigae> that's always the problem.
[20:51:24] <Casper> but really, your tungsten thing seems like a bright idea
[20:51:38] <Lambda_Aurigae> I could go pretty much off-grid here with one or two 30K USD wind turbines.
[20:51:56] <Lambda_Aurigae> coleman lantern mantels have tungsten in them to burn really bright.
[20:52:05] <Lambda_Aurigae> they are also ever so slightly radioactive.
[20:52:16] <Tom_itx> yup
[20:52:19] <Tom_itx> they make them here
[20:52:26] <Lambda_Aurigae> I remember checking them in high school with our school geiger counter.
[20:52:39] <Lambda_Aurigae> was neat to be able to set it off with a little bit of loosely woven cloth.
[20:53:35] <Casper> I wonder how expensive that material is
[20:53:51] <Tom_itx> for what?
[20:53:54] <Casper> and if it make good light source for solar..
[20:54:04] <Lambda_Aurigae> oh, it's thorium dioxide.
[20:54:34] <Lambda_Aurigae> I thought they used tungsten but I was mistaken.
[20:54:43] <Lambda_Aurigae> they use thorium...that's why the radioactivity.
[20:55:06] <Lambda_Aurigae> Casper, it's very bright! just buy mantles and use standard burning system with them.
[20:55:15] <Tom_itx> yeah
[20:55:22] <Tom_itx> they're darn bright
[20:55:35] <Tom_itx> i suppose they'd work with any gas wouldn't they?
[20:55:40] <Lambda_Aurigae> long as you have constant gas flow and it's burning the dang things last halfway to forever...till you touch them.
[20:55:50] <Lambda_Aurigae> Tom_itx, should work with methane/natural gas...
[20:55:54] <Lambda_Aurigae> they work with propane.
[20:56:08] <Tom_itx> yeah they take fair abuse too until you touch em
[20:56:31] <Tom_itx> i used to get em off the line here
[20:56:59] <Tom_itx> they tore the old plant down now though, i think it's still here but moved
[20:57:08] <Tom_itx> groundwater contamination
[20:57:12] <Tom_itx> they had to
[20:58:03] <Tom_itx> recent years they switched most their stuff to propane
[20:58:07] <Casper> is there anything simmilar to that but ain't touch sensitive?
[20:58:13] <Tom_itx> probably safer than the white gas they used
[20:58:31] <Lambda_Aurigae> every so slightly.
[20:58:48] <Lambda_Aurigae> Casper, well, you could do a limelight...but I think the lime gets used up.
[20:58:54] <Tom_itx> long time ago i worked there for a spell
[20:59:04] <Lambda_Aurigae> Tom_itx, so THAT's what's wrong with you.
[20:59:10] <Tom_itx> in the machine dept
[20:59:39] <Tom_itx> ran a row of 6 & 8 spindle lathes
[21:00:29] <Tom_itx> chances are good if you own a coleman product i made part of it
[21:01:09] <Casper> Tom_itx: is it the part that broke? or the part that will never break? :D
[21:01:21] <Tom_itx> i made good parts
[21:01:32] <Tom_itx> all the 'larger' machined parts
[21:01:43] <Tom_itx> there were 3 machine rooms for various size parts
[21:02:02] <Tom_itx> probably 100 or more machines all together
[21:02:23] <Tom_itx> i'm sure they've scaled way down nowdays
[21:02:57] <Lambda_Aurigae> outsourced a lot of it too.
[21:03:21] <Tom_itx> i know their plant is alot smaller than it was
[21:05:58] <Tom_itx> actually they had 3 in town
[21:06:30] <Tom_itx> the north plant was heating/ac, northeast was mostly stamped parts and plastics like coolers and canoes etc
[21:06:43] <Tom_itx> downtown was the lantern lines and machine shops
[21:08:21] <Lambda_Aurigae> kewl.
[21:08:29] <Lambda_Aurigae> ok..off to bedzies for bonzoies here.
[21:08:35] <Tom_itx> me too
[23:49:29] <ferdna> what type of laser does this thing uses:
[23:49:30] <ferdna> http://www.opticsplanet.com/leupold-rx-1200i-tbr-compact-digital-laser-rangefinder-with-dna-119360.html