#avr | Logs for 2014-11-17

Back
[09:29:11] <rue_more> malinus, I have a similar project, but I'm using an exdternal 24c515 for storage
[09:29:16] <rue_more> 8 bit, 8khz
[10:55:51] <Duality> looking at the bit/baud rate tables in the datasheet, and some have negative error rates
[10:55:55] <Duality> what does that meen ?
[10:56:01] <hypermagic> hello my friends
[10:57:22] <LeoNerd> Surely it just means the error is on the wrong side - i.e. too fast vs.. too slow
[10:57:48] <Duality> oh never thought about it that way
[11:00:52] <Duality> doesn't say anything about it in the datasheet though
[11:05:27] <twnqx> i think
[11:05:38] <twnqx> if you calculate the actual rates with the given formula
[11:05:46] <twnqx> you will come to exactly that result.
[11:11:51] <hypermagic> atmel wants to connect your car to a wireless grid btw
[11:12:07] <hypermagic> and they will do it if you want or not
[11:14:15] <asteve> those jerks
[11:22:15] <hypermagic> it will be quite easy to convince the civilians to buy this awesome self driving car that will take you to work every day.
[11:22:55] <hypermagic> only the masses matter, then you will become minority
[11:47:14] <Casper> Duality: technically, error rate can not be negative... it's how much off it is compared to the real value. However it does help to know in which direction is off, so you may be able to finetune the oscillator, or figure out if the other chip, that is also off, will work correctly
[11:48:16] <N1njaneer> Casper: No, the error rate being negative means the data is better than 100% perfect :D
[11:49:04] <Casper> like you must not go over 2% error.... what if you have 2 chip that run at 1.5% error and the other at 1.4%? will it run? if one is too slow and the other too fast, it won't... but if both are too slow or too fast it work. +1.5 and +1.4 is really only 0.1% off
[11:50:07] <N1njaneer> Part of that is also compounding error rate and drift, which sometimes has to do with how long the runs of bytes are between a paused return back to the marking state.
[11:51:23] <aandrew> a negative error rate means it's error correcting. :-)
[11:55:36] <Duality> that would be nice
[14:16:32] <hypermagic> btw guys, if somebody is interested what is not being mass produced : http://en.wikipedia.org/wiki/AMULET_microprocessor this is an asynchronous cpu that consumes microwatts in idle.
[14:23:55] <Casper> hypermagic: avr can go sub 1µA, and can go down under 2V... not surprised
[14:24:16] <N1njaneer> Yeah :)
[14:25:11] <twnqx> but not active
[15:08:15] <N1njaneer> Tech support in the middle ages: https://www.youtube.com/watch?v=aX0-nqRmtos
[15:11:33] <Casper> hehe
[15:52:08] <N1njaneer> Hooray ordering a connector that one one else in the world regularly uses.
[15:52:36] <LeoNerd> "one one" ?
[15:52:50] <aandrew> that's one way to ensure availability
[15:54:11] <malinus> who's one one?
[15:54:33] <LeoNerd> malinus: One one was a racehorse.
[15:54:52] <LeoNerd> .oO( Two Two was one too / One one won one race / Two two won one too )
[15:55:19] <malinus> hehe
[15:55:47] <N1njaneer> Sorry, "no one"
[15:55:54] <N1njaneer> Hooray ordering a connector that no one else in the world regularly uses.
[15:56:08] <LeoNerd> I guessed ;)
[15:56:13] <N1njaneer> Which means they're all machined to order, hence the 12-13 week leadtime.
[15:57:35] <LeoNerd> Eek
[16:04:51] <malinus> you should be sad, really
[16:14:38] <N1njaneer> Mmm?
[16:16:19] <malinus> why would one be happy about using some obscure connector? Shouldn't we be happy about standards?
[16:16:31] <malinus> standars make the world spin
[16:16:34] <malinus> d
[16:26:31] <N1njaneer> When you need a medical-grade connector because application durability is paramount there's really no way around it.
[16:26:42] <N1njaneer> It's just not a stocked part.
[16:28:23] <N1njaneer> We regularly use connectors which are $50-$100/each because they are medical or milspec for durability, as the customers demand it :)
[16:28:49] <N1njaneer> Thus every one of them is usually machined to order. Downside is that it's usually 4-12 weeks leadtime.
[16:29:08] <Tom_itx> bright side is the metal is nice and shiny
[16:29:35] <Tom_itx> not all tarnished from sitting on a shelf for years
[16:29:47] <N1njaneer> http://www.lemousa.com/en/products
[16:30:18] <N1njaneer> Tom_itx: Well, all of these are gold-plated contacts so never a worry there :)
[16:30:56] <N1njaneer> And more expensive than Lemo is Fischer -- http://www.fischerconnectors.com
[16:30:59] <N1njaneer> But REALLY nice
[16:51:18] <Tom_itx> so it's all your fault my medical bills are so high?
[16:52:05] <LeoNerd> Ohohoh! My board is shipping! :)
[16:52:28] <Tom_itx> cool! another 4 weeks and you'll have it!
[16:53:10] <LeoNerd> Hah.. yeeeah
[16:53:43] <Tom_itx> better practice your smt soldering skilz
[16:53:52] <LeoNerd> Yeah.. been working on it
[16:53:58] <LeoNerd> Got myself a big magnifier lamp
[16:54:06] <Tom_itx> those are handy
[16:54:11] <Tom_itx> get a scope
[16:54:17] <Tom_itx> those are even more handy
[16:55:30] <Tom_itx> http://tom-itx.ddns01.com:81/~webpage/scope/scope1.jpg
[16:56:33] <LeoNerd> I don't think I'll be going -quite- small enough for that :)
[16:56:53] <Tom_itx> good for finding boo boos
[16:57:15] * LeoNerd read all about reballing the other day
[16:57:45] <Tom_itx> http://tom-itx.ddns01.com:81/~webpage/scope/DSC_0001.JPG
[16:57:47] <Tom_itx> large pic
[16:57:56] <Tom_itx> but you couldn't see that with the naked eye
[16:58:25] <Tom_itx> those 2 traces were fused together under the mask
[17:12:45] <N1njaneer> https://www.youtube.com/watch?v=97v3gI9hx0U