Back
[00:16:08] <Casper> N1njaneer: have you noticed that the enter key don't work when you get a phone call?
[00:16:24] <N1njaneer> Yes, annoying like that. Need to check the wiring :D
[00:17:33] <Casper> as long as you don't try to check the wiring in a car :D
[00:17:45] <Casper> I was worried to have to do it this week end...
[00:18:20] <N1njaneer> Hey, at least it isn't as scary as having to check the 3-phase stuff here. XD
[00:18:23] <Casper> my ABS decided to kick in when braking normally, increasing the braking distance by like 20ft before disabling itself
[00:18:36] <N1njaneer> Oof, have had that happen before.
[00:19:11] <Casper> got 3 codes: FR signal is 0, FR excessive signal variation, FR circuit is open or short
[00:19:31] <Casper> no signal and too much.... short or open... very precise
[00:19:53] <N1njaneer> LOL you'd think the latter signal could at least tell you which, but I'm sure they saved $0.05 by not being able to differentiate
[00:19:57] <Casper> ended up to be a bad sensor (read 586kohms, supposed to be 800-1200 ohms)
[00:20:11] <Casper> and the fun part...
[00:20:17] <Casper> do you know where the sensor is?
[00:21:11] <Casper> in the f* wheel bearing
[00:23:52] <N1njaneer> Yeah, gotta love that.
[00:26:28] <Casper> so, needed to get a 36mm socket, remove the half shaft nut at the wheel hub, try to hammer it out to see if it move... it did! put back the nut, went buy the new hub. remove nut, partially hammer the shaft out, then jack, remove wheel, remove caliper and hang it, remove brake pad, remove brake bracket, remove disk, remove the 3 bolts that hold the hub, hammer out the hub, sensor broke in the hole... hammer it, nope, remove suspension, now can remove t
[00:26:28] <Casper> he shaft from the hub completly and work from behind, hammer out the sensor, steel brush clean the hole of corrosion...
[00:26:32] <Casper> then reassemble
[00:27:26] <Casper> meanwhile, since the trouble was intermittant, I pulled out the 50A abs fuse
[00:27:39] <Casper> so once I put it back, I cleared the codes...
[00:27:44] <Casper> 2 for abs, 1 for airbag
[00:27:57] <Casper> and yes, pull out the abs fuse, get airbag code :D
[00:28:40] <N1njaneer> Make sure you deploy the airbag to make sure it's working :D
[00:28:46] <Casper> the code was: U1040: lost communication with (brake module)
[00:29:30] <Casper> my guess is that when the airbag deploy, it use the ABS system to actually apply the brakes and the brake module is also what control the cruise control, so it disable the cruise at the same time
[00:31:41] <Casper> but man... garages around here are theif
[00:31:54] <Casper> 79.95$ to read the code...
[00:32:00] <Casper> got a code scanner...
[00:33:30] <N1njaneer> Casper: AutoZone in the US here is nice - free code reading.
[00:33:43] <Casper> knowing that I would have needed it twice, once to get the code, once to clear the codes...
[00:33:44] <N1njaneer> OBDC code readers are quite cheap online
[00:33:47] <Casper> here they all charge
[00:33:58] <Casper> readers that do ABS ain't cheap
[00:34:02] <N1njaneer> new AVR project for you!
[00:34:15] <Casper> oh the communication is simple
[00:34:23] <Casper> the problem is the protocol
[00:34:27] <Casper> undocumented
[00:34:41] <Casper> and vary from car model to another one
[00:34:44] <N1njaneer> I'm sure The Internet has reverse-engineered it. I'd look around.
[00:35:02] <Casper> for ECU yes
[00:35:08] <Casper> for ABS... nope
[04:15:08] <Jartza> does anybody know of dc boost converter, that would work with one aaa-battery and have dual-output, 3.3V and 9V? :)
[04:16:54] <Thrashbarg> with an AAA battery you'd want a flyback converter, not a boost converter... it'd be too inefficient
[04:17:06] <Thrashbarg> winding transformers is a delicate art however
[04:18:57] <Jartza> doesn't need to be that efficient
[04:19:02] <Jartza> but it needs to be small
[04:19:13] <Thrashbarg> how long do you hope to run off an AAA battery for? Minutes? :P
[04:19:26] <Jartza> 2-3 hours
[04:20:01] <Jartza> maximum load would be around 5mA on 3.3V and 10mA on 9V
[04:21:50] <Jartza> so if calculated with 100% efficiency, that would be around 71mA on 1.5V
[04:22:04] <Thrashbarg> except a boost converter isn't 100% efficient
[04:22:04] <Jartza> one aaa has capacity of 800-1200mAh
[04:22:13] <Jartza> well of course it isn't
[04:22:27] <Jartza> but I guess the efficiency can be 50% and I still achieve what I want
[04:23:16] <Jartza> 2 hours is fine, 3 hours would be awesome
[04:24:08] <Jartza> and I could do that with 2 dc-dc-boost converters, but I was wondering if anybody knows of a boost converter that has dual output in that range (with very small input voltage)
[04:25:51] <Thrashbarg> does the 9V line need to be regulated?
[04:26:25] <Jartza> not that accurately
[04:26:32] <Jartza> I guess anything between 8 to 15 is ok
[04:26:42] <Jartza> or not guess, that says so in datasheet ;)
[04:27:40] <Jartza> and the 3.3V will drive attiny and one i2c device
[04:27:56] <Thrashbarg> you might be able to get away with having a tapped inductor, so all of it is used to generate 9V and some of it is used to generate the 3.3V supply
[04:29:09] <Jartza> hmm
[04:29:10] <jacekowski> there are some avrs that have picopower capability
[04:29:19] <jacekowski> and have 3.3V output for supplying devices
[04:29:31] <jacekowski> so you get 1.5V battery that will get discharged as low as 0.7V
[04:29:46] <Thrashbarg> they use a charge pump I guess?
[04:29:59] <jacekowski> i'm not sure
[04:30:12] <Thrashbarg> if there's no external inductor they would
[04:30:15] <jacekowski> http://www.atmel.com/technologies/lowpower/default.aspx
[04:30:23] <jacekowski> you can have builtin inductors
[04:30:56] <jacekowski> maxim makes fully integrated RS485 ICs that have builtin SMPS for isolated side
[04:31:02] <Thrashbarg> yea
[04:31:46] <Jartza> sure I could use those, but this is basically few existing devices which I want to put together :)
[04:31:47] <jacekowski> http://www.atmel.com/Images/doc8206.pdf
[04:31:51] <jacekowski> yeah, external inductor
[04:31:53] <Jartza> and this has attiny85
[04:48:23] <Jartza> nah, maybe I'll just do it with 2 chips
[08:59:09] <Broot> morning
[09:18:16] <Jartza> well, good day :)
[09:19:12] <Jartza> seems like I had ncp1402 (3.3V version) and mc34063a chips on shelf
[09:19:45] <Jartza> made a quick'n'dirty AAA-battery -> ncp1402 -> mc34063a circuit and now I get both 3.3V and 9V from aaa battery :)
[09:24:04] <Jartza> seems to work ok
[09:24:12] <Jartza> and aaa can handle the load
[12:53:17] <Jartza> oh. seems I tackled the chinese datasheet
[12:53:21] <Jartza> everything seems to work
[12:53:51] <Jartza> although, I can also speak finnish, so no wonder? :D
[12:54:01] <Jartza> which reminds me of...
http://pics.kuvaton.com/kuvei/finnish.jpg
[12:58:57] <twnqx> http://oi61.tinypic.com/2ikfxb7.jpg first test setup: success
[12:59:22] <twnqx> now to wait for delivery of the other test parts.
[13:13:06] <twnqx> hm. is there no way to make an atmel an I2C slave with multiple addresses using the hardware receiver?
[13:14:57] <Jartza> no idea, I just made bitbang i2c
[13:16:19] <Jartza> I would've used USI, but I needed 2 i2c-devices which use same, non-configurable slave-address :(
[13:16:49] <N1njaneer> Jartza: Yeah, same here. :)
[13:17:04] <Jartza> I made a cruel hack as I don't have to speak to them at the same time :)
[13:17:08] <Jartza> they share the SCL
[13:17:18] <N1njaneer> As long as it works!
[13:17:20] <Jartza> and the code, of course
[13:17:39] <twnqx> but how do you get 400khz software :S
[13:17:43] <Jartza> I just change one flag which defines which pin to use as SDA
[13:18:05] <Jartza> luckily I don't need 400kHz
[13:18:18] <twnqx> well, i am a slave to whatever is thrown at me
[13:18:26] <twnqx> (i have no clue what speed to expect, really)
[13:19:00] <Jartza> I haven't checked my implementation speed
[13:19:03] <N1njaneer> Things only need to run as fast as you need them to :)
[13:19:13] <twnqx> i have no control of the master
[13:19:36] <Jartza> hmmh
[13:19:40] <Jartza> sure you have?
[13:19:47] <Jartza> it's called clock stretching :P
[13:20:05] <twnqx> from what i read that's only at the end of a byte?
[13:20:46] <Jartza> nope
[13:20:50] <Jartza> for any bit you need
[13:24:36] <Jartza> well, I have some fluctuation in my i2c-speed, but saleae says it's average 400kHz
[13:24:39] <Jartza> bitbanging
[13:26:34] <twnqx> hm ok
[13:32:50] * Tom_itx stretches his clock like a rubber band
[13:40:06] <RikusW> it won't add any more hours to your day ;P
[13:41:23] * N1njaneer starts a movement for adding leap-picoseconds, so leap-seconds aren't so much of a temporal bump in the road anymore.
[14:40:16] <rewolff1> USI doesn't have hardware slave addresses, so you can use hardware-assisted USI, right? Even for multiple slave addresses.
[14:40:35] <rewolff1> *USI -> I2C
[14:42:14] <rewolff1> Be careful with clock stretching.... Some implementations only expect that at the end of the byte.... (but as jartza says, should work inside a byte. I'm trying to avoid that if I can... ).
[14:42:50] <rewolff1> And if the slave might be a BCM2835.... it has a hardware bug in clock stretching....
[14:43:00] <rewolff1> *slave -> master
[14:45:05] <rewolff1> (USI on 'tiny you'd be able to use two hardware addresses.... with the I2C (called TWI iirc) on the 'megas you have the address recognition stuff -> you'd be limited to one address. )
[14:51:20] <twnqx> grml
[15:15:50] * RikusW don't particularly like USI....
[16:06:57] <Jartza> I don't like USI either
[16:07:19] <Jartza> so far I've done all the spi and i2c with bitbanging
[22:51:23] <rue_mohr2> I think the 34063 can go from 0-100% duty...
[22:52:54] <rue_mohr2> if I dial a char charger down to 3V I'll know
[22:53:04] <rue_mohr2> a car usb adapter
[22:55:19] <rue_mohr2> no it cant
[22:55:20] <rue_mohr2> drat
[22:56:39] <rue_mohr2> whats that current based one?
[22:56:45] <rue_mohr2> the 8 pin
[22:56:49] <rue_mohr2> 34.. no...
[22:56:54] <rue_mohr2> 3842
[22:59:30] <rue_mohr2> oo yea it can
[23:19:54] <rue_mohr2> wrong channel
