Back
[09:07:50] <eszett> hi
[09:08:25] <Chillum> hi
[09:08:32] <bss36504> hi
[09:08:57] <eszett> I have a question. The 10µF decoupling capacitor, that is necessary on the VBUS line (of my Atmega32U4). Does this cap have to be close to the usb connector, or can it be placed anywhere (as long as it is not too far from the MCU, 2 inches is max. distance AFAIK) ??
[09:09:44] <LeoNerd> I tend to put a bigger 10µF "somewhere" on the board; more important is the placement of the smaller (10-100nF) caps
[09:09:57] <eszett> all the keyboard pcbs I know have it placed close to the USB connector, and im asking if this is for a reason?
[09:10:15] <eszett> Leonerd: exactly, thats what i have already learned so far =)
[09:10:22] <bss36504> Generally it's good to place decoupling caps as close to the load as possible. If they are too far away, the trace inductance negates the decoupling abilities. 10uf is considered "Bulk" most of the time, so you can probably put it near-ish to the mcu or USB plug. Smaller caps should be near the micro
[09:11:46] <eszett> ye, the question i was discussing with neophyte was, if there is a forcing reason, why all the pcbs we know have this 10µF decoupl. cap for VBUS being placed directly close to the USB connector
[09:12:23] <LeoNerd> No real technical reason other than the general principle of keeping main power traces short
[09:12:34] <eszett> And the argument was, that it may be close to the USB connector to help with the unrush effect:
http://www.testusb.com/inrush_issue.htm
[09:12:39] <LeoNerd> So since the power probably comes in the USB connector, it makes sense putting the cap close to it
[09:12:58] <bss36504> I would imagine it's akin to having an input cap on a regulator. Might as well smooth the power out of any external disturbances before trying to do much else with it.
[09:13:10] <eszett> Leonerd: ah, thx that answers the question!
[09:14:06] <eszett> yes the power comes from USB connector so its just comfortable to put it there, since the region close to the MCU is crammed with traces and other stuff, and near to the USB connector there is plenty of space for placing parts
[09:14:44] <eszett> So it has nothing to to with the inrush effect?
[09:14:52] <LeoNerd> That's probably helped too
[09:14:58] <eszett> ok
[17:17:59] <CasperPhone> Boo!
[17:31:20] <cehteh> soo .. i've build a heating blanket for my next epoxy mold,
https://pbs.twimg.com/media/CoTKze7XYAUhMok.jpg
[17:32:08] <cehteh> will use an ardino nano to control the temperature w/ PID controller and PWM
[17:33:01] <cehteh> now the interesting question: would it make sense to put it in line with a fat inductor and diode to even out the current?
[17:34:03] <cehteh> duno if these cheap 13A 12v power supplies like 15A pwm switched
[17:34:08] <cehteh> 30A
[17:58:04] <cehteh> eveyone asleep :D
[17:59:20] <LeoNerd> I'm not
[17:59:54] <LeoNerd> I personally wouldn't try PWM'ing the bulk of the output of a switching PSU
[18:00:17] <LeoNerd> In your situation, given you need a lot of current and the voltage likely isn't critical, I'd just get a transformer/bridgerec/smoothing cap. Likely doesn't really need regulation even
[18:00:29] <LeoNerd> The heating controller has feedback I presume?
[18:01:45] <cehteh> yes
[18:02:17] <cehteh> well i have this switching psu's around, they are more efficient and cheaper than a 20A transformer
[18:02:49] <LeoNerd> Mm...
[18:02:56] <LeoNerd> Just put a real big cap in front then
[18:03:17] <cehteh> but i see that pwm high current will suck, thats why i ask if its reasonable to even that out
[18:03:43] <cehteh> cap wont work, hey how much does that need, lets say 200hz pwm
[18:04:02] <LeoNerd> Cap between the switching PSU and your PWM controller
[18:04:09] <cehteh> ok maybe i could do much higher frequency pwm
[18:04:11] <LeoNerd> So as to even out the load you draw from the PSU
[18:04:29] <cehteh> then parasitic inductance and capacitance would even the shit out
[18:04:52] <cehteh> http://tinyurl.com/jkw593r
[18:04:59] <cehteh> thats what i am just playing with
[18:05:53] <cehteh> the 0.01Ohms simulate the inpedance of the PSU, 0.8Ohms is my load
[18:09:08] <cehteh> but looking at it, the bigger the inductance the lower the total power it supplies
[18:09:22] <cehteh> posisbly wrong way of thinkin
[19:32:55] <LeoNerd> I wonder... presumably, a standard ATmega328P "arduino" bootloader is going to work on a 328PB, yes? The PB has all the same registers in the same place, so it ought to suffice as a bootloader..?
[19:47:13] <Tom_itx> should
[19:47:33] <Tom_itx> are the bits all in the same regs or any moved to unused regs?
[19:47:47] <Tom_itx> to make room for the new ones
[19:48:09] <Tom_itx> double check that then go for it
[19:51:43] <LeoNerd> I believe so.. I should doublecheck I guess. I'll read about it tomorrow
[19:51:51] <LeoNerd> at worst I just have to grab their source and rebuild it myself
[19:51:58] <Tom_itx> they're probably the same
[19:52:06] <Tom_itx> not that difficult
[19:52:17] <Tom_itx> i have it here if you want it
[19:52:24] <Tom_itx> app109?
[19:52:25] <Tom_itx> iirc
[19:52:27] <LeoNerd> I'm sure I can find it somewhere
[19:52:30] <LeoNerd> Yah that one
[19:52:43] <Tom_itx> i've got the cd with all the pdf and app notes etc
[19:52:50] <LeoNerd> I'll get my board up on Tindie tomorrow :)
[19:52:52] <Tom_itx> no longer distributed
[19:53:01] <LeoNerd> Got the first two copies ready to sell
[19:54:49] <Tom_itx> nice
[20:53:14] <cehteh> hah LeoNerd solution: LC filter
