Created Thu, 29 Dec 2011 23:53:57 +0000 by raypm
Thu, 29 Dec 2011 23:53:57 +0000
I was working on my board earlier today, and the voltage regulator tended to get a bit warm when running off a 9v wall dc power supply. I didn't think this was a problem. I made sure to unplug both the USB cable and the 9v ext supply when leaving for lunch, and now when I plug either back in the two LEDs near the reset button (LD1 and LD2) as well as the LD3 near the DC jack flash once when I plug in the USB or DC, but then they turn off and nothing else happens. I have the Max32 model.
I cannot upload sketches either. As I get "avrdude: stk500_2_ReceiveMessage(): timeout"
What's going on? Thanks for your insight.
Fri, 30 Dec 2011 00:01:02 +0000
Which board do (did?) you have?
Fri, 30 Dec 2011 00:07:45 +0000
The max32 model. According to this:
There is a problem connecting with the board but I double checked the serial port.
Normally there's a bit of LED blinking wasn't there? Like, when you plug it in, but it's not blinking at all.
Fri, 30 Dec 2011 00:11:25 +0000
I don't have one for reference, but looking at the schematics, you could try measuring the voltage between ground and the tab of IC4.
That should read 5V.
If it does, then check the voltage between ground and pin 8 of IC3.
That should read 3.3V.
Alternatively, just test the voltages of the +5V and +3.3V pins on the headers - they should be connected direct to those above locations anyway.
Fri, 30 Dec 2011 00:15:11 +0000
Good idea! The 5v pin reads correctly at 4.99, but the 3.3v pin reads at 0.38, so I guess that means the 3.3v regulator isn't working? Any ideas? Also both regulators seem to get awfully hot, is that normal?
Fri, 30 Dec 2011 00:19:31 +0000
The data sheet for that voltage regulator is here: [url]http://ww1.microchip.com/downloads/en/DeviceDoc/22026b.pdf[/url]
According to that the regulator has both short circuit current limiting and over temperature protection.
It should be almost impossible to cook it.
My guess is that you may have either developed a short circuit somewhere, or something has come adrift. (maybe due to excess heat melting solder?)
Check pins 1, 2 and 3 of IC3 to make sure they all read 5V (it is fed direct from the 5V regulator).
Fri, 30 Dec 2011 00:20:47 +0000
As for the heat, they should only get very hot if you are drawing lots of current.
The 3.3V one can supplu 500mA, and the 5V one 1A.
If you have a short circuit then the current may ramp up quite drastically causing lots of heat.
Fri, 30 Dec 2011 00:30:54 +0000
Yikes, IC3 is an SOIC! Hard to measure the voltages but none of them read 5v for sure. I unplugged my board from my setup and am powering it directly from USB with nothing else connected. IC5 gets hot to the touch as does IC3.
Fri, 30 Dec 2011 00:36:04 +0000
Also, I had 8 100k pots for knobs which were from the 3.3v pin to ground, but that shouldn't come close to the 800mA limit. But even when I unplugged everything and tried to power the board by USB or by 9V by itself, it didn't work out.
Fri, 30 Dec 2011 00:40:57 +0000
IC5? But that's just a simple low-power op-amp. It's purpose is to detect the voltage coming in through the power jack to see if the board should be self or bus powered. All it does is switch a MOSFET (Q1).
Hmmm... My guess is that possibly a power spike might have cooked IC5 which is causing a knock-on effect to IC3. The input power to the jack gets divided in half by a 10+10KÎ© voltage divider. This is fed into the non-inverting input of the op-amp IC5 and compared to the output of the 3.3V regulator IC3.
The inputs are rated at max Vdd+1V and Vss-1V, so with a 5V supply to it (which I think it is - it's not clear on the schematics) that would equate to between -1V and +6V, or, before the voltage divider, -2V and +12V. Anything outside that range coming into the jack could damage it and thus fry it.
Fri, 30 Dec 2011 00:43:49 +0000
Yes, I have deciphered that bit of the schematics, and IC5 is being supplied by 5V.
Fri, 30 Dec 2011 00:49:03 +0000
I see... yeah alright. I did fail to mention that I was soldering parts of the enclosure earlier and touched the solder tip to the enclosure casing (which is connected to ground) and there was a spark. This was not expected because soldering other parts of the circuit's grounding was fine, except now that I think of it, it was not plugged in at the time. I guess I did brick it?
I wouldn't have thought the solder tip would cause a spark when connected to ground since it's just a set of coils and a heating element, but I guess that's what bricked my board. Oh well, live and learn, I still have the enclosure and all the wiring mostly done. Guess I'll order me another board? Thanks again majenko!
Fri, 30 Dec 2011 00:55:55 +0000
Your pots should be consuming about 300ÂµA of current ;)
I too wouldn't expect a soldering iron to cause any problems of this sort. That may be a red herring. Were you soldering with the unit powered up?
TBH I would have designed a couple of clamping diodes and absorbing capacitors into the VCMP line so it can never exceed Vdd + Vdf or Vss - Vdf on that input to the op-amp. I haven't found many other cut corners on these boards (not like the Arduino boards - they cut so many corners on those ones I'm amazed they work at all. I mean, who's idea was it to use the FT232 chip as a 3.3v voltage regulator?!?!) but this seems to be the only one that could potentially cause issues.
If you're handy with that soldering iron of yours you could try replacing IC5 - but it'd take a very steady hand, and I can't be certain that that's really the issue (though it does sound like it).
Fri, 30 Dec 2011 01:08:42 +0000
Ack, not that handy! Can program, can design analog audio circuits, don't have small fingers, nerves of steel, or really even good lighting in here...
Anyways, yeah note to self: when soldering unplug everything and handle it separately. But in the mean time I'll finish up the enclosure stuff sans microcontroller. Glad to be doing this stuff again, and I remember why I switched to software! Thanks man.
Fri, 30 Dec 2011 18:36:13 +0000
I suggest you take a DMM and check out that soldering iron! The tip is typically grounded on a proper 3- wire powered iron. Is the iron (electronically) temp controlled?
Also wouldn't hurt to check your bench/shop wiring. Just saying...
Fri, 06 Jan 2012 01:08:43 +0000
TBH I would have designed a couple of clamping diodes and absorbing capacitors into the VCMP line so it can never exceed Vdd + Vdf or Vss - Vdf on that input to the op-amp. I haven't found many other cut corners on these boards (not like the Arduino boards - they cut so many corners on those ones I'm amazed they work at all.
You're right, there should be clamp diodes on that signal. That was an oversight. It will get corrected in the next rev of the board. Thanks for pointing it out.
Gene Apperson Digilent