Occasionally Hot UC3A3256

Log in or register to post comments
Go To Last Post
6 posts / 0 new
Author
Message
iwoloschin
iwoloschin's picture
Level: Hangaround
Joined: Wed. May 11, 2011
Posts: 337 View posts
Location: Boston, MA
#1
Posted by iwoloschin: Fri. Dec 2, 2011 - 04:55 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've got a custom UC3A3256 PCB and I've got an intermittent, odd problem. Every now and then, I plug the PCB into the USB port and nothing happens. Trying to reset into the bootloader doesn't seem to work either. If I leave it for a minute or two, the MCU actually gets too hot to touch (more I notice the smell first, but no smoke, thankfully!). At this point, I'll pull the PCB out of the USB port and let it cool on my desk, it's got a big exposed ground plane, so it cools relatively quickly. After it's cool, I tried plugging it back in, with a USB sniffer in between, and it came up working perfectly fine, though it loaded the bootloader instead of my application.

The only thing I did before plugging it back in was check for shorts on the power rails, but I didn't find any.

I've had this happen a couple of times now, the first I thought I just had a fluke, a bit of solder or dust stuck somewhere, but now I'm not so sure. I can't think of any hardware reasons this would happen, but software seems just as unlikely. I'm not sure I can do much right now other than monitor this and see if it happens again, unless anyone else has any ideas?

Tags:

Tools, Evaluation and Development Kits (AVR-related)
Top
gchapman
gchapman's picture
Level: 10k+ Postman
Joined: Tue. Jan 9, 2007
Posts: 16823 View posts
Location: Clay County, Texas, u.S.A.
Posted by gchapman: Fri. Dec 2, 2011 - 08:09 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Latchup?
The USB data lines maybe feeding the MCU thru the ESD diodes before the MCU's power is up.
Go through the Atmel schematic checklist.
A USB self-powered hub may be better than the laptop.
There's a somewhat recent post on an AVR32 sub-forum here about a bug that, IIRC, related to the power-up sequence (a race between core and I/O power).

"Dare to be naïve." - Buckminster Fuller

Top
leon_heller
leon_heller's picture
Level: Raving Lunatic
Joined: Fri. Jul 27, 2001
Posts: 7672 View posts
Location: St. Leonards-on-Sea (UK)
Posted by leon_heller: Fri. Dec 2, 2011 - 08:50 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

CMOS latchup would be my guess, as well. I've only had it happen with a couple of devices: the Inmos transputer and a Scenix chip.

Leon Heller G1HSM

Last Edited: Sat. Dec 3, 2011 - 04:11 AM
Top
gchapman
gchapman's picture
Level: 10k+ Postman
Joined: Tue. Jan 9, 2007
Posts: 16823 View posts
Location: Clay County, Texas, u.S.A.
Posted by gchapman: Sat. Dec 3, 2011 - 02:46 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

too hot to touch - the USB host should not supply that much power until the power handshakes are done between the device and host.
power sequencing - though a USB connector is a sequenced break/make (power on before data), if the MCU's voltage regulators (feeding it and within it) are too slow (loaded too much, capacitors are too large) then that's a concern.

"Dare to be naïve." - Buckminster Fuller

Top
iwoloschin
iwoloschin's picture
Level: Hangaround
Joined: Wed. May 11, 2011
Posts: 337 View posts
Location: Boston, MA
Posted by iwoloschin: Mon. Dec 5, 2011 - 04:05 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Does anyone have any good references on latchup? I've used the schematic checklist, and I'm pretty sure everything is correct there, but it doesn't necessarily account for other devices on the circuit (voltage regulator to generate 3.3V, peripherals, etc).

A race condition makes sense, since it doesn't happen often, guess I've got to go do some more research.

Thanks guys.

Top
gchapman
gchapman's picture
Level: 10k+ Postman
Joined: Tue. Jan 9, 2007
Posts: 16823 View posts
Location: Clay County, Texas, u.S.A.
Posted by gchapman: Mon. Dec 5, 2011 - 09:51 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

No handy reference for latchup.
TVSs on your signals (USB too) and power?
That should shunt impulses thru the TVS diode instead of an ESD diode.
Some voltage regulator datasheets will state that connecting or disconnecting the power cable can over-voltage the voltage regulator's input; solution is avoid ceramic capacitors or use a series resistor (on-the-cap or on power in). A power impulse can blow through the voltage regulator into the MCU.
The more reliable way to limit ESD diode current is to add resistance (if possible).
Edit: a reference -
What is latch-up and how can I prevent it? by Analog Devices.
Edit: a UC3L datasheet stated that latch-up is possible if the power supply rise rate (volts/microsecond) is too large.

"Dare to be naïve." - Buckminster Fuller

Top