23201a serial adapter overheats on applying power

Hi,

I have a Pololu 23201a RS-232 to TTL serial converter. When I apply power to it, (Gnd on pin 5, 5v on pin 10), the chip gets really hot. If left for more than a few seconds, it start smoking and of course dies.

Any thoughts what could be going on here. I’ve removed the part from the carrier board that I built, and have it plugged into a breadboard with only the 2 power pins connected - still it overheats.

Thanks for any suggestions.

Peter

Hello,

I just replied to your email; please don’t send or post the same thing to multiple places or at least indicate that you have done so. Since this might be of use to others, please continue the discussion here as opposed to replying to my email (if you have any follow up questions or comments).

I suspect the chip on your board is latching up in response to a voltage spike on your input (see this article about LC spikes for more info). What kind of power supply and connection are you using? The IC might already be permanently damaged (it sounds like one of them has not smoked yet), but you might be able to help things by putting an electrolytic capacitor across the power inputs near the board (e.g. on your breadboard), reducing your power wire lengths, or by lowering the operating voltage.

- Jan

I would suspect a spike would take the IC out in the few nanoseconds of powerup, and there would be no overheating. What I’m seeing is the IC rapidly heat in the first few seconds.

The first board definitely had the problem - I let the smoke out of it. The second board gets hot, not sure about #3 and #4.

I’m connected to a 30 watt/5v power supply, and it’s giving me a good solid 5v.
When power is applied to the 23201, it draws 400ma!

Why do you suspect it would have to happen in the first few ns? My (limited) understanding is that a latchup problem could have a continuum of manifestations since an arbitrary number of small gates could latch up, each drawing more power than usual but not necessarily enough for instant failure.

Have you looked at the 5V line, at the adapter board, with a scope? The 400mA is clearly too much, but I can’t tell anything specific from that beyond what I could guess from the chip heating up.

- Jan