Reverse socketing CMOS

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View
Just inadvertently reverse socketed a 4001 quad nor on a prototype. It got
stinking hot, after I turned the power off I decided to just swap it round
and try for luck. It works OK, just wondering how lucky this event was. Or
is CMOS really reasonably (unofficially) reverse-polarity tolerant?

The supply rail is fairly stiff at 12V, it would have held up with at least
an amp.

Re: Reverse socketing CMOS

Quoted text here. Click to load it
I've seen reversed 8748s (windowed microprocessor) with the internal
power links glowing brightly. Put them in right way around and they
work ;)

Reason is the substrate diode.


Re: Reverse socketing CMOS

Quoted text here. Click to load it

It happened once to me with a guitar effect pcb containing no less than 5
or 6 opamps. All sockets soldered backwards so that the inspection before
powering it up went ok. I kept that board powered a good half minute
before realizing what went wrong, but it worked like a charm once the
opamps were put back in the correct orientation.

Re: Reverse socketing CMOS
Quoted text here. Click to load it

You are lucky.  I once did this by mistake with a couple of 2716
ROMS.  They
didnt survive, and it was a huge hassle to source replacements from
the USA

(this was circa 1985, and there was no internet)

In your case reverse protection diodes on the chip probably saved it,
you turned it off before they got too hot and do physical damage to
the IC.

I remember a similar incident where a 6810 RAM had the negative pin
up under it when inserted in the socket.  This worked fine, even with
no earth.
It was discovered by accident some unknown time later (would be months/

I think this is another case of the protection diodes on the pins
saving the day by "making" a substitute
path to negative. Obviously the chip could cop the voltage drop

Site Timeline