[sdiy] destroying flash memory cell on a PIC

[Roman Sowa]

That's an interesting story!
At first I thought it's extremely unlucky to get that one single part 
faulty after so many good ones, but then I thought how lucky one has to 
be to make a light bulb out of windowed chip?

I'll be back to work today after 2 weeks, so I will replace the chip.

I feel your pain of early development. Around that year I made my own 
EPROM emulator. It was simply 32kB RAM chip with a capacitor soldered 
across power pins. I used to put it in simple programmer (written in 
assembler for PC) hooked up to printer port. It was basically a socket 
with wires to LPT port. Then I took the chip from the "programmer" and 
moved it to target circuit, which was of course powered but held in 
reset. After releasing reset the sytem started from that RAM. The 
capacitor was enough to keep RAM chip powered during the move.

And frankly, the stuff I did this way was more advanced than things I do 
today with fancy flash programmer :(

Roman


W dniu 2021-03-21 o 17:48, Richie Burnett pisze:
> I was going to ask if you had managed to try a pristine new chip yet?  
> It could just be a faulty chip and you were really unlucky.
> 
> I remember encountering an interesting failure in 1994 in my first job 
> programming PICs...  Back in those days there weren't many flash parts, 
> emulators were quite expensive and the simulator wasn't up to much.  So 
> I bought a bunch of ceramic EPROM based PICs with the little round glass 
> window in the top.  My firmware development cycle consisted of 
> programming up a blank chip, then testing it in a development board 
> whilst another chip was sat erasing in the UV box.  (It's crazy to think 
> what we had to do back then!?)
> 
> Needless to say these chips spent a lot of time being inserted and 
> removed from programmer and dev board sockets.  All had been going 
> smoothly for weeks, then one day I was suddenly greeted with a bright 
> light from the glass window when I put the programmed chip in the 
> development board and powered it up!  It was quite bright, like a small 
> filament bulb, and the firmware wasn't running.  I switched the power 
> off immediately, initially thinking I'd put the chip in the socket the 
> wrong way round. I got a magnifying glass and took a look through the 
> window.  A foreign object, (looked like a piece of metal swarf, but I 
> guess it would have been silicon,) was resting across a bunch of the 
> bond-wires, and two of them lit up like a light bulb when 5 volts was 
> applied!
> 
> I'd been using this chip for weeks without any problem and the offending 
> conductive object must have been rattling around inside since it was 
> manufactured.  They had been handled lots, but on that particular day it 
> decided to make it's presence known!  Just removing the chip from the 
> socket disturbed the short-circuit, and it was difficult to get the 
> offending object in view through the window again.  I marked up the chip 
> with a cross and gave it to the Microchip rep when he was next in, but 
> never heard any explanation back.  Hopefully QC is better at Microchip 
> now.  Maybe not.
> 
> -Richie,
> 
> 
> 
> -----Original Message----- From: Roman Sowa
> Sent: Sunday, March 21, 2021 12:29 PM
> To: synth-diy at synth-diy.org
> Subject: Re: [sdiy] destroying flash memory cell on a PIC
> 
> Thank you all for so many replies. Frankly I didn't expect that many for
> such mysterious matter.
> Let me reply to all in one email separating replies with dash line:
> 
> -------------------------
>