I did something stupid burning an EPROM

[ThumpNugget]

As the title says, I think I may have done something really stupid. I've been working on a game for the 2600 in my spare time for the last couple of months. It is working on an emulater and is far enough along that I wanted to try it on a real machine. I burned an EPROM and placed it in a PCB I purchased from AtariAge (the bank switching fully socketed type for developers to test). I then handed the PCB with EPROM installed to my son and asked him to stick it in the 2600 jr. nothing... I asked to look at the PCB and said "Oh, its in backwards! flip it and try it again". I was referring to the EPROM in the socket however my son assumed I meant the PCB so we walked over to the machine and placed the PCB in the wrong direction. It was only like that for a few second but it become very hot in those few seconds!

 

I flipped the EPROM so it was in correctly and tried again but again it did not work. I put the EPROM back in the burner and was able to read the data just fine and it mached my BIN file so the EPROM is fine and correct... So the only conclusion I have at this point is that I've destroyed the PCB but this is out of my area... Is it possible to destroy a PCB if its put in a machine backwards? (or if the EPROM was put in backwards?)

[Quadhorn]

Meh, these things happen. Even in industry, someone'll decide the machines should mount pairs of components 90 degrees to how they should be for no reason whatsoever. It gets pretty and the tantalum capacitors smell of fish...

 

Well, if the chip's ok and the Atari's ok, then as you say it can only be the PCB. Time to pull out the multi-meter and check the traces one-by-one. Yes it's possible to damage it if, for example, you end up shorting the power rail to ground through one or more of the traces.

 

If you don't have a spare PCB, I s'pose you'll have to check the traces and patch up any that have gone open-circuit. Shorted traces (? possible?) might be a little more awkward.

[supercat]

I flipped the EPROM so it was in correctly and tried again but again it did not work. I put the EPROM back in the burner and was able to read the data just fine and it mached my BIN file so the EPROM is fine and correct... So the only conclusion I have at this point is that I've destroyed the PCB but this is out of my area... Is it possible to destroy a PCB if its put in a machine backwards? (or if the EPROM was put in backwards?)

 

I'm surprised the EPROM survived. Usually they will be fried by VDD/VSS reversal.

 

The cart has grounds on diagonally-opposite corners (unlike the chip!) so putting a board in backward will usually not damage anything, though it will of course not work.

 

Not sure about the combo of backward cart and backward board. Wouldn't think that would be a problem, but it could.

 

Does the 2600 still work?

[supercat]

It gets pretty and the tantalum capacitors smell of fish...

 

I saw a board where someone had omitted polarity markings for the tantalums. One of the tantalums was directly over a via on a 4-layer board that ran off a 15 amp supply. Took awhile to figure out why boards were getting dime-sized holes in them.

[Quadhorn]

I saw a board where someone had omitted polarity markings for the tantalums. One of the tantalums was directly over a via on a 4-layer board that ran off a 15 amp supply. Took awhile to figure out why boards were getting dime-sized holes in them.

 

I'm kind of surprised I had as little damage to this particular set of boards as I did; we had swapped R/C pairs (on power lines, not as signal filters), the decouple caps 90 degrees out and even a 16-pin layout that should have been 14-pin. Still, we learn; once upon a time I thought tant's were non-polar, 'cos they didn't have negative signs showing the polarity ...

[Bruce Tomlin]

If nothing else, putting an EPROM in a socket backwards will probably fry the D4 data line. I blew one a few weeks ago this way by not paying attention.

 

I've even used EPROM programmers that could destroy a chip plugged in backwards. (the better ones can detect that something is wrong by seeing too much current draw)

Edited June 19, 2007 by Bruce Tomlin

[ThumpNugget]

I'm surprised the EPROM survived. Usually they will be fried by VDD/VSS reversal.

 

The cart has grounds on diagonally-opposite corners (unlike the chip!) so putting a board in backward will usually not damage anything, though it will of course not work.

 

Not sure about the combo of backward cart and backward board. Wouldn't think that would be a problem, but it could.

 

Does the 2600 still work?

 

I was surprised as well.. I thought the EPROM was the only possible loss when it happened but it was fine. The 2600jr is fine (played a quick game of space invaders afterwards and all was well). I just ordered a new PCB so I guess I'll know soon enough... I should have just assumed that the one thing that I only have one of would be the thing I toast!