Fried my 6128 - 3 Ohms between GND and VCC

[eto]

While doing some tests with a Scart cable I accidentally powered the 6128 with the 16V power supply of the monitor instead of the 5V bench power supply. A second of distraction and I just inserted the wrong connector. Of course that fried the 6128. Unfortunately I do know have a short somewhere. the resistance between GND and the VCC pins of the ICs is just 3Ohms. 

If I power the mainboard with the bench power supply it goes into overcurrent protection. 

I'm not really sure where to start the investigation and identify where the short is. Any hint is welcome, what I can do to narrow it down or in what order it makes sense to start.

[pelrun]

Augh!

Well you'll have your work cut out for you, but if you're really lucky the RAM died first and protected the GA.

First I'd pull every socketed chip you can, in case that clears the short.

If the short persists, look for any chips that are getting extra hot when power is applied (obviously set the current limit on your bench supply!). One or more RAM chips are almost certainly going to be. Remove those, and then continue with the remaining RAM chips and then everything else until the short is gone.

You can add back the non-RAM chips one by one testing for shorts and discarding until you've got every chip back in place that isn't shorted. Then you can start collecting replacements.

[eto]

Thanks. Yes, I hope it's the RAM. I was wondering if there are other options to find the fault that are easier than desoldering all ICs. If I limit the current, can I do the check for which IC gets hot first? Or is that a bad idea?

I'm not sure if this is a good sign but the resistance between VCC and GND on some RAM ICs is slightly lower than with other ICs. The closer I am to the middle area if the RAM banks, the lower the resistance is. Of course it's not much, just 3.04 vs. 3.08 Ohms, but I thought it might be a hint... I will see.

>Quote from: pelrun on 7/3/2023, 8:20:01 AMThen you can start collecting replacements.

I expected that at some point I will stupidly fry a machine, so I have most ICs already here, that's not the problem ;-)

[pelrun]

Without specialist equipment and techniques you're basically hosed if you want to find all the problems *in circuit*. Which is why I suggested an order that specifically minimised the amount of work involved.

[eto]

Which is why I suggested an order that specifically minimised the amount of work involved.

It was indeed the RAM. 10 RAM ICs were fried, only 6 are still working. Thanks for your clear suggestion.

Well... there's a bright side: since I had to desolder all RAM ICs anyway, I can now upgrade it to 512K ;-)

[GUNHED]

Glad you found the problem and hopefully the fix will work for you. Best luck! 

[eto]

It did take some time until the RAM ICs came from China - but here it is. Up and running with 512KB now :-) I was lucky to have another spare PCB lying around.

Another positive side effect: I have not documented the 512K upgrade yet, as I did not take appropriate care when I did this the first time and forgot where I stored the final version of the PAL sources. I could now verify which ones are correct and I will document that in the Wiki soon. 

As annoying as it was to fry this 6128 out of stupidity, the final outcome is nice :-)