Atari 520ST ram issues

[Dopy25]

I posted about this in the ST section but hoping someone in the hardware section might know something.

 

I am trying to upgrade the RAM in my 520ST. I removed the old chips and in doing so used too much heat on one area of the board. It caused a couple of the ceramic caps to burn up BUT all of the traces are fine. The caps seem to not be needed because they don't even come into contact with the new board but I could be wrong. The new ram is an upgrade by agranlund, fixed by a user and posted here on AA. I've double checked the traces on the RAM board, and the main board on the ST. Nothing *seems* out of place. There are no bridged traces on the RAM but I did accidentally bridge a couple once. Could I have blown something on the board by accidentally bridging say the vcc pin on any of the ram chips? Or would it only burn the ram? 

 

I goofed and didn't check if I had any bridging before trying 6 different chips on 3 different expansion boards. All have the same outcome. I do have 4 other chips I can try and will absolutely make sure no pins are bridged but honestly I'm curious to know if anyone may have a theory on this. All the expansion ram boards have the same outcome after I've checked no pins were bridged. 

 

[+Stephen Moss]

12 hours ago, Dopy25 said:

Could I have blown something on the board by accidentally bridging say the vcc pin on any of the ram chips? Or would it only burn the ram? 

Well it is possilbe but it really depends on what (if anything) was shorted to where, but you are more liekly to damage the RAM chips but putting them in the wrong way round, particualry if that reversed power pins so VVC goes the the GND pin.

 

Does it produce the same image with no RAM chips in?

Deen a long time since I dived into/had an ST, if I recall corectly from fitting my Marpet RAM upgrade the MMU is the 4 sided surface mount device and it is very easy to misscount the pins and connect to the wrong one (as I did) or short a couple of pins together when soldering.

[Dopy25]

3 hours ago, Stephen Moss said:

Well it is possilbe but it really depends on what (if anything) was shorted to where, but you are more liekly to damage the RAM chips but putting them in the wrong way round, particualry if that reversed power pins so VVC goes the the GND pin.

 

Does it produce the same image with no RAM chips in?

Deen a long time since I dived into/had an ST, if I recall corectly from fitting my Marpet RAM upgrade the MMU is the 4 sided surface mount device and it is very easy to misscount the pins and connect to the wrong one (as I did) or short a couple of pins together when soldering.

I will post a picture of the board. I don't think the ram is soldered on backwards but I will post before I try another one so I can get more eyes on it. I am sure I matched pin 1 to the right dot on the board.

 

It shows nothing on screen with no ram chips installed.

 

I can't remember which ones were bridged at this time or I would absolutely note that. I will look at them again and compare with the diagram to see if anything jogs my memory. 

[Dopy25]

Here are some pics of the board and RAM. 

 

After checking the diagram for this ram, I'm pretty sure I did not bridge vcc. I think I bridged pins 2&3 on one chip and somewhere around 15-17 (not all pins, just 2 of them) I could also be mixing it up as I didn't take notes on which were bridged, not even mentally. I just was kinda like "oh those shouldn't be there" and used a wick to get rid of it. I guess I could have bridged the other side and maybe it was around pin 37 which says it's VSS. I'm guessing that might not be a good thing either. 

 

I do have 2 boards that are not fully assembled yet, and 4 chips that have been completely unused. I'll finish assembling them and check for bridged pins before supplying power this time. It's just kinda weird to me that they all have the same result so far with those pink and black bars. 

[Dopy25]

Here are a few more pics of the Atari board. 

 

I'm still not sure those old ceramic caps are necessary but can anyone tell me what would be an acceptable replacement for the caps located at c36, c37, c38? When I first tried to remove the old chips, I used too much heat in this area and burned the old caps. One even fell apart. I didn't think they were necessary because the positive side no longer comes into contact with the new pins but I could be wrong. I double checked continuity in this area as well and everything still seems ok. 

 

Also uploaded a pic of the caps that I put too much heat on. 2 have the same label (3127) but the third ends in 3121. I don't know the difference and I can't seem to find a component list that includes c36-c38.

 

[Dopy25]

Looks like they should be 220nF based on the 224 above what you can see in the pic.

The whole label reads:

CGW 224 Z5J 8327 (OR 8321)

 

Looking up this info it seems the most likely is 220nF. Unfortunately I do not have this value or anything remotely close to it. Looks like I will order a few and hope any extra eyes will help in the meantime.

[RodLightning]

It might be a lot of work at this point, but reversing the computer back to 512kb would help you troubleshoot.  Is there a path back to factory condition you can follow or has that bridge been burned (no pun) already?  That route often rules out any damage done to the motherboard so you can focus your attention on the upgrade board.  I don't remember from your other thread if you started with a stock working 520ST or not.

[Dopy25]

It did in fact work before the mod, yes. The issue is that I started by clipping the legs off the old RAM chips before remembering I could use my heat gun. Hence the overheating in the middle of the rows of RAM chips. 

 

I suppose I could remove the 6 pin sockets and install full sockets, and buy more RAM, but I think that would be more work than I want to do, considering it *did* work before any modifications. Although if I can't figure anything else out, I guess that will be my only option 🤷🏼‍♂️. Ugh. 

[+Stephen Moss]

9 hours ago, Dopy25 said:

I can't seem to find a component list that includes c36-c38.

I found a parts list where C36-C39 are listed as 0.22uf (220nF), 25V Ceramic axial, if you have any 0.1uf you could put two in parallel to get 200nF.  

 

 

[Dopy25]

9 hours ago, Stephen Moss said:

I found a parts list where C36-C39 are listed as 0.22uf (220nF), 25V Ceramic axial, if you have any 0.1uf you could put two in parallel to get 200nF.  

 

 

I did have those. Unfortunately adding that in made no change. It's still pink and black bars only.

 

I even tried a third set of board/chips with the exact same outcome. This board had never been installed, and the pins never bridged, or at least powered while bridged. At this point I'm wondering if I have the wrong chips. It's odd to me that they would all have the same outcome.

 

Guess I'll try to revert everything unless there are other ideas. I'll be looking for replacement chips in the meantime.

 

Thanks for the advice. 

[Dopy25]

SUCCESS!!

 

The issue was the routing of the wires to the MMU. Something interfered with them. Not sure what it was but I had an idea to move the wires away from the edge and to again so I placed a spacer between the board and cabling and tried again and it booted! 

 

The other 3 assembled boards also work. 

[Bee]

So to make the upgrade you remove the old ram and make connections to the MMU?

[Dopy25]

15 minutes ago, Bee said:

So to make the upgrade you remove the old ram and make connections to the MMU?

Yes. Remove all the old RAM and solder the board in their place (most of their place), and you make 4 connections to the MMU.