Sam Coupé SAA1099 replacement.

[wub]

Hi all! This is my first proper retro repair and while it's not massively interesting compared to other projects on the forum, I thought it might be worth documenting as I made a few schoolboy errors. I'd love to get any advice on what I could have done better, I'm guessing everything! :)

My Sam has had a broken sound chip for at least 10 years and I've had a replacement, bought from Colin at Quazar, sitting around for about the same amount of time. Not long after I bought the replacement chip the second hand price of a Sam Coupé starting rising, and I quickly lost confidence. I didn't want to risk damaging something that I couldn't really afford to replace.

I've made a few repairs since then and I also have much better tools now, so I felt that the time had come to be bold. As things didn't go entirely to plan I forgot to take pictures of some of the more incriminating moments. You might think that's convenient, I'm going to say 'no comment' :)

This is the Sam PCB, I've circled the sound chip location in red:

SAA1099-1.JPG

Colin sold me both the chip and a socket and so my initial plan was to remove the broken chip without damaging it so I could take advantage of the socket to do some AB tests. I thought trying to diagnose the fault would be a good exercise as I'm learning how the chip works from a programming point of view at the moment.

SAA1099-4.JPG

SAA1099-2.JPG

As you can see, one side of the chip has a lot of space around it but the other side is parallel to 2 ceramic capacitors and 5 resistors, including 2 big disc resistors. I was going to join all the vias for each row of legs together using a low melting point solder, and then use 2 soldering irons to heat up both sides, but at the last minute I bottled it and decided to cut the chip off and remove the legs one at a time.

In hindsight I think I would have been ok using the 2 iron method but I still think being cautious was the right approach, despite the fact it didn't go perfectly. Ironically, while I managed to avoid damaging the big disk resistors when cutting the legs, I somehow managed to crush not one but both of the tiny ceramic caps. They were so far out of the picture that I wasn't really watching to make sure I was clear of them. I was trying to save as much of the legs as possible, hoping that I could still do something with the chip. I should have just cut them as flush as possible to the chip, coming at them from above rather than at an angle.

I also managed to catch the electrolytic cap next to the chip with the soldering iron while clearing off some old solder from that side of the PCB. It was just the outer plastic coating but I decided to replace it as it didn't look good.

SAA1099-3.JPG

Removing the old solder gave me the opportunity to try out some new flux that Johnny gave me when he worked on my GP32 last year. I've only ever used a fluid flux before now and the new stuff is more of a paste. This meant I was able to put a small bead of flux along an entire row of vias and clean them all off in one hit, without needing to keep reapplying flux after each one. It was way easier and faster this way and I was really pleased with the results. It worked so well that I actually managed to remove a bit of solder from some nearby vias by mistake. Easy to fix but I'll need to be careful with that in the future!

SAA1099-5.JPG

Next I soldered all the new parts on, and this is where I made my second mistake. I was constantly worried about using too much heat and damaging the PCB as I've heard that old PCBs are easier to damage this way than modern ones. So despite having no problems so far I started lowering the temperature of my iron by a few degrees after doing each component. So the caps, that I did first, all soldered up perfectly but by the time I got to the chip I was at about 230 degrees (according to the digital display on my iron, I have no idea how accurate it is!).

At this point the solder wasn't flowing as nicely and despite upping the temperature I never got back into the flow with it. The picture above was taken when I was about halfway through re-flowing the solder on the chip socket, so there's still a few horrors. Don't look at those.

SAA1099-6.JPG

The above is the last picture I took because things were about to get a bit tense.. Everything looked good so it was time to plug in the chip. I've never used a chip and socket like this before and I assumed that the chip would plug in easily. I didn't check to see if that was the case before soldering the socket to the PCB though, and in hindsight I should have plugged in the chip first and then soldered the whole thing on together.

As it was I had to use *way* more force than I was comfortable with and, despite making sure that the chip went down cleanly and level to the socket, it took so much pressure that it went in suddenly with a massive CRACK! I immediately assumed I'd cracked the PCB but thankfully that wasn't the case!

At this point I reassembled it, fully expecting it to go bang or for the chip to be damaged somehow. It was quickly obvious that everything was ok and I've made a video comparing the audio, before and after the repair:





I already knew from playing with E-Tracker that it was channel 3 that was defective. It was still processing sound data but something was getting shifted or modified which really messed up any sound playing on that channel.

So in conclusion, all's well that ends well! There's nothing else I want to repair that would be as expensive if it went wrong so I'm looking forward to doing a few more repairs this year.

Thanks for reading! :)