Ambika SMR4 voicecard loud and distorted?

Hi All,
So I bought one of the last Ambika kits many years ago and lost my bottle after soldering a resistor network the wrong way round so left it until I got better at soldering (on smaller projects), I had to replace the motherboard pcb in the end trying to remove that resistor network resulting in removing a pad.

I’ve managed to get it all working and passing the checks apart from one voicecard (unsurprisingly the first one I made) which is much louder, to the point of distorting. Lots of the previous posts have been resolved after incorrect resistors, I’ve checked against a working voicecard and I can’t see any difference (it doesn’t help of course that I wasn’t consistent with my resistor orientation at this point).

I’ve pulled my scope out. So I’ve got the faulty one and a working one and I’ve got the ambika in mono with one oscillator playing a sine.
Here’s the output of the working voice card

Here’s the output of the faulty one

I went looking through the pins and I’ve found that pin 13 on ic4 (LM13700N) on the faulty voice card had a sine

while the working voice card didn’t (or at least it’s very quiet)

I then noticed that it seems like R29 (10k) on both have a sine going into it but on the working voicecard the resistor brings it down to the previous pictures but on the faulty voicecard it stays the same.

I checked the soldering, the voicecard with the green band and the awful soldering is the faulty one, I’ve not managed to find any bridges or shorts, just this gunk (is it flux?)

Here’s a close up of the pins of that resistor and IC (might be a red herring)

Let me know if there’s any more photos I can provide, sorry if this is a really dumb mistake I’ve made that I’m not seeing.

The bottom of that PCB is very dirty (flux). Clean it off with isopropyl or PCB degreaser of some sort. It makes it difficult to look at the solder connections. My guess would be check out the resistors around IC4. The resistor color code can be used to determine values. and some resistors may be measured in circuit depending on what they are connected to. However, you may be able to measure differences between the good board and the bad one.

Thanks for the response! Yeah I guessed it might be too dirty to check but I didn’t know isopropyl alcohol would fix that, I’ll go through everything and clean it.

I’ve tried reading the colours and it all looks correct measuring the resistance doesn’t always give the resistance I’d expect (I don’t have the measurements to hand annoyingly but the 18ks say they’re a single digit resistance, the 10ks are fine) but it gives the same measurements across both boards so, maybe naively, I assumed that the resistors are correct and it’s just the fact they’re in the board that throws it off.

When I’m next by the board I’ll put the scope round the resistors and see what the differences there are.

Well I cleaned it up using isopropyl, couldn’t see anything other than some of the solder joints looked a bit funky so I reflowed the solder. Then the output was a clean sine as expected!

Then I put it back in and I noticed that layer B on the voice cards wasn’t trigger (I’ve set it to mono). I’ve swapped all the voice cards round and tried different combos and it just seems like none of the voice cards on layer B are triggered, they have power going to them but no data. Both the headers have the jumpers on B on the stacked ones and A on the ones directly on the mobo. I don’t know if this is an improvement or not haha.

Well I’m an idiot, I had those voices off for some reason…it’s all working it seems!!