exxos's Atari, Amiga & retro forum

I am also working on 4 ZX Spectrums I have obtained and have found a strange mod in a 48K issue 3 which I can't work out what it actually would do?

Anyone have any ideas on this mod?

I have installed a Cap Composite mod but the Yellow wire from this little board goes to the +5v for the modulator and as you can see the other two wires go to the buzzer. There are no other connections that I can see.

Doing some digging and taking the fact that it is connected to 5v and the speaker, my best suggestion would be that it is some sort of amplifier to boost the speaker.

Or maybe audio injection into the RF output?

stephen_usher wrote: 15 Aug 2026 17:30 Or maybe audio injection into the RF output?

Definitely maybe, especially since we can only see 5v and the speaker connections, there may be something else connected we can't see.

There is definitely only the 3 connections on this little board as I have removed the heatsink which looks like it's been modified by have a couple of chunks cut out of it and I also found IC26 which is a 74LS158N has pin 11 cut and is fitted into a socket (blue) with pin 10 of the IC connected to the pin 11 location on the socket?
This looks to be changing the Hi Low jumper T1 which is set to TI but the chips upper ram chips are Oki?

And it should be a 74LS157 I believe not a 74LS158? Edit: hmm IC25 is also a 158 but should be a 157. From what I understand the 158 has the same footprint as the 157 but then output is inverted?

Acsi wrote: 16 Aug 2026 00:23 There is definitely only the 3 connections on this little board as I have removed the heatsink which looks like it's been modified by have a couple of chunks cut out of it and I also found IC26 which is a 74LS158N has pin 11 cut and is fitted into a socket (blue) with pin 10 of the IC connected to the pin 11 location on the socket?

And the yellow wire definitely goes to the 5v of the modulator? Yellow is often used for video, and also if it is 5v where is GND? Or is taken from the speaker GND? The schematic seems to suggest this could be the case.

I like @stephen_usher's idea that maybe it is injecting audio into the video signal for Audio via the TV.

Acsi wrote: 16 Aug 2026 00:23 This looks to be changing the Hi Low jumper T1 which is set to TI but the chips upper ram chips are Oki?

And it should be a 74LS157 I believe not a 74LS158? Edit: hmm IC25 is also a 158 but should be a 157. From what I understand the 158 has the same footprint as the 157 but then output is inverted?

Yes, they should be 74LS157s going by the schematic and IC25 & IC26 are used in addressing, but the 157 is a non-inverting multiplexer and the 158 is an inverting multiplexer. If there are other changes to the system then they may be done to account for the difference between the two.

Looking at the datasheet the pin 11 being cut may be to counter the inversion of the 158. Perhaps they couldn't get a 157?

And the yellow wire definitely goes to the 5v of the modulator? Yellow is often used for video, and also if it is 5v where is GND? Or is taken from the speaker GND? The schematic seems to suggest this could be the case.

Yep definitely +5v and no connection to the RF signal input on the modulator unless somethings already been removed before I got it.

Perhaps they couldn't get a 157?

I do wonder about this but surely that would also invert the other gates or maybe because the other gates which go through IC25 are also inverted given it's also a 158 it actually works somehow? This is beyond me and I do have a few 157's I believe so will put it back either by changing the links to Oki or I think I have enough TI chips to do a complete swap.

With the audio mod unfortunately the sponge which is glued to the bottom of the little circuit board is stuck quite well and the sponge of course after all this time and heat is crumbling so I need to remove it to clean it up so I can see the circuit wiring which may give a better idea of what it is..

Acsi wrote: 16 Aug 2026 09:38 Yep definitely +5v and no connection to the RF signal input on the modulator unless somethings already been removed before I got it.

Could have been I suppose, but I'm guessing we'll probably never know.

Acsi wrote: 16 Aug 2026 09:38 I do wonder about this but surely that would also invert the other gates or maybe because the other gates which go through IC25 are also inverted given it's also a 158 it actually works somehow? This is beyond me and I do have a few 157's I believe so will put it back either by changing the links to Oki or I think I have enough TI chips to do a complete swap.

I 'think' it works because pin 11 is cut and that is giving an inverted output as it only has one input, but that doesn't really fly for the other gates unless something else has been done.

TBH I don't bother with 4532's I drop in 4164s and usually they are fine.

Acsi wrote: 16 Aug 2026 09:38 With the audio mod unfortunately the sponge which is glued to the bottom of the little circuit board is stuck quite well and the sponge of course after all this time and heat is crumbling so I need to remove it to clean it up so I can see the circuit wiring which may give a better idea of what it is..

While IPA is usually good for cleanup of this sort of thing, WD40 (The lubricant) can actually do a fantastic job of removing sticky residue.

I had to replace a keyboard surround on a BBC B and it did the job of getting the residual adhesive off in quick time.

So WD40 took the sponge off in lots of pieces (Think it just disintegrated as the middle had crystallised) but the glue is a solid mass which isn't budging yet.

Anyway I removed the 3 wires attaching it to the board and then tested the board. Low and behold the noise I had on the video output has almost completely gone and the colours are now stable so it must have been injecting the audio as @stephen_usher suggested.

I have also removed the 2 74LS158's and put in 2 74LS157's and changed the jumpers to Oki and Low to match the installed ram and the machine tests fine so far.

I also had a heatsink from a previous 48k which I installed a Traco in and it matched the original so I have put this one in instead of the modified one.

Acsi wrote: 17 Aug 2026 10:54 So WD40 took the sponge off in lots of pieces (Think it just disintegrated as the middle had crystallised) but the glue is a solid mass which isn't budging yet.

Anyway I removed the 3 wires attaching it to the board and then tested the board. Low and behold the noise I had on the video output has almost completely gone and the colours are now stable so it must have been injecting the audio as @stephen_usher suggested.

I have also removed the 2 74LS158's and put in 2 74LS157's and changed the jumpers to Oki and Low to match the installed ram and the machine tests fine so far.

I also had a heatsink from a previous 48k which I installed a Traco in and it matched the original so I have put this one in instead of the modified one.

Cool, glad to hear it's working