Issue 2 with short to GND on MREQ and A10

[MrClump]

I have an Issue 2 that was not booting, IC2 did not appear to be functional and was getting hot; IC1 appears fine when testing in a known good Spectrum.

Upon further examination, with both IC1 and IC2 removed there is a short to GND on both A10 and MREQ. When the Spectrum is powered up neither IC5, IC23, IC25 or any of the upper ram chips are getting hot.

I'm not really sure what my next step should be, I've studied the board and so far haven't found any dodgy solder joints, at least not any that look like they are causing this issue.

Could it be the upper ram, should I start removing upper ram chips?

Any pointer would be appreciated.

[zx64]

Is it constantly 0v on these pins when Spectrum is on, or did you measure the resistance between pins and ground when Spectrum is off?

[MrClump]

Is it constantly 0v on these pins when Spectrum is on, or did you measure the resistance between pins and ground when Spectrum is off?

I’m measuring resistance to ground with the Spectrum off.

[1024MAK]

If you can’t find a short by careful visual inspection (under a good light, and with magnification if you have old eyes like me), then you will need to use electrical testing and remove all possible suspects one by one.

It’s unusual for two very different signals to be both shorting to a supply line. Are you sure they are not shorted together?

Also what is the actual resistance value of this short?

And finally, I presume you measured this with all socketed chips fitted. One at a time, remove any of the socketed chips that you think may be the cause.

Also /MREQ and A10 don’t directly go to any of the DRAM chips.

Mark

[MrClump]

It’s unusual for two very different signals to be both shorting to a supply line. Are you sure they are not shorted together?
Mark

No, i'm not sure.

Also what is the actual resistance value of this short?
Mark

I was only checking for continuity at this stage, but if i measure the resistances it is:

A10 - 13.7ohms
MREQ - 27.3ohms

And finally, I presume you measured this with all socketed chips fitted. One at a time, remove any of the socketed chips that you think may be the cause.
Mark

So far I have been taking measurements with both the ULA and Z80 removed. The Z80 was non-functional, confirmed when I put it into a known working Speccy, perhaps fried from getting so hot (too hot to touch) if the previous owner left it powered up for some time? The ULA is working just fine when I put it into another Speccy.

I did briefly put a known working Z80 in along with the ULA and powered it up and the Z80 very rapidly started to get hot, so immediately switched it off. The Z80 fortunately still worked afterwards when returned to its normal home. I don't intend putting any more power through this board unless I have to until I've located the issue.

Also /MREQ and A10 don’t directly go to any of the DRAM chips.
Mark

Noted. I am going to get my even bigger magnifiers out and have another look over the board to see if there are any whiskers of solder or bridges.

[1024MAK]

/MREQ goes to:
IC1 (ULA) pin 38,
IC2 (Z80) pin 19,
IC5 (ROM) (via the H / N links)(pin 20 or 27),
IC23 (74LS32) pins 1 & 2 (controls the ‘upper’ RAM),
The edge-connector expansion port.

A10 goes to:
IC2 (Z80) pin 40,
IC3 (74LS157 or 74LS257),
IC5 (ROM) pin 21,
IC25 (74LS157 or 74LS257),
D2, (keyboard circuitry),
The edge-connector expansion port.

Mark

[MrClump]

/MREQ goes to:
IC1 (ULA) pin 38,
IC2 (Z80) pin 19,
IC5 (ROM) (via the H / N links)(pin 20 or 27),
IC23 (74LS32) pins 1 & 2 (controls the ‘upper’ RAM),
The edge-connector expansion port.
Mark

I have made some progress on /MREQ. I have now also removed IC23 along with IC1 & IC2 and there is continuity on all the pins (MREQ to my ROM is on 27) but with NO continuity to ground on any of them any more. So I guess IC23 was bad?

I did have to cut IC23 out to remove it, but at least it is cheap to replace.

[zx64]

I was only checking for continuity at this stage, but if i measure the resistances it is:

A10 - 13.7ohms
MREQ - 27.3ohms

Noted. I am going to get my even bigger magnifiers out and have another look over the board to see if there are any whiskers of solder or bridges.

It's not a complete short, so no need to use magnifiers, your board is fine. Some of ICs are bad

[zx64]

I have made some progress on /MREQ. I have now also removed IC23 along with IC1 & IC2 and there is continuity on all the pins (MREQ to my ROM is on 27) but with NO continuity to ground on any of them any more. So I guess IC23 was bad?

Correct. Now repeat the same with ic3 and ic25. Toss a coin which one you cut first.
You may try to cut just suspicious pins, no need to remove entire chip - most likely one of ic3,ic25 is still good.

[MrClump]

Correct. Now repeat the same with ic3 and ic25. Toss a coin which one you cut first.
You may try to cut just suspicious pins, no need to remove entire chip - most likely one of ic3,ic25 is still good.

It looks like both IC3 and IC25 were also bad.

I cut the leg on pin 11 of IC25 first, short still present, did the same on IC3 and short gone! So I then decided to experiment and temporarily bridged the gap on IC3's cut leg and the short returned!

I've now completely removed, desoldered and cleaned the area and have soldered in two new sockets ready for replacement chips.

I'm hoping that now resolves all the issues with this poorly Speccy, TR4 & TR5 have also been replaced (they were also shot), the VRM has been replaced and I have a new replacement lower ram module to install. Hopefully, once it has been recapped and a replacement Z80 installed all will be good again. I'm not going to apply power now until the recap has been completed. I will post an update once that is done (could be a few weeks)!

Thank you ZX64 & 1024MAK for your advice it has been invaluable. I have to say this is starting to get addictive!

[1024MAK]

For so many chip to have died is unusual, unless either the power rails went significantly over their normal voltages or they were subjected to reverse voltage (or AC).

TR4 and TR5 die with a reverse voltage, as does the 7805 voltage regulator.

Certainly don’t fit any of the socketed chips when you do the first power up, as all the supply voltages need testing.

Mark

[MrClump]

For so many chip to have died is unusual, unless either the power rails went significantly over their normal voltages or they were subjected to reverse voltage (or AC).

TR4 and TR5 die with a reverse voltage, as does the 7805 voltage regulator.

Certainly don’t fit any of the socketed chips when you do the first power up, as all the supply voltages need testing.

Mark

I had identified that TR4 had failed before applying any power or even realising the shorting was an issue, so had already replaced that along with TR5 and the 7805 as a precaution. With that, the logic chips and the fried Z80 something has clearly gone very wrong here! I guess I could yet have more problems when I eventually get to the point of being able to power it up and will definately be having a good poke around with the multimeter before either the ULA, CPU or lower ram go back in.

But if I can get what would otherwise have likely ended up in landfill working again then I will be happy.

[zx64]

But if I can get what would otherwise have likely ended up in landfill working again then I will be happy.

I'm pretty sure you will make it work. You already tested ULA and it's good. Everything else is easily replaceable.

[MrClump]

Replaced the memory logic chips, checked some voltages without the CPU, ULA and lower ram in place, all looked good.

Socketed all the chips, fired her up........ nothing!

Jeez, what now? Then I spotted the ULA chip on the bench, if it was there what had I put in the Speccy? I’d accidentally put in my faulty ULA! Put in the one the correct one and then, 1982 Sinclair Research Ltd

Many, many thank yous!

[zx64]

[1024MAK]

Well done MrClump, you’ve just brought a Speccy back from the dead

Now go and use it for fun

Mark

[MrClump]

Well done MrClump, you’ve just brought a Speccy back from the dead

Now go and use it for fun

Mark

I want to do another one now.