A tale of 2 Spectrums...

[JetSetWally]

Hello to you all, its my first post so please bare with me.

Many years ago, i picked up a couple of spectrums from a car boot sale, got home plugged them in, played for a while, discovered small faults but i never got around to fixing them, until now

1 - 128K toastrack ( issue 6U ) - New keyboard memebrane, clean, changed the Voltage reg. new CAPS, checked operations with a retroleum diag. rom , run a couple of games --- GREAT one fine working Spectrum saved.

2 - 48k (issue 2 ) Booted up but would crash so I thought I fix it-----that's where my decent in madness ( and poverty ) started about a month ago. It has been a rollercoaster ride from joy to dispair, I have spent many hours following video's, soldering, desoldering, measuring with my multimeter, building up only to tear back down and STILL NO JOY.

Everytime I seem to fix one thing, another problem rears it ugly head.

From working but unstable, to corrupted video, no video, dark video, black border , white border, multi coloured borders, lines, b&w only video, then working again , not working , blocks on screen, messed up chr$ on screen, sound, no sound, you name it - Its done it

I have had it run & pass Retroleum Diag rom tests - only to fail them the next day !

Now every time I plug it in I get a different random screen / border . I have performed a complete visual check of both sides of the board with a jewellers loup and can't see any damaged / bridged tracks.

I have purchased and replaced the following

Upper RAM x 8
Lower RAM x 8
ULA * nebula ic1
ic2
ic3
ic4
ic23
ic24
New Electrolytic Caps
TR1,2,4,5
VR1,2
d9,10,13,15
Voltage Reg.
Kbd. memebrane
c5,6,7,8
Comp. Video Mod.
Kb1 connector

And as that wasn't enough, I bought a Oscilloscope two days ago!--- I don't even know how to use one ( but i'm learning hopefully )

This board is swiftly turning into "Triggers Broom"

So far I have spent a small fortune and have gone nowhere. Only ic's 5, 25 & 26 to be replaced.

Also the RAM voltages are OK - 0v, 12.3v, 5.01v, -4.38v ( a bit low but within limits )

Please can someone advise me to give up and buy a working board.

ps- used the scope on pin 39 of the ULA and i think it was reading about 3Mhz. - it should be 14Mhz I think ?

I'm in the same sort of mood as Sir Clive was when he saw the C5 sales figures.

Its Prob. going to be something small & simple that I'v missed

Thanks for listening

[1024MAK]

I don’t recommend testing ULA pin 39. If the ULA produces a video picture, then the 14MHz oscillator is running. If you want to see that the ULA is working, test the Z80 clock signal on pin 6 on the Z80. But note that it may vary slightly as the ULA contention to the Z80 is applied via the CPU clock.

So you listed rather a lot of symptoms.

Which chips are in sockets?
And which of these are original sockets?

And what are the current symptoms?
Include photos of the board and the screen display showing any problems.

Mark

[1024MAK]

Oh, my bad, I forgot to welcome you

Time for me to correct that, so...

Welcome on board JetSetWally!



Welcome to Sinclair ZX World

Mark

[JetSetWally]

Hi Mark, thank you for your kind reply. I will try to get all the info you have requested ASAP, I work early in the morning so I will try and post tomorrow after work tomorrow. I hate to let a retro computer die and even if it kills me, I'll get this working.

Frustrating thing is, at one point I had it working with the original 16k RAM (upper removed) for at least 20mins. running code !

Now I'm back to square one.

But I'll keep trying....

[JetSetWally]

Right , I've removed thr diag. ROM and restored the board to the following picture's

This is the original upper / lower RAM.

When power is applied I get the following...

Then I removed the upper RAM - video still corupted

Removed both the upper and Lower ram and got a white border, white paper with a rolling pattern cycling around.
I took some video but its file size is too large, I'll re-do the video soon and post it

I changed the lower RAM to my new stock, but no luck

Voltages - -4.35v 12.37v 5.01v 0.00v ( Old RAM chips )

Then I got some clue as to the posible problem - as I rechecked the voltages with the new RAM on IC13 ( -4.38v 12.32v 5.01v 0.00v ) I THEN decided to check at IC6 and got 8v at ground ! Measured again and got 0v, measured again and got 7.6 and rising, then 0v again

Now I just need to figure out WHY !

I suspect a faulty ram module leaking volts to ground ?

I will investigate further this weekend.

originaly IC1 and IC5 were socketed, Now I have added sockets to all the RAM (upper/lower) & IC14.

I'll try again.....

[1024MAK]

I THEN decided to check at IC6 and got 8v at ground ! Measured again and got 0v, measured again and got 7.6 and rising, then 0v again

Now I just need to figure out WHY !

I suspect a faulty ram module leaking volts to ground ?

Hmm, what point are you using for the negative lead (black probe) for your multimeter?

The heatsink tab of the 7805? The case of the modulator? Somewhere else?

It’s far more likely that there is a break in the 0V/GND network rather than anything “leaking volts” into it (as that would be a short circuit).

Mark

[RobG]

Is it possible that one of the traces is damaged?

[1024MAK]

Is it possible that one of the traces is damaged?

Yep, that’s what I was getting at!

It is possible to test for that by using the continuity range or the 200 ohm (or equivalent) resistance range on your multimeter. Do the tests with the power disconnected from the computer. Put the negative (black) probe on the heatsink tab of the 7805 voltage regulator. Then use the positive (red) probe to test each and every 0V/GND pin on all the 4116 (or equivalent) DRAM chips. Do the same for any other chips where you suspect there may be a problem.

If the tracks / traces to the chips you test are good, the meter should indicate good continuity (normally via a tone) or show a resistance of less than 2 ohms resistance.

Mark

[JetSetWally]

Hello again everyone, and thank you for your kind help, I was working at the weekend so I decided to not touch the spectrum until I could concentrate on the matter.

So, this afternoon after work, I decided to give it another go, I checked the continuity of the ground to pin 16 of all the lower RAM, no problems were detected, again I checked for shorts with a jewellers loup all the PCB tracks, again all good. I checked the 8 lower RAM chips ( out of the sockets )with a multimeter for resistance on pins 1,8,16 ( I followed instructions from another thread onthis forum ) again all good.

I then took out all the socketed IC's to check the tracks, again all good. I then cleaned the board, and dried it with compressed air, paying attention to the DIL socket's.

I then re-examand the front face of the board, I noticed that on IC's 3,4,24,25 five or six of the chip pins, the solder had not flowed all the way through from the rear of the board - I quickly re-flowed the solder on these pins.

Next I reinstalled the ROM, Nebula ULA, IC14. I then powered up the spectrum and got thr expected screen display. All good again.

Now for the lower RAM chips ( the original ones ) . Switched on and I was welcomed with the correct Sinclair logo.

I switched it off again and connected the keyboard and rebooted. Again all working. I quickly typed in a short program to find max RAM, random paper / border colour changes -OK again.

I hooked up a cassette player and loaded an MC assembler program. first attempt failed half way through - I adjusted the volume and it then worked.

And thats where I have left it . I've been here before with this machine so I'm in no rush to put in the upper RAM in case it causes another problem.

Now does anyone know a way to check the upper RAM while its out of the board i.e. via resistance with a multimeter, or am I going to have to bite the bullit and put them in and risk upsetting the machine again ?

Thanks again for all the help and support

Happy now we're getting somewhere

[JetSetWally]

...forgot to say the BEEP command produces no sound but the coil noise changes when it is called. also when I loaded the program from a cassette I could faintly hear the speaker on the spectrum......new speaker perhaps

will test it more this week.

Thanks again