ZX Spectrum - Issue 4B - problem between chair and keyboard

[unhuzpt]

Hello everyone!

So here is my call for help and kindly ask if people can help me find and solve whatever problem i created.
For starters i have a very basic understanding of electronics but still trying to figure things out; then i bought the basic tools to try to help finding the problem: digital multi-meter, more or less decent microscope and solder stuff.

Here is a description of how i got to this situation:

- it was working fine all stock i could load games with mobile phone and play, sometimes some would crash but seemed ok
- then i found there is a replacement for the power input converter - TSR1-2450 - that we could swap in to remove the heating and the big heatsink so i replaced it and it was working just like was before
-then i found the composite mod and i did it and all worked as before no issues
- then i found the advise to recap that seemed logic to me due to very old caps so i ordered a recap kit; here i made a mistake as i replaced all without 1 at the time and test (i think i tested after the first 2 that were the ones near the video out circuit and was ok)
- after replacing all it does not work anymore correctly.... i had very few some good boots, with color in the image and dandanator board giving ok in the tests, them random failed boots with not even sinclair text and quite random results. Then i did my best to find any solder issue or short, then i resoldered many many pins - all lower memory, ula, caps, and many more.
- then i put an lower memory ic backwards and things got worse so i ordered the lower ram board that replaces the ics

So after learning the hard way the "if ain't broke don't fix it" i now struggle with this. I have no idea why and how but every time i retouch the solder of pins i get different behavior after.
Until this morning it would boot again to the normal sinclair text, then after like 10s make some beeps and change to a white border and black screen in middle.

Then I re-soldered some more pins (i think around the high memory ics) and now i have boot to the sinclair text, then it beeps a few times and nothing else - no colors shown also.
Ah and if i boot with dandanator board it boots run those checks and then "ram fail 01234567" and no colors

Now this morning i recaped again (new caps), did my best to check for shorts or bad solders, re heated most pins in the lower/upper ram, ula, whatever else.

Now i try to start to look into power delivery to the board and i think i have issue there. But as now i have the lower ram board from retroleum i am not sure how to check the voltage on lower ram. I suspect i have some issue on the board power delivery somewhere.

In C79 i read 0v
In C44 i read 1.9v
In C45 i dead 1.9v

No idea if these are correct

I have uploaded to my google drive a couple of videos and picture of my machine:
https://drive.google.com/drive/folders/ ... sp=sharing

[dfzx]

- then i put an lower memory ic backwards and things got worse so i ordered the lower ram board that replaces the ics

Just to be clear, you put a 4116 in with the wrong orientation, then powered up? That would have connected +12V through that IC and to goodness knows where. You might have fried part of the ULA. Is the ULA socketed? Do you have a working Spectrum you can raid for test parts?

You should probably start fault finding from scratch. Take out the lower RAM module and disable the upper RAM. Check voltages. -5V doesn't matter, but the others do. If voltages are OK, does your test ROM run tests and seem generally stable? Ignore the results (with no RAM they won't get far), you need to see if the program is running for a few minutes, the video seems OK, etc.

[unhuzpt]

Just to be clear, you put a 4116 in with the wrong orientation, then powered up? That would have connected +12V through that IC and to goodness knows where. You might have fried part of the ULA. Is the ULA socketed? Do you have a working Spectrum you can raid for test parts?

You should probably start fault finding from scratch. Take out the lower RAM module and disable the upper RAM. Check voltages. -5V doesn't matter, but the others do. If voltages are OK, does your test ROM run tests and seem generally stable? Ignore the results (with no RAM they won't get far), you need to see if the program is running for a few minutes, the video seems OK, etc.

Hello dfzx!

Yes i put a 4116 the wrong orientation and powered it on.... yes the ULA is socketed but i do not have another to test... for the moment as i ordered one today I am going to check as i saw some post on how to disable the upper RAM and test the dandanator test rom.

Also i did not realize the google link was not correct... can you please take a look?

https://drive.google.com/drive/folders/ ... hEvbzEt1LD

[dfzx]

So, booting with the external ROM card sees the Spectrum come up normally? The video seems to demonstrate that? Is the machine stable for that minute or so, in that you can type stuff and make it BEEP, etc?

If it seems OK, then that seems to point to a building heat issue. Can you find any device which seems to get very hot? Use a fingertip on the ICs and transistors. The ULA will always get pretty hot, the rest of it shouldn't get much above warm. Best not run it too long (although it does seem the damage is already done).

If you boot up the Dandanator card to the RAM Fail message, then leave it for 2 minutes, does it go weird like at the (C) message, or does it just sit there happily?

Check the voltages before you go any further. Then disable the upper RAM to eliminate issues where the stack ends up in unreliable memory.

[1024MAK]

It's no good hiding here as well as posting elsewhere! I'll still find you!

Okay, best not to do anything else until I've had a chance to think about this...

You say you have a digital multimeter, do you know how to measure resistance, continuity and voltages?

Mark

[unhuzpt]

It's no good hiding here as well as posting elsewhere! I'll still find you!

Okay, best not to do anything else until I've had a chance to think about this...

You say you have a digital multimeter, do you know how to measure resistance, continuity and voltages?

Mark

Hello Mark

I have replied to you on the first topic.

Yes i can measure resistance, continuity and voltages but also i have a very basic understanding of electrical circuits so i may ask basic things.
I have just done what was proposed that was to short IC23 pin 5 to the plus side of C34 where i measure +5v. Hope this is not another fail....
So now i removed the lower ram board so no lower ram ics are in.

[unhuzpt]

Update:

Following kind Marc instructions i have:

Measurements (without ram board plugged in):

Across C50 -> 11.83v
Across C34 -> 5.01v
Across C44 -> 1.92v
Across C45 -> 1.89v
Across C79 -> 0v

Went further if would be useful:
Across C80 -> 9.90v
Across C46 -> 11.56v
Across C74 -> -0.17v
Across C27 -> 4.90v
Across C25 -> 4.88v
Across C65 -> -1.59v

[1024MAK]

For those following along at home (or wherever else you are browsing this forum), the OP also opened a topic over on Sinclair ZX World here.

Mark

[unhuzpt]

Updated info for the next steps:

So, the +12V and -5V are not working. Without a +12V supply, the video will only be monochrome.
It's likely that TR4 and maybe TR5 are damaged. But do not remove them yet.

Before hunting for other non-power supply problems, the power rails need sorting out.

--------------------------

Then after this i went to check TR4 and 5 for continuity:

Does it mean anything that TR4 has continuity between the center leg and lower leg?
Is between base and collector? Tried my best to find that

In terms of solder no short either up or down on the board.

[unhuzpt]

So yeah, TR4 is dead as it seems.
Ordered and waiting for new to arrive and then the saga continues, just hope i did not fry the new ram module board....

[unhuzpt]

Latest update:

So i replaced TR4 and now yes there is color in the image and first time it booted into normal sinclair screen but after some seconds started beeping and went to black center with some red stripes in the top of the square. I had the lower ram board plugged in.

Now i removed the ram board and went again to measure the voltage on the caps as you told me to do before:

Across C50 -> 11.73v
Across C34 -> 5.02v
Across C44 -> 12v
Across C45 -> 11.70v
Across C79 -> 13.68v


Across C80 -> 2.60v
Across C46 -> 15.63v
Across C74 -> 0.66v
Across C27 -> 4.90v
Across C25 -> 4.88v
Across C65 -> 0.60v