"Dead" ZX81 Issue 1

[iturbez]

Well... Good and bad news!

I've measured signals with a 10X test probe... Oscillator is OK. CLOCK signal appears at PIN 6 of Z80... but no video signal at PIN 16 of ULA...

I change Z80 CPU... VIDEO SIGNAL APPEARS AT PIN 16 OF ULA... So I power my TV on... I get a white screen, but no "K" cursor...

What do you think I must try next: ROM? RAM?

Thanks, guys...

[PokeMon]

Try to measure A15 signal with you scope, should look like this:

zx80core_ram-test3_s.jpg (83.33 KiB) Viewed 4843 times

This is with time resolution 20us / div.
You can see the video output (A15 high) with a period of about 42ms and a silent period of about 22ms, all in all 64 us (one horizontal line).
If you see this pattern, you can be sure the ROM and RAM are working.
So you can see about 35 pulses in a package for the 32 chars and some extra pulses for newline and interrupt acknowledge.

If you have this, try to measure out the ULA output if you have 3 voltage levels (sync, the lowest, black (middle) and white for highest voltage).
Most defective ULA outputs will deliver sync level and white level but due to internal failure no more black level.

See attached signal:

zx80core_video-backporch_s.jpg (100.22 KiB) Viewed 4843 times

Here you can see all 3 levels and beginning with sync level, than backporch black level (you maybe don't have a backporch depending on the ULA type) and in the middle you can see the black level in picture for the K cursor, you maybe press simply on P for PRINT to see some more characters. This signal is not easy to trigger, you need some experience for it and a good hand at the trigger level. It's with 10us/div. As the characters appear only in the last line the brightness of the black level is not very good (about 4% of the other signal).

zx80core_video_verpos_s.jpg (110.64 KiB) Viewed 4843 times

Here we have same signal with 2 ms/div to see one full frame with a vertical sync at the beginning and at the end. You see the PRINT signals / black level maybe better as this is positioned in line 24, the last visible screen line (start with line 1 after the vertical sync pulse of about 400 us).

[iturbez]

Very good news! The Zeddy has come to life again!

RAM was faulty. I've installed a 62256 chip and now I get a brilliant inverted "K" at the screen...

I've connected an external Memotech keyboard, and I can see the little thing is alive!

Another one saved from the rubbish. The Zeddy and me are happy now!

Thanks for your help!
Greetings.

[1024MAK]

Mark

[gozzo]

Woo-hoo!! ps. was the original RAM 2x2114's or 1x4118(or 2016 - my original ZX81 used a 2016 2k chip but only half connected/used - maybe part duff?!) Never come across any RAM failure (yet..!)

[iturbez]

Mark

Thanks, Mark.

Woo-hoo!! ps. was the original RAM 2x2114's or 1x4118(or 2016 - my original ZX81 used a 2016 2k chip but only half connected/used - maybe part duff?!) Never come across any RAM failure (yet..!)

It was 1x4118. And it seems to be dead, as CPU was...

[TMAOne]

Congratulations! I love hearing Zeddy rescue stories.

I adore those 62256 mods. Did you figure it out yourself or find & use the ZX-Team mod page?

I like having the 2nd 16K as a swap area for flashing a copy of the primary 16K in case of a crash. The reset button I've added does not wipe out the 48 to 64 K area, so if I've fouled up I just load a stub program with a machine code copy routine, and "restore" the 16 to 32 K area from the high 16K. Works great.

Not to mention there's no piggyback RAM module to consume gobs of energy, wobble and crash, or fail outright.

Ian

[iturbez]

Did you figure it out yourself or find & use the ZX-Team mod page?

I used ZX-Team mod, in its simplest form. Only 16K available. but no wobble at all...