Timex Sinclair 2068 - Troubleshooting

[Diegorand]

Need some help here, hope you can shed a light on this.

I received this TS 2068 from a friend. He told me that the original PSU was connected to 220V, so it blew.

I have some experience repairing many ZX and C64 systems, including soldering, socket replacements, recaps, etc., so I accepted the challenge. Upon inspecting the motherboard, there were no signs of burned pcb, bloated caps, or burnt resistors or diodes. I decided to power it on with a 19V switching PSU, negative center. EDIT: (its 16V actually).

The image displayed was distorted and unstable, with waving black and white vertical lines, so I immediately switched it off.

The next step was checking the voltages. I clamped the negative probe of the multimeter to the RF chassis and started checking with the positive probe, in DC mode.

At the LM1889N, I got 6.99V instead of the required 12V. The TMS4416 RAM voltages showed 11.47V at pin 9 ! The Z80 pin 11 also had 11.47V.

I then switched it off and thought that was the end due to the higher voltages (12V where 5V supposed to be).

Went to the store and replaced the following components: 78L12, 1N5821 diode, swapped D43C1 for a TIP42C (the one the store had), both C3 and C40, and tried again. Unfortunately, there was no change. Same voltage range in the RAM and Z80.

Today, I decided to give it another try. I removed the Z80, tested it on another machine, and it worked! So, fortunately, the Z80 is fine. I also removed the RAM chips and put new sockets in.

Before placing the new 4416 chips in, I turned on the machine to check the voltages again, and once more, 11.5V at pin 9 in the RAM sockets!

I am very confused. Maybe I am doing it wrong by testing the voltages with the multimeter's negative probe on the RF chassis? If so, how should I do it?

Any clues about how to properly check the UA78S40PC regulator? I haven't found a replacement for it locally. My original idea was to change every component involved in voltage regulation but this one wasn't available.

I am limited by the tools I have—just a multimeter and a component tester (TC1); I don't have an oscilloscope.

Any help will be appreciated. Thanks!

[jdfan1000]

The 2068 power supply was 15v; 19v is a bit more than it likes.

The 5v circuit in the 2068 is generated by the 78S40 and D43C1. I have never tried powering the 2068 with anything higher than the original 15v but I'll guess that putting 19v it will result in higher than 5v from that power supply.

You can replace the entire 5v and 12v sections with modern, tiny switcher power supplies like this: https://www.aliexpress.us/item/3256806299955878.html

David

[1024MAK]

U1 (µA78S40), Q1 (D432C1), CR1 (1N5821), the inductor L2 and associated components form a buck converter (step-down DC-to-DC converter) which decreases and regulates the output voltage to a set amount. It has the limitation that often a fault with it will result in nearly the full input voltage being passed through to the output. This may be happening here.

As this is a type of switch mode PSU, a ‘scope would be helpful...

Okay, so the first thing to say, is do not refit any removed chips until the +5V supply is fixed. If there are any remaining chips that are in sockets, remove these as well.

For fault finding. I suggest you feed the board with a 7V to 8V DC regulated supply. Or a regulated 9V supply via a couple of rectifier diodes. If you don’t have a regulated PSU, a smoothed DC u regulated type will do, as long as the actual output voltage is no more than 8V off load. Doing this will limit the stress on the remaining chips.

The following should be checked without the board being powered:

Locate resistors R5 (3320Ω 1%) and R6 (1100Ω 1%) and check that their okay. Keep in mind that if tested in circuit, the meter may show a lower value due to parallel paths. Also test that one end of R5 is connected to the +5V rail. And that the junction of these connects to U1 pin 10. R5 and R6 form a potential divider to sample the +5V (“output”) and feed to the error corrector amplifier in U1. Hence providing a feedback loop so that the output stays at the correct voltage. When the output is at +5V, pin 10 should be between 1.2V and 1.3V.

With the power still off, check R2, R4 and R9.

If all the above resistors are okay, move on to the next step.

Feed in the 7V to 8V supply, and test the voltages on the following pins of U1:
5, 8, 9, 10, 12, 13, 14, 15, and 16.

Next test the voltage on Q1 base, collector and emitter terminals.

Keep the board powered up for only as long as needed to carry out the tests.

Post your results here.

Do you have a regulated +5V PSU?

Mark

[Diegorand]

Thanks Mark & David for your help.
Mark, I really appreciate your detailed guide!! I've followed it step by step.
Found a regulated 5V PSU Hi link Hlk-10m05 5v 2a 10w, and fortunately a new UA78S40 with socket.
Powered the board with that 5v PSU, and without any memory on the sockets started probing U1 with the following results:

PIN 05 -> 4.9V
PIN 08 -> 1.2V
PIN 09 -> 1.2V
PIN 10 -> 0.39V
PIN 12 -> 0.73V
PIN 13 -> 4.9V
PIN 14 -> 4.9V
PIN 15 -> 1.6V
PIN 16 -> 1.6V

Q1 BASE -> 4.4V
Q1 COLLECTOR -> 1.7V
Q1 EMITTER -> 4.9V

What do you think of the values ? Should I put the chips on the sockets and try with a 15v regulated psu?
Thanks again for your valuable help!!!
Diegorand.

[1024MAK]

Hi.

When the board is being fed via a 5V power supply, those voltages are in the right ball park.

Now, before going for broke with a 15V PSU, it's best to try it with a 7V to 9V DC regulated supply.

Do the same as before. Connect it up and test the same points. But also test the +5V rail.

Mark