Low output from voltage regulator

[nugat]

Hello all!

I'm currently having my first experience with a ZX81, which isn't going too well unfortunately but I'm determined to get it working, any advice would be much appreciated

I've tried tuning in two different CRT televisions to the UHF modulator, and while there is something going on, it's just a mess of interference and the vertical sync is not holding. The board has the newer ULA2C210E fitted (it's in a socket fortunately), but I don't have a spare one to try.

I've checked the output from the power supply is 9v exactly when under load, but the output of the original 7805 voltage regulator was 4v. Which is too low right?

Last night I fitted a new 7805 voltage regulator, and while the output has gone up, it's still only 4.3v. This seems strange as I understand it should be much closer to 5v?

I've tried re-seating the chips that are in sockets but this hasn't made any difference.

The next thing I was going to try is tapping onto pin 16 of the ULA and see what's going on there. But my main concern is the low voltage output of the 7805. Is this normal, or is something funky going on?

Thanks in advance.

Image of main board :
https://drive.google.com/open?id=1XwgLc ... lqqbJyhco4

[Paul]

Welcome to our forums nugat.
How much current is it taking?
Kind regards Paul

[1024MAK]

Hello and welcome to our forum

4.3V is far too low for a 5V supply (which should at worst be 5V +/-5% meaning between 4.75V to 5.25V).

Are you using an original Sinclair power supply unit (PSU)?
If yes, which model / rating is it? If no, please supply details of the PSU including ratings.
Which multimeter do you have?
Do you have, or have access to (own/beg/borrow/steal...) any other test gear, such as an oscilloscope?

To work out what is going wrong, some tests are needed.
Can you please test the following and post up your results:

1. With the PSU disconnected from the ZX81, test the output with the meter switched to the 20V DC voltage range (or other suitable range). Expected value between 12V and 16.5V.
2. Same as above, but switched to the 2V AC voltage range (if the indicated voltage is low and you have mV ranges, switch down to a lower range). If you have an oscilloscope, use this to test for AC ripple.
3. With the PSU connected to the ZX81, test the output with the meter switched to the 20V DC voltage range. Expected voltage between 10V and 12.5V.
4. Same as above, but switched to the 2V AC voltage range. Again, if you have an oscilloscope, use this to test for AC ripple.
5. Measure the 5V rail at the output of the 7805 voltage regulator. Expected voltage between 4.75V and 5.25V.
6. Same as above, but switched to the 200mV AC voltage range if your meter has such a range. If not, use the 2V range. Same again if you have a ‘scope.
7. It’s normal for the metal heatsink connected to the 7805 to run hot. The ULA also runs hot (middle of chip), but you should be able to touch the rest of it and it should only be warm. The Z80 runs warm in the middle, the rest of it should be cool. The other chips (ROM, RAM) should be cool.
8. Can you measure the DC current in the 9V input, between the PSU and the ZX81, connect the meter in line (in series) with either the positive or the negative supply wire/connection. Use a DC current range of 500mA (0.5A) or greater. You will have to make up an adaptor cable for this test. Expected current is around 330mA DC.
9. With the PSU disconnected from the ZX81, power up the PSU for ONE minute. Then disconnect from the mains/power off. Leave both the input AND output disconnected. Test the output voltage with the meter switched to the 20V DC voltage range. Expected value between 12V and 16.5V. Disconnect the meter. Leave for 15 minutes. Repeat the voltage test.

The above tests are designed to find out how healthy the PSU is and to work out if the ZX81 is drawing more current than it should do.

Mark

[nugat]

Wow, thanks for the warm welcome guys and the detailed test schedule! It's going to take me a day or two find the time to run through that, but to start with....

Yes the PSU is an original Sinclair model 'UK700'. It's rated for 9v at 0.7A. When it's not under load it's outputting ~14.3v, when connected to the ZX81 it's outputting 9.0v. I haven't tested how much current the ZX81 is drawing.

My meter is nothing fancy (Draper automotive), but it has served me well for over 20 years I do have a small digital oscilloscope and I also have access to some more expensive scopes and test equipment (courtesy of my father).

But again, thanks for confirming that the 7805 output is too low. I'll work through the tests as soon as I'm able and post up the results!
Cheers, Tim.

[nugat]

Hi

I had another chance to take a look at the ZX81 yesterday and I'm pleased to say it's now working (almost).

Although the original Sinclair PSU appeared to be fine (14v disconnected and 9v under load with 400mA power draw), we connected the ZX81 to a bench supply to double check, and what do you know it starts working!

It turned out that the 1000μF smoothing capacitor in the Sinclair power supply was providing a less than adequate 100μF. We replaced it with a new capacitor and now when connected to the ZX81, the output of the voltage regulator has gone up to the expected 5v. Fantastic!

I now have a different problem in that five of the keys are not working (on the bottom right of the keyboard from 'B' to 'Space'), but I assume this is down to the dreaded ribbon cable. It hasn't cracked as far as I can see but I didn't have time to investigate further.

Cheers, Tim.

[1024MAK]

Good to see that you have solved the power problem

There are four common faults with Sinclair 9V ZX PSUs:

• A break in the output cable conductor (often intermittent)
• Live expired electrolytic smoothing capacitor
• Faulty rectifier diode
• The thermal fuse has blown

The keyboard fault is most likely to be the membrane.

As the membrane is just a matrix of conductors and contacts, you can simulate it with a piece of wire (thin solid core wire with about 5mm of the insulation stripped off the ends, such as fixed telephone or ethernet cable).

Momentarily connect it between one of the contacts in the eight way keyboard connector and one of the contacts in the five way keyboard connector. If you try every combination, you should be able to simulate every key (although shift on it’s own will have no visible effect).

Mark

[nugat]

As the membrane is just a matrix of conductors and contacts, you can simulate it with a piece of wire (thin solid core wire with about 5mm of the insulation stripped off the ends, such as fixed telephone or ethernet cable).

Momentarily connect it between one of the contacts in the eight way keyboard connector and one of the contacts in the five way keyboard connector. If you try every combination, you should be able to simulate every key (although shift on it’s own will have no visible effect).

That's really useful information, I'll try that next. Thanks again!

[nugat]

Final update (I hope) Looking at the board schematic, we noticed that all of the non-functional keys were connected via D8. Looking at the ribbon cable at this position there was clearly a break. Fortunately as the break it was close to the socket I was able to use some scissors trim back the broken section and re-insert.

This little machine is now issue free (apart from the slight burning smell when it's switched on, but that's normal )