48K ZX Spectrum Repair (No. 5)

[@discoHR]

"Error, found 4, expected 4" :) BTW, great videos, I'm learning a lot from them.

[@michaelhawthorne8696]

I really do like your videos, no messing about or talking about the weather, just getting in there and fault finding. There is one thing that makes me smile and that's your greeting, "A good of time of day to you", Your English is perfect but that's a strange thing to say. maybe " Good day to you" or "Hi there". But this isn't criticism just an observation hopefully taken in the right way. Keep the videos coming !!

[@JoulesperCoulomb]

In all tests the Spectrums were powered by a regulated bench power supply set to output 9 V DC. I have no reason to suspect the power supply was not working properly and if it wasn't, it only affected one particular Spectrum/ULA combination. P.S. KZbin is being strange again and not presenting the option to reply to certain comments.

[@michaelhawthorne8696]

Just watched this again and thought maybe it could have been a seating problem in the socket, maybe a dirty pin of the IC making seldom but intermittent connections, a touch of corrosion or pitting maybe. Can't fault your logical reasoning as you went through the circuit. good in-depth knowledge of this device., does amaze me there are still so many of these comps still around.

[@MrWudzi]

I love electronics repair videos and yours are up there with the best. Please keep it up :)

[@Beaps73]

Love you vids mate. Where did you get the ram check program from, also what brand millimetre do you use (the yellow one)

[@GadgetUK164]

It's great when problems like that manifest themselves lol. It's likely the difference between board revisions (track layout, capacitance) and something degrading in the ULA - perhaps its more susceptible to noise and having been installed on a different revision the track layout alone can make enough difference. You get similar wierd things in Atari ST's where the revision of the board combined with a particular chip revision can mean that a 'faulty' chip actually works in a different revision board. I wonder if its maybe a degraded internal pull up resistors or something - these things don't last forever. I would be tempted to measure from gnd to address pins of the ULA (out of circuit) and between vcc and the address pins, and compare to another working ULA - It's a long shot but sometimes a difference can be seen that way. Great work! Love your approach =)

[@GadgetUK164]

Sweet!!! Love the channel - keep the vids coming =)

[@TheBezumax]

Is it ok if my TR4 shows me 0,6 at one side and 0,9 at other side(leg)?

[@lawrencemanning]

Love all your videos. Very interesting and great presentation.

[@michaelhawthorne8696]

Did you check the PSU noise level or voltage level, things do seem to be a tad random, noise could do this.

[@TheBezumax]

I have black centre and white border and think problem is on upper RAM how to check it by multimetr? Should I replace all chips with TMS4164-15NL or just wich was broken?

[@jaycee1980]

I'd have probably put it down to some sort of marginal timing on the lower ram IC's.. Those 4116's are more trouble than they're worth quite frankly! I would have been willing to bet that if you had replaced the lower RAM with one of your replacement modules, the problem would have disappeared.

[@UTRG-UnderTheRain]

not sure if anyone is watching these comments anymore but does the spectrum actually use 9v anywhere? doesn't it have a 9v psu? It seems it uses 5v +/- 12v so why a 9v supply?

[@SkillTimO]

I have started looking into my spectrums power supply. It says 9v but outputs 13v. This is normal. They are made to cost. Having the regulator do the work in the spectrum keeps the psu cost down. If only they made single voltage chips in the 80s. Heck Intel are only now pushing their 12VO standard.

[@mykedindeal]

Best of thebest.Thank you

[@EmmittBrownBTTF1]

4164 memories were used because they were faulty in one of two banks, and were wired to use only the good bank.

[@crumplezone1]

Fantastico ! really enjoyed that and you definitely have no memory corruption faults at all maestro ! :)

[@AstAMoore]

Great job! (Love your videos.)

[@sciexp]

Hello, I wanted to ask something. Recently I bought second hand ZX spectrum, and I have found your videos. So, the ZX spectrum boots ok, it shows the Sinclair logo, but there is a fluctuation on the screen. There is a horizontal bar that moves slowly from top to bottom, it is kind of black or grey. I guess that it could be some fluctuation that it is not getting filtered or something. The test ram shows 65535, so it is ok. What can it be? Have you had this problem before? I hope it is nothing serious. Thank you all for your attention. Regards.

[@JoulesperCoulomb]

How is the computer connected to the display device? If you're using the original modulated RF output note that the modulator isn't of the best quality and it usually causes some extra interference on the picture, though a very good picture is possible from some. A poor quality or unreasonably long connecting lead or dirty connectors on the lead can cause problems too. Also note that the modulator's carrier frequency isn't especially stable and it's usually necessary to adjust the television's tuning from time to time to maintain a reasonable picture, especially while the Spectrum is warming up. If you've adapted the Spectrum to provide a composite video output, again a poor quality or poor condition connecting lead can cause problems. Whatever the video output type, make sure it is soundly connected inside the Spectrum. For example, it isn't too unusual for the connection to the output socket inside the modulator's metal can to fail. Does the Spectrum still have its original (or other old) electrolytic capacitors? If these are degraded and failing, all sorts of strange interference can appear on the picture. I recommend they be replaced for this reason and because failing capacitors are a very common cause of faults in ZX Spectrums and can lead to the lower RAM being damaged.

[@sciexp]

I am using the AV connection. I did the mod. I connected a capacitor to the TV connection and disconnected the 5v supply to the modulator.

[@sciexp]

So, I detected that the DC power plug socket is very sensitive to movements, and after a quick look it is apparent that it is not in a good state. I removed it this afternoon, and checked the image in the screen without the power plug socket. Now, the image is better, it is fixed, but a shadow bar still appears in the screen, although now it does not move. Another observation I made is that I have to set the power adapter at the 12 V position, not in the 9V one in order to boot the ZX spectrum. My guess is that maybe something in the power supply circuit is not working really properly. ¿Maybe the transistors?

[@JoulesperCoulomb]

Your discovery that the computer is sensitive to the input voltage in the way you describe makes me think it very likely the voltage converter circuit isn't working. Perform the tests described at kzbin.info/www/bejne/f6vKdKCimdJlfpY and in particular, check for the presence of all three power supply voltages. Look for 5 V, -5 V and 12 V at the appropriate power pins of the lower RAM ICs. There is a high probability of damaging the lower RAM if you power the computer on while the voltage converter circuit isn't working, so other than once more for checking the supply voltages, I recommend you keep the power off until you are reasonably sure the voltage converter circuit has been repaired.

[@sciexp]

Well... I did the tests and everything seems ok. When I plug it, there are one or two grey shaded horizontal bars that keep moving. I have changed already C44, C45, C34, and C50. I have an issue 3b.