I’ve done 2 computer science degrees, and you and Jan Beta have me learning more about computing and given me a thirst to learn more after 10 years of focusing on my job. I forgot why I got into ICT and computing but now I have a re-energised desire again. Thanks for this!
Really enjoying these videos. I'm currently trying to have a go at repairing a previously repaired unit so this is helping than starting blind! I have replaced the 12v with a separate PSU so the composite circuitry send to be firing up. Problem seems more fundamental as it's just a screen full of garbage.
I really like your videos. You have a polite demeanor, your videos are well-edited and cut, your fingernails are trim, and you don't veer off topic. Your skill level is way above mine, but if there's one piece of advice I could give you it would be not to use "Multicomp" capacitors from Farnell. I cringe everytime I see their triangular logo on the capacitors you use for replacement. On this model, I noticed you used some panasonic capacitors as well, so I commend you for that. "Multicomp" isn't a brand, it simply stands for "Multi Company", and is Farnell's in-house brand. They just outsource to the lowest bidder and have a custom sleeve put on, so they're usually bottom-of-the-barrel quality. Another thing I want to touch upon is whether you test capacitors properly. For example, in this video, you mentioned that you removed the old capacitors, eventhough they tested well. I'm sure you're aware that in order to properly examine whether a capacitor is up to spec, you need to test the impedance/ESR besides just testing the capacitance. Lastly, with regards to this model ZX Spectrum, I'm not sure whether you found the cause of the abnormal albeit slight excess power draw, but if I may offer a suggestion, it might have something to do with those crusty leaded ceramic capacitors. I realize the board is very fragile and it may very well be impossible to remove them without ripping up pads/tracks, but nonetheless, you should consider it. Thank you for all your fantastic videos.
I have a few different marked brands of electrolytic capacitors in my stock of new parts. I don't doubt that Multicomp ones are not the very best available, but neither have I any reason to think they are particularly poor. I've used them in all sorts of situations for many years and haven't yet noticed a new one that is faulty and am not aware of any having failed in service, though no doubt they eventually will, like most electrolytic capacitors do. I choose from the range available those with the highest life expectancy rating and voltage rating (provided the physical size is not too great) to increase the chances of a long life. Even manufacturers worshiped by some for creating high quality capacitors have had some bad batches and bad designs. I would give more careful thought to selecting a capacitor to use in a situation where it is highly stressed or its characteristics are critical to correct and safe operation, but I don't think any in the ZX Spectrum fall into those categories. Just for the sake of completeness I did a full set of tests on the capacitors I removed from the +2's voltage converter circuit: capacitance, ESR and leakage current. None of the results indicated any faults. The largest part of the increased current draw was due to the faulty line driver IC. However, this +2 does still draw a bit more current than the one I have been using as a comparison. It's unusual for ceramic disc capacitors to exhibit any significant leakage current, but of course it can happen. I do not plan to replace them, but I am about to subject this computer to an extended test period to see whether any further faults emerge.
Excellent video as usual. Looking forward to part 3 and how the external damage can be resolved. I'm very tempted to now attempt to fix the Grey in my loft. But don't worry. If its not a very trivial fix I'll be minded to seek your services.
I was more than half expecting to find the audio carrier set to where it would be on system G when I saw the Spanish menu. As you mentioned, this Spectrum underwent repair in Spain, so the Spanish ROM and adjustment for system G are not unexpected. A bit puzzled that there seems to be no heatsink (if that's what it is) on the ULA.
That was very unfortunate to have the Z80 short out on you mid-repair. It's possibly been stressed out by the intermittent power, but still fairly unusual for a CPU to short out, I normally get +5v shorts on 40xx CMOS or bypass caps
I changed the C23 value to 47 uF, C22 at 100 uF, and that should raise the +12 value, but R55 made it lower again by the feedback, increasing the -12V too.And i don't think any consumption in mA on the + or -12V should affect R79. The way i see it, some AC voltage drops on it, between the 9V rail and the AC rail resulted from the oscillator. I think it's around 3V AC
I don't think I have ever seen such a clear picture from my Grey before. It has always had some ghosting or color bleeding, even when it was just unboxed. It looks awesome now! Very instructive. What is still there to be fixed? If feels quite okay right now. :)
If a part gets worm, it means, it is working. If it gets hot, it îs burning. I think you should look for a short. If your rezistor, gets hot, then there îs the short somewhere.
Very enjoyable & informative as usual, thank you I'm interested by the engineering steps you've taken to adjust the voltages back/closer to spec: Would you put this necessity down to overall component value drift & their lifespan - or do you suspect the original build was perhaps never that exacting in the first place ? Warm regards
I didn't make any permanent changes to adjust the outputs of the voltage converter circuit. The -12 V output isn't regulated and I shouldn't think it'd cause a problem so long as it stays within +/- 2 V of -12 V. The voltage will probably rise a bit (towards 0 V) when the interfaces are driving an external device too. P.S. I've got no idea why TH-cam puts a line through some of the text above, that's not how I entered it!
What is the purpose of R79 (100 ohm), anyway? Mine is extremely hot and I've changed it with a 2W one (and even further considering replacing it with a 3W resistor). IC3 is only slightly warm.
Following a quick glance at the circuit diagram, I'm not absolutely sure why R79 was added to the circuit (it isn't present in earlier versions of the Sinclair voltage converter). Perhaps the waveform at T1's output has a slightly too large amplitude without R79. However, it shouldn't get hot (perhaps slightly warm, but not hot). There is probably a fault in the voltage converter circuit or something is overloading one of its outputs.
@@JoulesperCoulomb Do you think is it safe enough to simply remove it and watch what happens? Or there is too high a chance of exposing memory chips to overvoltages?
I recommend trying to find what is causing the excessive power dissipation in R79. Removing R79 won't cure the underlying fault (and might even cause other problems).
@@JoulesperCoulomb There is no obvious fault; the only chip that is getting hot is Z80 CPU. (and the ULA - a lot less) I have replaced IC3 with no apparent changes.
There obviously is a fault; R79 is getting very hot and it shouldn't. I can certainly believe that the fault that causes it to get very hot is not obvious. However, I recommend you try to find and correct that fault.
You mentioned originally that this spectrum was designed for spain, which explains the audio carrier difference: Spain used PAL B-G, rather than PAL-I, Pal B and G had the audio carrier at 5.5MHz not 6.0MHz :-)
It's a UK Spectrum, but was repaired (badly) in Spain. So the audio carrier frequency could well have been adjusted then or prior to that so it'd work in Spain. It was at about 5.3 MHz before I changed it.
Is it correct to say that -13.8 Volts is "higher" than -12 Volts? I've always called them relative to 0V, as a negative supply is still a supply in technical terms, so if the supply is "low", then it means the negative supply output is low (ie. closer to 0V). I thought that was pretty much the standard terminology, because positive and negative supplies are two sides of the same coin (so to speak)? Strangely, the true current flow actually goes from the negative rail to Ground / 0V though. :p www.mi.mun.ca/users/cchaulk/eltk1100/ivse/ivse.htm
There's no question that -13.8 volts is lower than -12 volts. I can see why it'd be tempting to change that definition in the situation I was investigating, but it'd lead to other equally confusing scenarios. It would be reasonable to say the absolute voltage of the former is higher than the latter. Perhaps it'd be less confusing, but more long winded, to always refer to the voltage between two points instead of relying on an implied reference. On the final point, the way I think of it is that current is an abstract concept of a net movement of charge and is always positive in the direction from a higher voltage to a lower one. Of course, when the current is due to a flow of electrons (as it usually is in electronic circuits), then the charge carriers (the electrons) are moving in the opposite direction to the current flow.
Good answer. :) Yep, I suppose it's more about the potential difference between a lower voltage and a higher one, but I've always used the convention that -13.8 Volts is "higher" than -12 Volts. I've heard that used a lot in other repair / engineering vids, but I guess either convention can be used. As you say, it all depends on the reference point of course, but if the negative voltage is being measured with reference to Ground / 0V, then I would generally go with the "greater potential difference is higher" method. Could just be that it's more common to hear of the "higher" voltage convention for negative voltages when it's on a supply rail maybe? Hmm. I've never really looked into it too deeply before tbh. lol
I’ve done 2 computer science degrees, and you and Jan Beta have me learning more about computing and given me a thirst to learn more after 10 years of focusing on my job. I forgot why I got into ICT and computing but now I have a re-energised desire again. Thanks for this!
If I have not said it already, your videos are exceptionally well presented!
Really enjoying these videos. I'm currently trying to have a go at repairing a previously repaired unit so this is helping than starting blind!
I have replaced the 12v with a separate PSU so the composite circuitry send to be firing up. Problem seems more fundamental as it's just a screen full of garbage.
I really like your videos. You have a polite demeanor, your videos are well-edited and cut, your fingernails are trim, and you don't veer off topic.
Your skill level is way above mine, but if there's one piece of advice I could give you it would be not to use "Multicomp" capacitors from Farnell. I cringe everytime I see their triangular logo on the capacitors you use for replacement. On this model, I noticed you used some panasonic capacitors as well, so I commend you for that. "Multicomp" isn't a brand, it simply stands for "Multi Company", and is Farnell's in-house brand. They just outsource to the lowest bidder and have a custom sleeve put on, so they're usually bottom-of-the-barrel quality.
Another thing I want to touch upon is whether you test capacitors properly. For example, in this video, you mentioned that you removed the old capacitors, eventhough they tested well. I'm sure you're aware that in order to properly examine whether a capacitor is up to spec, you need to test the impedance/ESR besides just testing the capacitance.
Lastly, with regards to this model ZX Spectrum, I'm not sure whether you found the cause of the abnormal albeit slight excess power draw, but if I may offer a suggestion, it might have something to do with those crusty leaded ceramic capacitors. I realize the board is very fragile and it may very well be impossible to remove them without ripping up pads/tracks, but nonetheless, you should consider it.
Thank you for all your fantastic videos.
I have a few different marked brands of electrolytic capacitors in my stock of new parts. I don't doubt that Multicomp ones are not the very best available, but neither have I any reason to think they are particularly poor. I've used them in all sorts of situations for many years and haven't yet noticed a new one that is faulty and am not aware of any having failed in service, though no doubt they eventually will, like most electrolytic capacitors do. I choose from the range available those with the highest life expectancy rating and voltage rating (provided the physical size is not too great) to increase the chances of a long life. Even manufacturers worshiped by some for creating high quality capacitors have had some bad batches and bad designs. I would give more careful thought to selecting a capacitor to use in a situation where it is highly stressed or its characteristics are critical to correct and safe operation, but I don't think any in the ZX Spectrum fall into those categories.
Just for the sake of completeness I did a full set of tests on the capacitors I removed from the +2's voltage converter circuit: capacitance, ESR and leakage current. None of the results indicated any faults.
The largest part of the increased current draw was due to the faulty line driver IC. However, this +2 does still draw a bit more current than the one I have been using as a comparison. It's unusual for ceramic disc capacitors to exhibit any significant leakage current, but of course it can happen. I do not plan to replace them, but I am about to subject this computer to an extended test period to see whether any further faults emerge.
Excellent video as usual. Looking forward to part 3 and how the external damage can be resolved. I'm very tempted to now attempt to fix the Grey in my loft. But don't worry. If its not a very trivial fix I'll be minded to seek your services.
I was more than half expecting to find the audio carrier set to where it would be on system G when I saw the Spanish menu. As you mentioned, this Spectrum underwent repair in Spain, so the Spanish ROM and adjustment for system G are not unexpected. A bit puzzled that there seems to be no heatsink (if that's what it is) on the ULA.
That was very unfortunate to have the Z80 short out on you mid-repair. It's possibly been stressed out by the intermittent power, but still fairly unusual for a CPU to short out, I normally get +5v shorts on 40xx CMOS or bypass caps
I changed the C23 value to 47 uF, C22 at 100 uF, and that should raise the +12 value, but R55 made it lower again by the feedback, increasing the -12V too.And i don't think any consumption in mA on the + or -12V should affect R79. The way i see it, some AC voltage drops on it, between the 9V rail and the AC rail resulted from the oscillator. I think it's around 3V AC
I don't think I have ever seen such a clear picture from my Grey before. It has always had some ghosting or color bleeding, even when it was just unboxed.
It looks awesome now!
Very instructive.
What is still there to be fixed? If feels quite okay right now. :)
If a part gets worm, it means, it is working.
If it gets hot, it îs burning.
I think you should look for a short.
If your rezistor, gets hot, then there îs the short somewhere.
Very enjoyable & informative as usual, thank you
I'm interested by the engineering steps you've taken to adjust the voltages back/closer to spec:
Would you put this necessity down to overall component value drift & their lifespan - or do you suspect the original build was perhaps never that exacting in the first place ?
Warm regards
I didn't make any permanent changes to adjust the outputs of the voltage converter circuit.
The -12 V output isn't regulated and I shouldn't think it'd cause a problem so long as it stays within +/- 2 V of -12 V. The voltage will probably rise a bit (towards 0 V) when the interfaces are driving an external device too.
P.S. I've got no idea why TH-cam puts a line through some of the text above, that's not how I entered it!
What is the purpose of R79 (100 ohm), anyway? Mine is extremely hot and I've changed it with a 2W one (and even further considering replacing it with a 3W resistor). IC3 is only slightly warm.
Following a quick glance at the circuit diagram, I'm not absolutely sure why R79 was added to the circuit (it isn't present in earlier versions of the Sinclair voltage converter). Perhaps the waveform at T1's output has a slightly too large amplitude without R79. However, it shouldn't get hot (perhaps slightly warm, but not hot). There is probably a fault in the voltage converter circuit or something is overloading one of its outputs.
@@JoulesperCoulomb Do you think is it safe enough to simply remove it and watch what happens? Or there is too high a chance of exposing memory chips to overvoltages?
I recommend trying to find what is causing the excessive power dissipation in R79. Removing R79 won't cure the underlying fault (and might even cause other problems).
@@JoulesperCoulomb There is no obvious fault; the only chip that is getting hot is Z80 CPU. (and the ULA - a lot less) I have replaced IC3 with no apparent changes.
There obviously is a fault; R79 is getting very hot and it shouldn't. I can certainly believe that the fault that causes it to get very hot is not obvious. However, I recommend you try to find and correct that fault.
You mentioned originally that this spectrum was designed for spain, which explains the audio carrier difference: Spain used PAL B-G, rather than PAL-I, Pal B and G had the audio carrier at 5.5MHz not 6.0MHz :-)
It's a UK Spectrum, but was repaired (badly) in Spain. So the audio carrier frequency could well have been adjusted then or prior to that so it'd work in Spain. It was at about 5.3 MHz before I changed it.
JoulesperCoulomb That would make sense too :-)
Is it correct to say that -13.8 Volts is "higher" than -12 Volts?
I've always called them relative to 0V, as a negative supply is still a supply in technical terms, so if the supply is "low", then it means the negative supply output is low (ie. closer to 0V).
I thought that was pretty much the standard terminology, because positive and negative supplies are two sides of the same coin (so to speak)?
Strangely, the true current flow actually goes from the negative rail to Ground / 0V though. :p
www.mi.mun.ca/users/cchaulk/eltk1100/ivse/ivse.htm
There's no question that -13.8 volts is lower than -12 volts. I can see why it'd be tempting to change that definition in the situation I was investigating, but it'd lead to other equally confusing scenarios. It would be reasonable to say the absolute voltage of the former is higher than the latter. Perhaps it'd be less confusing, but more long winded, to always refer to the voltage between two points instead of relying on an implied reference.
On the final point, the way I think of it is that current is an abstract concept of a net movement of charge and is always positive in the direction from a higher voltage to a lower one. Of course, when the current is due to a flow of electrons (as it usually is in electronic circuits), then the charge carriers (the electrons) are moving in the opposite direction to the current flow.
Good answer. :)
Yep, I suppose it's more about the potential difference between a lower voltage and a higher one, but I've always used the convention that -13.8 Volts is "higher" than -12 Volts.
I've heard that used a lot in other repair / engineering vids, but I guess either convention can be used.
As you say, it all depends on the reference point of course, but if the negative voltage is being measured with reference to Ground / 0V, then I would generally go with the "greater potential difference is higher" method.
Could just be that it's more common to hear of the "higher" voltage convention for negative voltages when it's on a supply rail maybe?
Hmm. I've never really looked into it too deeply before tbh. lol
Lol
You have that thing on life support.