twistedpneumatic
Experienced Member
I'm recapping my analog board as a suggestion from someone at VCF. Should I just replace all the electrolytic with equivalent ones or are there any other types of caps that I should look out for and replace?
- VCF South West - June 14 - 16, Davidson-Gundy Alumni Center at University of Texas at Dallas
- VCF West - Aug 2 - 3, Computer History Museum, Mountain View, CA
- VCF Midwest - Sept 7 - 8 2024, Schaumburg, IL
- VCF SoCal - Mid February 2025, Location TBD, Southern CA
- VCF East - April 2025, Infoage Museum, Wall NJ
-
Please review our updated Terms and Rules here
Recapping Pet analog board?
- Thread starter twistedpneumatic
- Start date
Chuck(G)
25k Member
First order of business, electrolytics.
Then tantalums, if they're failing.
Then tantalums, if they're failing.
ngtwolf
Veteran Member
Just out of curiosity, what's the problem that prompted the suggestion of recapping? Might narrow down where to focus.
What was the actual fault? I've never "recapped" anything unless the actual capacitors are bad (high ESR, leakage, etc) OR they're the RIFA with cracked epoxy (or about to crack anyway).
Frank IZ8DWF
twistedpneumatic
Experienced Member
First of all, the machine doesn't work. It made a pattern of a small dot that would bob up and down slowly then a large raster flash. I showed it to someone at VCF and they said recap the analog board. I figured it was worth a shot. I don't have an ESR meter so I'll probably just recap everything I see.
spaghetticode
Member
Massive recaps usually do not pay out IMHO.
I think I was never able to fix anything by doing recaps. I tried a few times, then I realized it was wasted time & money: unless you know a cap is failing, you're basically replacing good old caps with good new caps.
That being said, if you can't make a proper failure diagnosis and if you're confident you won't add new issues (damage PCB tracks, solder caps with inverted polarity) then it may be worth a try... still it's a longshot.
I think I was never able to fix anything by doing recaps. I tried a few times, then I realized it was wasted time & money: unless you know a cap is failing, you're basically replacing good old caps with good new caps.
That being said, if you can't make a proper failure diagnosis and if you're confident you won't add new issues (damage PCB tracks, solder caps with inverted polarity) then it may be worth a try... still it's a longshot.
Dwight Elvey
Veteran Member
This is my thoughts as well. I didn't want to keep saying the same thing over and over. It is true that tantalum caps are a common failure point. Electrolytics, used as active components in circuits and not as bypass, are a common source of failure. Electrolytics that fail as bypass usually have visible failures.
Still, unless the person is an experienced repair person with excellent rework soldering skills, the chance of doing damage with a soldering iron are greater than one might think. Unsoldering takes different skills than soldering components in.
There are different parts of the circuit. Replacing parts that have nothing to do with the failure are surly a waste of time and money. It is like your car radio is acting up so you replace the gas cap, the spark plugs and the thermostat because these are common failure points on a car but have nothing to do with the radio.
There is always someone that has had success with blanket replacement.
I agree with removing old RIFA caps. They will make a smelly failure. They do show visual potential failure if you look at them for cracks in their cases. Still, I doubt one of these is the cause of your failure. Flashes are often caused by dust or cigarette smoke deposits in the high voltage circuits. It can also be caused by a bad picture tube, with cathode to filament arcing.
Dwight
Chuck(G)
25k Member
My view on this is that the OP (TP) is bent on replacing electrolytics no matter what anyone says and wanted advice on other capacitors.
So if he was going to do it anyway, what's the point in saying anything other than "yes, it's true that the bulk of capacitor failures involve electrolytics". I think I've seen a grand total of 2 ceramic discs fail, for instance, in all my years. of dinking with this stuff. I don't think I've ever seen an oil or silvered mica cap fail--ever.
Electrolytics in power circuits that involve self-heating (e.g. filter caps, motor start/run caps, etc.) do fail quite often.
So if he was going to do it anyway, what's the point in saying anything other than "yes, it's true that the bulk of capacitor failures involve electrolytics". I think I've seen a grand total of 2 ceramic discs fail, for instance, in all my years. of dinking with this stuff. I don't think I've ever seen an oil or silvered mica cap fail--ever.
Electrolytics in power circuits that involve self-heating (e.g. filter caps, motor start/run caps, etc.) do fail quite often.
Dwight Elvey
Veteran Member
Exposed silvered mica capacitors fail quite often. The example is the capacitors used in IF transformers made in the 50's to 60's. The silver forms oxide ( or maybe sulfate ) that shorts sections of these caps. I've cleaned them on several radios to bring them back to life.
The only ceramic caps that I've had issues with are the glass encapsulated ones used in military equipment. Still the ceramic disk with arc slots cut in them can fail as a result of dust and cigarette smoke deposits. Other than that, the only ceramics I've seen fail were mechanically damaged.
Tantalums and electrolytics used in timing circuits can have issues. Most places that have high capacitive AC currents, as Chuck mentions, are common failures location ( LCD screen power supplies come to mind ). Old oil/paper capacitors are often bad ( not usually found in modern computers ) are know to fail.
Tantulums used as power supply filters often fail when first powered up after having been stored for some time, although, I've had one fail on a disk drive that had been on continuously for more than a year. Since these short out power rails and stay shorted, they are not likely to be the cause of the OPs problem.
If you think a bypass electrolytic capacitor has a high ESR failure problem, it is easier to tack a capacitor in parallel with the capacitor to see if the failure goes away than to replace it first. If beefing it up with a paralleled one makes the unit function, replacing that capacitor makes sense. Do note, this only works for ESR issues with bypass capacitors, not shorted, coupling or timing capacitors.
Dwight
The only ceramic caps that I've had issues with are the glass encapsulated ones used in military equipment. Still the ceramic disk with arc slots cut in them can fail as a result of dust and cigarette smoke deposits. Other than that, the only ceramics I've seen fail were mechanically damaged.
Tantalums and electrolytics used in timing circuits can have issues. Most places that have high capacitive AC currents, as Chuck mentions, are common failures location ( LCD screen power supplies come to mind ). Old oil/paper capacitors are often bad ( not usually found in modern computers ) are know to fail.
Tantulums used as power supply filters often fail when first powered up after having been stored for some time, although, I've had one fail on a disk drive that had been on continuously for more than a year. Since these short out power rails and stay shorted, they are not likely to be the cause of the OPs problem.
If you think a bypass electrolytic capacitor has a high ESR failure problem, it is easier to tack a capacitor in parallel with the capacitor to see if the failure goes away than to replace it first. If beefing it up with a paralleled one makes the unit function, replacing that capacitor makes sense. Do note, this only works for ESR issues with bypass capacitors, not shorted, coupling or timing capacitors.
Dwight
Chuck(G)
25k Member
The most common electrolytic failure mode that I encounter is with drying out due to heating. It's not at all unusual to measure a bad filter cap and find that it's little more than an empty can. Shorts do occur, but less frequently than the former case. Ether case can usually be tested for without removing the old cap.
If you need a bit of equipment to test components and are on a tight budget, I recommend the LCR-T1 for less than $20 shipped. The probes are garbage, but the unit itself is remarkably accurate and will test a wide variety of components. There are less-expensive versions of the same unit, since the MCU source is open, it's more a matter of cosmetics.
I keep one on my bench for quick tests.
If you need a bit of equipment to test components and are on a tight budget, I recommend the LCR-T1 for less than $20 shipped. The probes are garbage, but the unit itself is remarkably accurate and will test a wide variety of components. There are less-expensive versions of the same unit, since the MCU source is open, it's more a matter of cosmetics.
I keep one on my bench for quick tests.
Last edited:
ngtwolf
Veteran Member
Oddly, I agree with everyone in this thread. I personally wouldn't replace the caps because I don't think it'd replace the fault, and even if it did, it was likely one bad cap, not the entire board. I do have a C= Pet though and i've reflowed the solder joints on most of the analog board and the connectors on the logic board to clean-up the aging solder joints Unless you're really bad at desoldering, the likelyhood of lifting traces or ruining the board are slimmer than they are with some other systems (like a c64 for instance), so i don't personally think you'd risk wrecking much by swapping the capacitors out with this particular system.
I have personally fixed an Atari ST (the built-in one on the STFM) by recapping its power supply. The issue with that system was completely random crashing that could never be attributed to any one thing, though it was more common during disk access. In that case, I did test the capacitors I removed and while none of them were completely bad, most of them were off and a number of them had a higher than expected ESR. So, that said, it could be the analog board in this situation and while I think recapping the entire thing might (using an example similar to dwights) be like overhauling the entire engine instead of replacing just the oil or a gasket, in theory both would accomplish the same goal. Of course, in the car example, it could entirely be where the issue is in the fuel line and not the engine at all... but hey, you do have a new engine in the end... and in this case, you have a nice new recapped analog board, either way.
I have a version of one of these also, it was about $12 from aliexpress but didn't have the probes. I'm not sure how accurate it is, but i agree that it's very handy as it tests ESR and transistors, which my meter doesn't.
I have personally fixed an Atari ST (the built-in one on the STFM) by recapping its power supply. The issue with that system was completely random crashing that could never be attributed to any one thing, though it was more common during disk access. In that case, I did test the capacitors I removed and while none of them were completely bad, most of them were off and a number of them had a higher than expected ESR. So, that said, it could be the analog board in this situation and while I think recapping the entire thing might (using an example similar to dwights) be like overhauling the entire engine instead of replacing just the oil or a gasket, in theory both would accomplish the same goal. Of course, in the car example, it could entirely be where the issue is in the fuel line and not the engine at all... but hey, you do have a new engine in the end... and in this case, you have a nice new recapped analog board, either way.
I have a version of one of these also, it was about $12 from aliexpress but didn't have the probes. I'm not sure how accurate it is, but i agree that it's very handy as it tests ESR and transistors, which my meter doesn't.
Chuck(G)
25k Member
The probes are nothing much; just some cheap "grabbers" attached to short length of ribbon cable, terminated in male pins. My things-to-do-list includes adding some real probe jacks to the thing.
I've been surprised once or twice when testing a "house numbered" device and finding out that what it really was, as opposed to what I thought it was. (e.g. JFET vs. BJT or even SCR)
I've been surprised once or twice when testing a "house numbered" device and finding out that what it really was, as opposed to what I thought it was. (e.g. JFET vs. BJT or even SCR)
ngtwolf
Veteran Member
Interesting. I never thought about the probes since the numbers on the socket seem to account for most pitches, and i can usually just bend them a bit to test if needed. I guess the only hard ones are the really small leads left on the caps... Maybe i'll make up a set of probes as well. The other handy tester i have is the TL866 eprom programmer. Ignoring the main 'eprom programming' use, it also tests most logic chips which is useful.
Chuck(G)
25k Member
I had some trimpots that I wanted to pre-adjust before I soldered them onto the PCB. The leads on those are too short to clear the ZIF socket ridges, so I did have a use for the probes. Also, consider the need to test semiconductors in TO-3 or TO-208 packages.