Hi all. I have this Heathkit AR-15 here, and it has had this symptom: One channel takes a good fifteen seconds to come fully alive. It moves from zero sound through distorted sound to full-on sound. But then at intervals it gets distorted again.
I think it has finally decided to die, though, because it will not come fully up out of distortion-land anymore.
It has this massive output capacitor, a 4000 mfd giant cylinder. Well, it HAD one; the previous owner (I am a recent Ebay buyer of this much-treasured receiver. At least by me) replaced it with a nice new one, I think Nakamichi? I can look up the brand but I've just put it all back together and don't feel like taking it back apart just now. The point is that it's been replaced so I am inclined to disregard it as the culprit. But I'm in over my head; I'm just an amateur; I do not know how to determine which component (because I'm not sure it's a capacitor, except that this problem tends to be due to a dying capacitor, from what I've heard).
I know it's advisable to do a complete re-cap, or at least it is for some people. For the moment, though, I'd like to isolate the culprit, replace it, and get back to listening.
I'm posting the schematic for the power amplifier section. I ran a test to make sure the problem is in the power amp (took the pre-amp output and ran it through another amplifier. It sounds OK).
Could somebody take a look at the circuit and tell me, in their opinion, which capacitor (assuming it is a capacitor) would be the one you'd suspect of having died?
Or what else could the problem be?
I'd really appreciate anybody further along in this hobby than me - I'm just a dilettante.
I've circled the electrolytics in the circuit on the schematic.
Thanks for any help.
Bob H.
I think it has finally decided to die, though, because it will not come fully up out of distortion-land anymore.
It has this massive output capacitor, a 4000 mfd giant cylinder. Well, it HAD one; the previous owner (I am a recent Ebay buyer of this much-treasured receiver. At least by me) replaced it with a nice new one, I think Nakamichi? I can look up the brand but I've just put it all back together and don't feel like taking it back apart just now. The point is that it's been replaced so I am inclined to disregard it as the culprit. But I'm in over my head; I'm just an amateur; I do not know how to determine which component (because I'm not sure it's a capacitor, except that this problem tends to be due to a dying capacitor, from what I've heard).
I know it's advisable to do a complete re-cap, or at least it is for some people. For the moment, though, I'd like to isolate the culprit, replace it, and get back to listening.
I'm posting the schematic for the power amplifier section. I ran a test to make sure the problem is in the power amp (took the pre-amp output and ran it through another amplifier. It sounds OK).
Could somebody take a look at the circuit and tell me, in their opinion, which capacitor (assuming it is a capacitor) would be the one you'd suspect of having died?
Or what else could the problem be?
I'd really appreciate anybody further along in this hobby than me - I'm just a dilettante.
I've circled the electrolytics in the circuit on the schematic.
Thanks for any help.
Bob H.
Attachments
I would begin by measuring the DC voltage on the two main stages of the amp and seeing how they compare between faulty state and working. Do they vary as the fault alters.
So Q202 Collector and Pin 10 of those right hand numbers.
So Q202 Collector and Pin 10 of those right hand numbers.
OK. To be clear, I should measure voltages at the C of Q 202, and ... By "Pin 10 of those right-hand numbers" you mean the speaker output line prior to the coil/capacitor junction, right? My only confusion is that it's not precisely a "pin"; it's just a point in the circuit. But I think I get your meaning. It seems like a logical place to measure voltage. Also: the manual does specify the correct voltages, so I have that to compare with.
I'll dig in and get back to you. Thanks for the suggestion!
- Bob
I'll dig in and get back to you. Thanks for the suggestion!
- Bob
You have to measure the output line before the big speaker coupling cap. We are not looking for small differences, we are looking for big change in voltage over time as the amp moves from no sound to distorted sound etc.
The voltage on the speaker coupling cap should be approx half the supply voltage give or take a couple of volts but it should be steady.
The voltage on the speaker coupling cap should be approx half the supply voltage give or take a couple of volts but it should be steady.
It would be good though to know which one exactly is failing. If I replace them all, and it fixes the problem, I won't know which one was the really bad one. Is there one that seems more likely to have given out/died than the others? Just from the point of view of theory I wonder which capacitor you would look at first. I suppose I could replace one at a time and see which one fixes the problem, though.
OK. I'll report back. Thanks for your help.
OK. I'll report back. Thanks for your help.
The way to fault find is always by methodical measurement wherever possible. Changing parts in hope is often a recipe for disaster for the unwary as there are so many possibilities to introduce new issues.
The fact it is distorted suggests in the first instance that the DC conditions are changing. It may or may not be a cap.
If the two checks show a problem they will also show which half of the amp the issue is.
The fact it is distorted suggests in the first instance that the DC conditions are changing. It may or may not be a cap.
If the two checks show a problem they will also show which half of the amp the issue is.
My idea is what I usually do at job: change some few caps, say, 4, 5, and test the amp for changes in its behavior. If something goes wrong, it is in the lasts 4 or 5. Also use a current limited power supply during tests. It can be done as follows:
https://www.diyaudio.com/forums/power-supplies/348381-current-limited-outlet-socket.html#post6051747
https://www.diyaudio.com/forums/power-supplies/348381-current-limited-outlet-socket.html#post6051747
Thanks for the idea. A propósito, viví en Buenos Aires por 3 años... Un lugar inolvidable y tan diferente de México, donde vivo actualmente (although I am a North American by birth). Fui unas tres o cuatro veces a Burzaco a visitar a una persona allí... Tomaba el tren desde Constitución.
Anyway, thanks again!
Bob
Anyway, thanks again!
Bob
As Mooly said , and especially for newbies :
first measuring , twice if needed , and what ever happens never change two parts at the time , unless you like confusion 😉
.
first measuring , twice if needed , and what ever happens never change two parts at the time , unless you like confusion 😉
.
Nobody has given you the basics of repair:
Make measurements with only one hand. 25 v across your heart can stop it. Connect negative of DVM to speaker ground or another convenient ground, with alligator clip leads. Put your other hand in your pocket if you are impulsive.
Wear no jewelry on hands or neck, 1 v @ 30A through a ring can burn your flesh to charcoal.
Wear safety glasses when desoldering or trying out with power. Parts can explode, solder can splash to your eye.
Other possible cause of distortion volume changes, oxidized headphone switch.
All (but one) your electrolytic caps are suspicious, as all are the same age built out of the same *****y rubber seal. If you replace one, your amp will be broken most of the time as they fail one by one.
But don't replace everything at once. Replace a part or two, then listen whether better or worse. If worse, you did something wrong, and you know with 50% accuracy where the new problem is.
Voltage on rail cap C508 is pretty easy to reach, it should be huge. this should be 75% of voltage rating & stable. Not fading in & out with the volume.
Other likely culprits, C202 & C208.
As the box from a real distributor costs $10-12 freight, it is best to load it up with parts. I found when buying e-caps from a TV parts store, they lasted 7-10 years. The 3000 to 10000 service life rated caps from newark/digikey/mouser are going on 12 years no problems, all 300 of them. newark(farnell) and digikey show your the service life in the selector table. Mouser makes you download the datasheet for each cap. Arrow is reputable but I don't know anything about them. RS of britain may operate there, also reputable.
You can't buy these old values anymore. Caps are stocked in 10,12,22,33,47,56,62,68,75,82 series only. Pick something close, these are not timing circuits so exact value isn't critical. Screw terminal caps are rediculously expensive, make a little board and glue snap on caps to it for values over 2200 uf. Glue or screw the board to the case.
Axial lead caps can't be bought in useful life spec. Only 1000 hours or less. Convert to radial lead caps or do it again in 5-10 years.
You can test electrolytic caps with a ESR meter like the Peak for $120. Most of the time you have to remove them to get a good reading. Once you've removed it, why on earth would you put it back? to save $.33? It is of the same quality as a 40 year old tire, rubber that has been attacked by oxygen for all those years. Plus removing & replacing stuff can damage lands on PCB's. Do it once & forget it. Use old caps for experiments maybe, not inside an amp you have to take the case off of.
The pro repairman replace ONE part if possible, because that makes their repair cheaper than anybody else's, and keeps the customer coming back. Over and over again as the caps fail one at a time until the customer chucks the device & buys a new one. I don't want the cheapest repair, I want the repair that keeps me out of the device for 15 more years.
Happy hunting.
Make measurements with only one hand. 25 v across your heart can stop it. Connect negative of DVM to speaker ground or another convenient ground, with alligator clip leads. Put your other hand in your pocket if you are impulsive.
Wear no jewelry on hands or neck, 1 v @ 30A through a ring can burn your flesh to charcoal.
Wear safety glasses when desoldering or trying out with power. Parts can explode, solder can splash to your eye.
Other possible cause of distortion volume changes, oxidized headphone switch.
All (but one) your electrolytic caps are suspicious, as all are the same age built out of the same *****y rubber seal. If you replace one, your amp will be broken most of the time as they fail one by one.
But don't replace everything at once. Replace a part or two, then listen whether better or worse. If worse, you did something wrong, and you know with 50% accuracy where the new problem is.
Voltage on rail cap C508 is pretty easy to reach, it should be huge. this should be 75% of voltage rating & stable. Not fading in & out with the volume.
Other likely culprits, C202 & C208.
As the box from a real distributor costs $10-12 freight, it is best to load it up with parts. I found when buying e-caps from a TV parts store, they lasted 7-10 years. The 3000 to 10000 service life rated caps from newark/digikey/mouser are going on 12 years no problems, all 300 of them. newark(farnell) and digikey show your the service life in the selector table. Mouser makes you download the datasheet for each cap. Arrow is reputable but I don't know anything about them. RS of britain may operate there, also reputable.
You can't buy these old values anymore. Caps are stocked in 10,12,22,33,47,56,62,68,75,82 series only. Pick something close, these are not timing circuits so exact value isn't critical. Screw terminal caps are rediculously expensive, make a little board and glue snap on caps to it for values over 2200 uf. Glue or screw the board to the case.
Axial lead caps can't be bought in useful life spec. Only 1000 hours or less. Convert to radial lead caps or do it again in 5-10 years.
You can test electrolytic caps with a ESR meter like the Peak for $120. Most of the time you have to remove them to get a good reading. Once you've removed it, why on earth would you put it back? to save $.33? It is of the same quality as a 40 year old tire, rubber that has been attacked by oxygen for all those years. Plus removing & replacing stuff can damage lands on PCB's. Do it once & forget it. Use old caps for experiments maybe, not inside an amp you have to take the case off of.
The pro repairman replace ONE part if possible, because that makes their repair cheaper than anybody else's, and keeps the customer coming back. Over and over again as the caps fail one at a time until the customer chucks the device & buys a new one. I don't want the cheapest repair, I want the repair that keeps me out of the device for 15 more years.
Happy hunting.
Last edited:
I agree 🙂
And tbh (sorry indianajo) but I take exception to this:
The 'pro repair man' ensures that he finds the exact cause of the problem without diving in at random and replacing parts ad-hoc. As little as possible is disturbed while at the same time detailed measurements are made.
Fault finding is all about gathering evidence wherever possible.
Replacing a bunch of caps might even get it working but that might not be because any of the parts randomly replaced were faulty... and the pro can see why that situation is possible.
Only when a definite diagnosis and then hopefully a repair has been made is it wise to then look at what other restoration work may be needed and inform a customer accordingly.