Hi,
The Siliconix J555 is a current regulator diode (actually a J-FET selected for that current rating in a two lead package). It should regulate the current to about 2 mA that remains pretty constant with fluctuating voltage across it. These break down at around 50 volts (this would be a bad thing).
The LM329 is a 7.0 volt, low noise reference. Something like a zener diode in how you use it, but it is compensated for temperature changes. This uses a subsurface zener diode, plus other diodes in series to make it more stable with temperature changes. The fact that it's a subsurface zener diode means that it is also very low noise. Much less noise than a normal zener diode, and far more stable with changing temperature. Don't just use a normal zener diode to replace these.
-Chris
The Siliconix J555 is a current regulator diode (actually a J-FET selected for that current rating in a two lead package). It should regulate the current to about 2 mA that remains pretty constant with fluctuating voltage across it. These break down at around 50 volts (this would be a bad thing).
The LM329 is a 7.0 volt, low noise reference. Something like a zener diode in how you use it, but it is compensated for temperature changes. This uses a subsurface zener diode, plus other diodes in series to make it more stable with temperature changes. The fact that it's a subsurface zener diode means that it is also very low noise. Much less noise than a normal zener diode, and far more stable with changing temperature. Don't just use a normal zener diode to replace these.
-Chris
Yeah, having been all over that circuit, it is clearly designed for a zener reference, and not a current reference. And it works great with the LM329 part.
Scott
Scott
Progress on 565 #2:
I cleaned up the control board, replaced the caps and one of the zeners (the one with the corroded lead), and tried out the amp.
It was inoperative.
A little sleuthing, and I found that the fusible resistors in the +/-14 volt regulator circuit were off. One of the 27K resistors was measuring about 35K, and the 10 ohm resistor that "grounds" the center if that circuit was open. The result was the negative rail was at -21 volts the #3 "ground" was at -7 volts, and so on. Once I replaced these resistors, the low voltage rails went right back to +/- 14 volts. Of course, this mistreatment had also cooked the op amp. Once that was in, the amp worked.
However, I am still fighting an offset that seems to come in somewhere around the driver stages. The Op Amp output is at about -8 volts, which tends to offset the bias on all of the downstream circuits. The result is that the amp works great up to about 50 watts. Above that I get a transformer hum, and the signal starts clipping in one direction (because it is running with this offset).
I am going to replace the top diodes in the bias stack (D103 and 104), which I did on my other 565, and see if that helps. I think the issue with these amps is that they designed the bias of the drivers very close to the edge, so if one side is offset relative to the other it basically biases off. When this happens the output bias circuit bleeds the signal across to the opposite output side (through R139), so the amp works, but not very well (because it is not really running in Class AB). With the drivers biased on a little higher, the outputs combine properly and the operation is AB. At that point the servo can do its job, and the class A drivers don't get railed or cut-off. I used 1.5 volt LEDs for this. The specs on the diodes are 1.2 volts, but the ones in the circuit measured out at .68 volts, and so did the replacements.
I'll post the results later today.
I cleaned up the control board, replaced the caps and one of the zeners (the one with the corroded lead), and tried out the amp.
It was inoperative.
A little sleuthing, and I found that the fusible resistors in the +/-14 volt regulator circuit were off. One of the 27K resistors was measuring about 35K, and the 10 ohm resistor that "grounds" the center if that circuit was open. The result was the negative rail was at -21 volts the #3 "ground" was at -7 volts, and so on. Once I replaced these resistors, the low voltage rails went right back to +/- 14 volts. Of course, this mistreatment had also cooked the op amp. Once that was in, the amp worked.
However, I am still fighting an offset that seems to come in somewhere around the driver stages. The Op Amp output is at about -8 volts, which tends to offset the bias on all of the downstream circuits. The result is that the amp works great up to about 50 watts. Above that I get a transformer hum, and the signal starts clipping in one direction (because it is running with this offset).
I am going to replace the top diodes in the bias stack (D103 and 104), which I did on my other 565, and see if that helps. I think the issue with these amps is that they designed the bias of the drivers very close to the edge, so if one side is offset relative to the other it basically biases off. When this happens the output bias circuit bleeds the signal across to the opposite output side (through R139), so the amp works, but not very well (because it is not really running in Class AB). With the drivers biased on a little higher, the outputs combine properly and the operation is AB. At that point the servo can do its job, and the class A drivers don't get railed or cut-off. I used 1.5 volt LEDs for this. The specs on the diodes are 1.2 volts, but the ones in the circuit measured out at .68 volts, and so did the replacements.
I'll post the results later today.
Last edited:
Hi Scott,
You might still have an electrolyte problem there. With heavy offset correction, you may find that the bias current is low as well. I don't think you have a problem in the output stage as that normally causes excessive current flow - or turns everything off. It seems more like an offset problem.
-Chris
You might still have an electrolyte problem there. With heavy offset correction, you may find that the bias current is low as well. I don't think you have a problem in the output stage as that normally causes excessive current flow - or turns everything off. It seems more like an offset problem.
-Chris
Yeah, I'll give the diodes a try, and if that doesn't work, I'll try cleaning some more. This one definitely had electrolyte issues...even the screws holding the board to the frame were corroded!
Thanks!
Scott
Thanks!
Scott
Hi Scott,
Yeah, I'm sorry I had to rain on your parade. Once you get that stuff right into the PCB, it can be extremely difficult to get rid of it.
-Chris
Yeah, I'm sorry I had to rain on your parade. Once you get that stuff right into the PCB, it can be extremely difficult to get rid of it.
-Chris
Yeah, the amp is really not stable yet. Changing the diodes did clean up the output, but I noticed that the output sort of drifts around, and goes through periodic "glitches" (not sure what else to call them....). where it drifts off zero, and then sort of jumps around on the scope,and then settles back down. It's funny that the circuits in the areas that were most affected by the electrolyte (the servo and the low voltage regulators) seem to be performing perfectly +/- 14.00 volts on the nose..
It is very difficult to troubleshoot in this mode since with the feedback loop connected, it "looks" like it is working properly, but, not quite (feedback covers up a lot of ills!!). As soon as I disconnect the feedback line (and reduce the input level by 10:1), the output goes bonkers, with a weird square wave (even at millivolt input signal levels). My other amp, in this mode has nice clean sine waves, so something is clearly amiss in this amp, and it is not the same issue that the previous one had.
I did notice that even now, when I solder things on the board, I still get the funky black residue, and a bit of the electrolyte smell.
So, it is cleaner, but not yet clean.. 🙂... What did you say you used to clean yours?
I also laid out a new control board using Eagle schematic capture. I did it with 4 layers in order to get a good ground and power plane, so I have a top signal layer, internal +V and -V layers, and a bottom ground plane with a few signals. It was easier to do it this way with the autorouter than to try to copy the board art. I laid out the parts in more or less the same arrangement.
I am thinking I may double check the schematic, and if it is accurate, I'll get the board made. I have all the parts, so then I can try swapping it in and see if the new one works better. Seems like an easier fix than farting around trying to get the old one working with this invisible gunk soaked into 1/3 of the board and components.
Meanwhile I have a line on yet another ailing 565, so that will complete my 3 amp setup.
Cheers,
Scott
It is very difficult to troubleshoot in this mode since with the feedback loop connected, it "looks" like it is working properly, but, not quite (feedback covers up a lot of ills!!). As soon as I disconnect the feedback line (and reduce the input level by 10:1), the output goes bonkers, with a weird square wave (even at millivolt input signal levels). My other amp, in this mode has nice clean sine waves, so something is clearly amiss in this amp, and it is not the same issue that the previous one had.
I did notice that even now, when I solder things on the board, I still get the funky black residue, and a bit of the electrolyte smell.
So, it is cleaner, but not yet clean.. 🙂... What did you say you used to clean yours?
I also laid out a new control board using Eagle schematic capture. I did it with 4 layers in order to get a good ground and power plane, so I have a top signal layer, internal +V and -V layers, and a bottom ground plane with a few signals. It was easier to do it this way with the autorouter than to try to copy the board art. I laid out the parts in more or less the same arrangement.
I am thinking I may double check the schematic, and if it is accurate, I'll get the board made. I have all the parts, so then I can try swapping it in and see if the new one works better. Seems like an easier fix than farting around trying to get the old one working with this invisible gunk soaked into 1/3 of the board and components.
Meanwhile I have a line on yet another ailing 565, so that will complete my 3 amp setup.
Cheers,
Scott
Hi Scott,
Sorry, your board is exhibiting the classic signs of electrolyte poisoning.
I used a product called "Simple Green". I don't know if it was that effective. You know what you might try? Hot dish washing fluid in a sink, soak that darned board really well (hours). IPA might help after that. Then dry the board out in an oven at the lowest temperature setting. Obviously you need to remove capacitors and pots that you want to keep. Pull your new parts as well. This shouldn't hurt your resistors and sealed film caps.
Caution with your new PCB! If you're going to do a new PCB, copy the old one exactly! Trace layout can make a design great, or an abysmal failure. Without knowing how many iterations they went through and why, I wouldn't suggest changing anything. So do as accurate a copy as you can. It will then perform similarly, look the same and agree with the existing documentation. A few years down the line, you may be very happy you stayed with the original layout. Otherwise you might regret your decision to go your own way instantly on power up.
Layout is critical. Look for the Symasym threads as they went through the development and how much difference the PCB layout made while using the exact same circuit.
-Chris
Sorry, your board is exhibiting the classic signs of electrolyte poisoning.
I used a product called "Simple Green". I don't know if it was that effective. You know what you might try? Hot dish washing fluid in a sink, soak that darned board really well (hours). IPA might help after that. Then dry the board out in an oven at the lowest temperature setting. Obviously you need to remove capacitors and pots that you want to keep. Pull your new parts as well. This shouldn't hurt your resistors and sealed film caps.
Caution with your new PCB! If you're going to do a new PCB, copy the old one exactly! Trace layout can make a design great, or an abysmal failure. Without knowing how many iterations they went through and why, I wouldn't suggest changing anything. So do as accurate a copy as you can. It will then perform similarly, look the same and agree with the existing documentation. A few years down the line, you may be very happy you stayed with the original layout. Otherwise you might regret your decision to go your own way instantly on power up.
Layout is critical. Look for the Symasym threads as they went through the development and how much difference the PCB layout made while using the exact same circuit.
-Chris
Good to se you got it working again. Some "repair men" seem to think they are building legos, cause they´ll put in what ever compnent that fits. And if it doesn´t, there is always glue and copious amounts of solder.
Since you have cleaned and resoldered the board, it really doesn´t matter any more... but am I seeing a dab of stray solder shorting two legs on the transistor, or is it a trace underneath?
The picture of the cleaned and unpopulated board was to low res for me to make out the traces.
Since you have cleaned and resoldered the board, it really doesn´t matter any more... but am I seeing a dab of stray solder shorting two legs on the transistor, or is it a trace underneath?
The picture of the cleaned and unpopulated board was to low res for me to make out the traces.
Hmmm. Interesting. This one is working fine now, so I probably caught that at some point. Good eye!
That's why you should always clean off the old flux and whatever. Gives you ca chance to see defects like that. It was easy to see right off. But then, it is something I often see in units that are "unrepairable".
-Chris
-Chris
I dont know the name for it, not even in norwegian, but it looks like a dental pick. After soldering, I scrape and make sure there aren´t any strands of copper, dabs of solder or other material that might make a short and destroy even more or in best case scenario, leave me scratching my head.
I have no education other than middle school, so I tend to be overly cautious when it comes to these things. Im still trying to get a grip on all this, but it´s by far the gratest challenge I have faced when it comes to learning things. (except from female logic, ofc)
Messing with tube amps and SMPS is not something I should be doing at all, but I have found out that a lot of things are possible if i use my brain cell and read up on the subject first. Not to mention the exercise of caution. Especially when it comes to matters which I dont fully understand.
I hate the feeling of getting electrocuted, but I hate the smell of a four figure amplifier going up in smoke even more.
I cant find my keys or the remote if its right in front of me ... strange how we tend to miss the simple things the more we learn.
I have no education other than middle school, so I tend to be overly cautious when it comes to these things. Im still trying to get a grip on all this, but it´s by far the gratest challenge I have faced when it comes to learning things. (except from female logic, ofc)
Messing with tube amps and SMPS is not something I should be doing at all, but I have found out that a lot of things are possible if i use my brain cell and read up on the subject first. Not to mention the exercise of caution. Especially when it comes to matters which I dont fully understand.
I hate the feeling of getting electrocuted, but I hate the smell of a four figure amplifier going up in smoke even more.
I cant find my keys or the remote if its right in front of me ... strange how we tend to miss the simple things the more we learn.
Hi Forman313,
Some people have education, but no common sense. The reverse is also true. Remember that. Darwin favors the later group, that's all I'm going to say.
-Chris
How can someone be overcautious around high voltage?
Some people have education, but no common sense. The reverse is also true. Remember that. Darwin favors the later group, that's all I'm going to say.
-Chris
Yeah, I have gotten a jolt here and there after 50 years of working on audio gear. The worst was when I was a kid and accidentally touched the secondary wires of a tube amp transformer.. YEOW!!! 450 volts..burned a small hole in my finger.. Never forgot that.
Chris;
I laid out the new board with the components all in the same basic locations as the original. I also placed "polygons" around the edges, where the original has large area traces for +V, -V, Ground, and the low voltage ground. The auto router I have only supports top and bottom routing, so I ended up with a 2 sided board that is somewhat reminiscent of the original, just not as sinuous (that board was probably laid out by hand back in the day). The more angular approach should not matter at audio frequencies. The upside of that is that there are no jumper wires. I fattened up traces in the higher current sections of the board.
Here is the current layout:
I think I'll give it a try.. The board is relatively cheap, and I have all the parts other than the big heat sinks and 3 pin headers.
I'll keep you all posted.
Scott
Chris;
I laid out the new board with the components all in the same basic locations as the original. I also placed "polygons" around the edges, where the original has large area traces for +V, -V, Ground, and the low voltage ground. The auto router I have only supports top and bottom routing, so I ended up with a 2 sided board that is somewhat reminiscent of the original, just not as sinuous (that board was probably laid out by hand back in the day). The more angular approach should not matter at audio frequencies. The upside of that is that there are no jumper wires. I fattened up traces in the higher current sections of the board.
Here is the current layout:
I think I'll give it a try.. The board is relatively cheap, and I have all the parts other than the big heat sinks and 3 pin headers.
I'll keep you all posted.
Scott
Hi Scott,
The new board looks good, but you never know until you actually put it into service. You're going to have to run the amp into different loads and phase angles while looking for bursts of oscillation. I am hoping for success.
You know what? If that board works well, you could easily sell it for others who have had the same problem. I would have bought a board right out of the gate with a couple of the more ugly ones I've fixed. That board would have been a life saver.
-Chris
The new board looks good, but you never know until you actually put it into service. You're going to have to run the amp into different loads and phase angles while looking for bursts of oscillation. I am hoping for success.
You know what? If that board works well, you could easily sell it for others who have had the same problem. I would have bought a board right out of the gate with a couple of the more ugly ones I've fixed. That board would have been a life saver.
-Chris
Yeah, I want to be sure I have it right, and as close to the original as reasonable. Getting the first few made is expensive, but I agree, once it is proven out, it is probably a cheaper solution that screwing around with cleaning invisible electrolyte off 25 year old board with half rotted components...only to then find that it still doesn't work! (ask me how I know this.. and I haven't done nearly as many of these as you have!).
I have gone through it again this morning, continuing to fatten up traces that carry higher current loads. I also redid the package definition for the big heat sink transistors, so they match the heatsinks that are currently available.
I'll probably get a couple fabled in the next month (for small runs the cost for a few is pretty much the same as the cost for one..). I'll let you know how it comes out!
As I noted, the layout is slightly different since Eagle only does 2 to 16 layer boards.. I guess few folks mess with one layer boards anymore, since 2 layers is solo much easier to route.
Despite its somewhat quirky interface, I have grown to like using Eagle for schematic capture and board layout. The manual is semi helpful, and it is widely enough used that there are lots of part libraries, and online hacks available.
I have gone through it again this morning, continuing to fatten up traces that carry higher current loads. I also redid the package definition for the big heat sink transistors, so they match the heatsinks that are currently available.
I'll probably get a couple fabled in the next month (for small runs the cost for a few is pretty much the same as the cost for one..). I'll let you know how it comes out!
As I noted, the layout is slightly different since Eagle only does 2 to 16 layer boards.. I guess few folks mess with one layer boards anymore, since 2 layers is solo much easier to route.
Despite its somewhat quirky interface, I have grown to like using Eagle for schematic capture and board layout. The manual is semi helpful, and it is widely enough used that there are lots of part libraries, and online hacks available.
Hi Scott,
Hey! I resemble that remark!
I will prefer a single sided PCB with a couple jumpers over double sided PCB any day. But, I'm old and used to laying these out by hand. I'm trying to use DipTrace, and it seems to be good. The biggest problem I have is with libraries and finding the darn footprints I need.
I am serious about you selling those boards to others if they turn out fine. Just think of how many people you could help out with those.
-Chris
Hey! I resemble that remark!
I will prefer a single sided PCB with a couple jumpers over double sided PCB any day. But, I'm old and used to laying these out by hand. I'm trying to use DipTrace, and it seems to be good. The biggest problem I have is with libraries and finding the darn footprints I need.
I am serious about you selling those boards to others if they turn out fine. Just think of how many people you could help out with those.
-Chris
Eagle has huge libraries, and has the facility for creating/editing your own. I did this, on the attached for the heatsink mounted TO126 transistors. I drew the outline of the heatsink with the proper placement of the solder pins, and then dropped in a TO126 footprint for the silkscreen and the pads, solder stop, etc.
I have found that the routers (they have 10 different ones) are pretty good, but still not as good as a human. I manually edited this one (below) and eliminated about 25% of the top metal traces, and re-organized the bottom metal a bit.
I also set up a layout for one that has a buried ground plane between the two layers. I suspect that's the better board, since sit will have less noise and less potential for oscillations. You will notice that the vents for the heatsinks are a series of small holes rather than one large hole. This is because the cheapest fab solution I found has a max hole diameter 1/4", but that doesn't really fit. The holes on the original are about 3/8", so I used five 1/8" holes instead. Interesting issues when you get down to the brass tacks of making a board! This setup has slightly less area than the original, so I may end up cutting between the holes. If I fab more boards, then I can specify milling, and do a nice big opening under the heat sinks.
I have found that the routers (they have 10 different ones) are pretty good, but still not as good as a human. I manually edited this one (below) and eliminated about 25% of the top metal traces, and re-organized the bottom metal a bit.
I also set up a layout for one that has a buried ground plane between the two layers. I suspect that's the better board, since sit will have less noise and less potential for oscillations. You will notice that the vents for the heatsinks are a series of small holes rather than one large hole. This is because the cheapest fab solution I found has a max hole diameter 1/4", but that doesn't really fit. The holes on the original are about 3/8", so I used five 1/8" holes instead. Interesting issues when you get down to the brass tacks of making a board! This setup has slightly less area than the original, so I may end up cutting between the holes. If I fab more boards, then I can specify milling, and do a nice big opening under the heat sinks.
Last edited: