megajocke said:
Amplifier has soft start circuit currently , my original question was if i went ahead and increased power supply capacitance would the original soft start work or would i have to upgrade or add another..
Chris argument is/was that doing this would be a negative gain ...
Even if the extra capacitance wouldn't hurt there is probably very little to be gained if there already is enough capacitance.
OK, how much is enough , is 70,000 enough ? over the years this has not proven to be true for a high current amplifier operating at low impedances...
Okay, one at a time.
Hi unclejed613,
You handled a dicey concept very well. Let me restate what I think you are getting at a different way.
The power supply of any amplifier is in series with the output signal. The contribution of the impedance of that supply is swamped by other contributing factors in that series circuit. I'd like to add connection resistance in every connection. As the load impedance is reduced, all impedances become more important, however your power supply contribution is still swamped by the other circuit impedances.
Now for the nasty bit. We all agree we're talking about impedances. As the frequency goes up, larger power supply capacitors perform less well. This explains the rise in popularity of "audio grade" capacitors, and the excellent practice of using many smaller units in parallel to combat that effect. Bypassing these larger filter capacitors close to the output stage is a necessary step. Adcom does this BTW.
-Chris
Hi unclejed613,
You handled a dicey concept very well. Let me restate what I think you are getting at a different way.
The power supply of any amplifier is in series with the output signal. The contribution of the impedance of that supply is swamped by other contributing factors in that series circuit. I'd like to add connection resistance in every connection. As the load impedance is reduced, all impedances become more important, however your power supply contribution is still swamped by the other circuit impedances.
Now for the nasty bit. We all agree we're talking about impedances. As the frequency goes up, larger power supply capacitors perform less well. This explains the rise in popularity of "audio grade" capacitors, and the excellent practice of using many smaller units in parallel to combat that effect. Bypassing these larger filter capacitors close to the output stage is a necessary step. Adcom does this BTW.
-Chris
Hi a.wayne,
Although this amplifier is rated into a 2 ohm load, it will not perform at it's best into a 2 ohm load. At high power levels into 2 ohms, the quality of your AC supply will affect the performance of any amplifier much more that you may expect. My amplifier testing was done on it's own 15 A circuit in an industrial unit where the BX cable was over sized.
Most amplifiers may be rated for 4 ohms minimum, 2 ohms stable. This refers to the ability of the amplifier to remain stable at 4 ohms, even though the impedance may dip as low as 2 ohms at some frequencies. Many people mistakenly assume that amplifier can be used on 2 ohm loads. This is an incorrect assumption.
Finally, in my opinion, any speaker that is designed to present an impedance less than 4 ohms is irresponsible on the designer's part. Why build a speaker that causes most amplifiers out there to not work well? That makes zero sense at all. If you are running woofers in parallel to achieve this impedance in search of higher power levels, you are shooting yourself in the foot. You are further ahead to run each woofer on it's own channel. You loose maybe 3 dB in power (big deal), but you will gain more damping from the amp and run the voice coils at a lower temperature. BTW, you will lose around 2 dB due to hot voice coils.
-Chris
Edit: I just noticed this part ...
I have. This is a standard test for low impedance Carver units and also Nakamichi. This test typically trips the breaker and not the protection network in the amp under test.
Although this amplifier is rated into a 2 ohm load, it will not perform at it's best into a 2 ohm load. At high power levels into 2 ohms, the quality of your AC supply will affect the performance of any amplifier much more that you may expect. My amplifier testing was done on it's own 15 A circuit in an industrial unit where the BX cable was over sized.
Most amplifiers may be rated for 4 ohms minimum, 2 ohms stable. This refers to the ability of the amplifier to remain stable at 4 ohms, even though the impedance may dip as low as 2 ohms at some frequencies. Many people mistakenly assume that amplifier can be used on 2 ohm loads. This is an incorrect assumption.
Finally, in my opinion, any speaker that is designed to present an impedance less than 4 ohms is irresponsible on the designer's part. Why build a speaker that causes most amplifiers out there to not work well? That makes zero sense at all. If you are running woofers in parallel to achieve this impedance in search of higher power levels, you are shooting yourself in the foot. You are further ahead to run each woofer on it's own channel. You loose maybe 3 dB in power (big deal), but you will gain more damping from the amp and run the voice coils at a lower temperature. BTW, you will lose around 2 dB due to hot voice coils.
-Chris
Edit: I just noticed this part ...
This amplifier is not rated for 1 ohm impedances. Not many are. I used to have a Carver Lightstar. 1200 watts per channel into 2 ohms. It required two circuits. It was not rated for 1 ohm duty and I never ran it into anything less than 4 ohms nominal Off hand, I can't think of any amplifier that is rated into 1 ohm, except maybe a Krell. Even then I don't know if it is.
Hi megajocke,
This "more capacitance" thing is another one of those fads. A green marker is more likely to improve the bass than bigger capacitors.
-Chris
Edit: Sorry, I'm very tired.
Finally, the quality of an amplifier has nothing at all to do with it's ability to drive lower than 4 ohm loads. In fact, design comprises move away from high fidelity to handle low impedances. Either that or the price shoots up for very little gain in performance.
Like bridging an amplifier, running low impedance loads just dosn't buy you much. There are more than enough negative things involved with this practice to go around. It's your amp, just don't talk about it.
You are right. Adding more capacitance will bring you very little gains. It will cause trouble in the form of higher transformer core temperatures (thinking of low impedance operation also) and they have a thermal fuse buried in the primary windings. The higher pulses of current will create stronger magnetic fields that tend to interfere with the low level circuitry. The soft start will need to be modified as well.
This "more capacitance" thing is another one of those fads. A green marker is more likely to improve the bass than bigger capacitors.
-Chris
Edit: Sorry, I'm very tired.
Well, no kidding! These impedances are not reasonable. Did you know that the old standard impedance was 16 ohms? Transistor amps don't use a transformer normally, so the standard impedance was changed to 8 ohms. Now they are playing around between 6 ohms and 4 ohms.
Of course! These amps were not designed to run at these levels. You are shorting out the feedback network as you go lower in load impedance (not good for standard feedback type amplifiers). Even wire gages and trace thicknesses were not designed to handle this current. Then there are those pesky current surges that creep into the safety factor built into the design.
Finally, the quality of an amplifier has nothing at all to do with it's ability to drive lower than 4 ohm loads. In fact, design comprises move away from high fidelity to handle low impedances. Either that or the price shoots up for very little gain in performance.
Like bridging an amplifier, running low impedance loads just dosn't buy you much. There are more than enough negative things involved with this practice to go around. It's your amp, just don't talk about it.
anatech said:
Chris... 2 discoveries of mine that I have found oppsed to conventional wisdom (I have not found the logic or reason yet but it is observable)
1) I always found the sound of a balanced/ bridged amp to sound less grainy than the single ended version when it came to the 555ii's and 545ii's. This scenario did not entail running the amp into full power but anywhere from 10w to 200w levels...
2) The other thing I have found is that reducing the rails of the amp and bumping up the bias up a tad gave it a less harsh sonic signature...
It is no small wonder that at one time I was running converted 555ii's to run off 50v rails and then bridged to give me about 300w into 8 ohms.
I think it was an experiment that paid off for me in the past...
anatech said:
Hello Chris,
Appreciate the technical Input , Are you Julian Hirsch in disguise ? 🙂
1. You have now acknowledge that operating @ 2 ohm the power supply voltage and stability becomes important. ( stiff supply ) where an increase in power supply stiffness ( caps ) offsets the other negatives previously discussed. In audio as most other worldly things compromise is the name of the game.
2. Have you ever bench tested an amplifier that doubles ? ( constant voltage ) If so you need a 20 amp breaker to fully test 1 side to clipping. when operating in this manner you would see the difference between 50,000 and 150,000 uf .If testing on a fully resistive load , the reaction and stability is completely different to a simulated RLC load.....
3. Double up speakers to lower impedance gives an increase of 3 DB and an additional 3 db on amplifiers that double ( min 5 db gain) in reality ...a substantial difference in dynamics
4. Designing speakers below 4 ohm , while not practical is neither irresponsible as the same can be said about dinky toy amps that cannot handle below 8/4 ohm nominal. For the record , the Adcoms can drive a 2 ohm nominal load all day long, one of the few ,it is what it is ....horse's for courses
Regards,
Hi Arif,
I haven't tried testing Adcom 555 in bridged mode, but the human ear is fallible when it comes down to listening tests. Could be that the extra power is what you needed and was the biggest problem. One thing is for sure, the output impedance just doubled. That means that in your case the damping factor was not an issue.
In this case, it worked out well for you. The thing is ... you know what you are doing and most people do not.
Now, looking at the 555II service manual, I find that it is only rated down to a 4 ohm load, not 2 ohms. Also, as you know, when you bridge an 8 ohm speaker, it "looks" like a 4 ohm load to each channel.
Keep in mind that the 2 ohm capability was mostly a marketing thing. They didn't actually expect people to run 2 ohms. That actually implies that the customer is using more than one set of 4 ohm speakers on one amp. Not common or recommended. Also, keep in mind that some Adcom amps blew outputs due to driver transistor failure into low impedance loads. Then there is how much power you are expecting out of the amp. You did touch on that.
BTW, good to "see" you!
-Chris
I haven't tried testing Adcom 555 in bridged mode, but the human ear is fallible when it comes down to listening tests. Could be that the extra power is what you needed and was the biggest problem. One thing is for sure, the output impedance just doubled. That means that in your case the damping factor was not an issue.
In this case, it worked out well for you. The thing is ... you know what you are doing and most people do not.
Now, looking at the 555II service manual, I find that it is only rated down to a 4 ohm load, not 2 ohms. Also, as you know, when you bridge an 8 ohm speaker, it "looks" like a 4 ohm load to each channel.
Keep in mind that the 2 ohm capability was mostly a marketing thing. They didn't actually expect people to run 2 ohms. That actually implies that the customer is using more than one set of 4 ohm speakers on one amp. Not common or recommended. Also, keep in mind that some Adcom amps blew outputs due to driver transistor failure into low impedance loads. Then there is how much power you are expecting out of the amp. You did touch on that.
BTW, good to "see" you!
-Chris
Hi a.wayne,
Out ... the .... front ... panel. You know, where the meters used to be. They also had to [hand] file the openings larger. This ruined the resale value forever. What did the customer get for his money? A wrecked amplifier. The extra capacitance did nothing. Mind you, the Marantz 500 only stores about 2.2 joules of energy in its capacitors. Well, something along those lines anyway. We ended up reversing the modification and putting the meters back. Couldn't fix the ragged holes around top and bottom of those meters.
Building speakers that will stress most amplifiers is just silly. You lose market share and gain a reputation as a difficult speaker. The intelligent thing to do is build speakers that do not stress the amplifier. They will sound better and more dynamic. Remember Infinity Kappa 9? They didn't sound very good anyway, but they sounded worse on otherwise competent amplifiers. For example, I have a Marantz 300DC that sounds really good. It throws my PSB Stratus Golds (4 ohm) around easily. Just because it isn't designed to drive 2 ohm speakers does not detract from it's quality in any way. Same thing with my Cyrus Mono X amplifiers. Knowing that the standard low impedance is 4 ohms, and having the technology to easily design for 4, 8 or 16 ohms, it is not real smart to design for 2 ohms unless you are going for notoriety. I also believe that such a designer is deliberately short changing his customers.
Now for my question to you. Are you going for sound quality and SPL, or just high power? You see, power is not your friend. It heats up speaker components, drags down your AC power which may cause trouble with your signal sources. What you should be doing is going for efficiency. You really want to decrease the power required. Part of that is to allow the voice coils to run cooler. Last point. making speakers is also a lesson in compromise. Having to make the voice coil larger increases weight and reduces cooling (larger gap). Also, the spider and surround must become stiffer. These things make the driver less efficient and raises the resonant frequency.
That's much like buying a car because it's hard on gas. You tell me what is more intelligent.
-Chris
No. 😀 I'm just cranky.
In a way, yes. I also mentioned that the difference is very small and probably not noticeable. Your AC quality is even more important.
Can't argue with that at all. In that light, I'll share something that someone did that really made no sense at all. Into the shop one day came a nice little Marantz 500 power amp (only rated down to 4 ohms). The owner was talked into adding extra capacitance by another shop. So, leaving the original parts installed, they added two more large caps at great expense to that individual. Where did they put them you ask ??
Out ... the .... front ... panel. You know, where the meters used to be. They also had to [hand] file the openings larger. This ruined the resale value forever. What did the customer get for his money? A wrecked amplifier. The extra capacitance did nothing. Mind you, the Marantz 500 only stores about 2.2 joules of energy in its capacitors. Well, something along those lines anyway. We ended up reversing the modification and putting the meters back. Couldn't fix the ragged holes around top and bottom of those meters.
Yes I have, often. The Marantz 500 is an example of that, as is the Carver Lightstar I mentioned.
No, in both cases, you need the same constant power in. You will still trip the breaker. You may get a couple watts more with music before the breaker trips. Certainly nothing really worth adding that extra amount of capacitance.
Yes, but only with dynamic signals. Otherwise all you are going to do is heat up the output stage as the reactive part gets larger, and possibly bring the protection circuits into action. The extra capacitance does not come into play here. Not in a meaningful way.
In a perfect world - yes, 6 dB more. Life just isn't fair though. 4 dB is more realistic. All your losses get higher. The temperature of you voice coils and crossover will reduce that gain further. The speaker will no longer be tuned to it's box (the Qt will change). With almost twice the current, connection and wire losses may become significant - especially against your 2 ohm load.
A little, but you will also pick up more dynamic compression due to heat. Your damping factor is going to be much lower too.
Sorry, I have to disagree with you here. Most quality amplifiers will handle a 4 ohm load. The cheap chip amp receivers suffer from a small heat sink and power supply. These in particular would benefit from both more capacitance and a larger heatsink together. If not, the supply will not sag any more (used to protect the amplifier stage to some degree). However, we are talking about good equipment here.
Building speakers that will stress most amplifiers is just silly. You lose market share and gain a reputation as a difficult speaker. The intelligent thing to do is build speakers that do not stress the amplifier. They will sound better and more dynamic. Remember Infinity Kappa 9? They didn't sound very good anyway, but they sounded worse on otherwise competent amplifiers. For example, I have a Marantz 300DC that sounds really good. It throws my PSB Stratus Golds (4 ohm) around easily. Just because it isn't designed to drive 2 ohm speakers does not detract from it's quality in any way. Same thing with my Cyrus Mono X amplifiers. Knowing that the standard low impedance is 4 ohms, and having the technology to easily design for 4, 8 or 16 ohms, it is not real smart to design for 2 ohms unless you are going for notoriety. I also believe that such a designer is deliberately short changing his customers.
No they can't. Most of them are not rated for less than 4 ohms. Believe me, I not only checked one manual, but I was authorized service for Adcom until they were sold.
Now for my question to you. Are you going for sound quality and SPL, or just high power? You see, power is not your friend. It heats up speaker components, drags down your AC power which may cause trouble with your signal sources. What you should be doing is going for efficiency. You really want to decrease the power required. Part of that is to allow the voice coils to run cooler. Last point. making speakers is also a lesson in compromise. Having to make the voice coil larger increases weight and reduces cooling (larger gap). Also, the spider and surround must become stiffer. These things make the driver less efficient and raises the resonant frequency.
That's much like buying a car because it's hard on gas. You tell me what is more intelligent.
-Chris
New problem, different amp
Hello Chris and Arif, it has been quite a while! I'm glad to see that this thread is alive and thriving with lots of information and knowledge to be shared.
A gentleman approached me and asked for my help in repairing and rebuilding his GFA-565, it appears to be in impeccable condition and I did my thing, almost exactly the same way as I did my own.
I stripped, cleaned (with ultrasonic) and rebuilt the input board with brand new electrolytic capacitors, IC101 (OP97 op-amp), and variable resistor, replaced all other electrolytic caps in the amp in addition to the bridge rectifier) and then powered up.
Here is the problem:
At powerup, the DC voltage at the speaker terminals is 600mV and immediately starts climbing over the period of a few minutes to 1.7V!
The amp bias sits at 47mV when the variable resistor is set all the way down (when the variable resistor is set to the middle, i get a reading of about 60mV!!)
As soon as I switch off the amp, the DC voltage at the speaker terminals jumps to about 5VDC and climbs (yes, CLIMBS when the amp is powered off) to about 6VDC and then slowly comes back down.
Do you have any idea what is going on? Why is the amp biased so high when I have the resistor turned all the way down? Why is there so much DC offset at the speaker terminals? And why does the offset climb significantly on poweroff?
Thank you gentlemen once again for your amazing patient help!
Hello Chris and Arif, it has been quite a while! I'm glad to see that this thread is alive and thriving with lots of information and knowledge to be shared.
A gentleman approached me and asked for my help in repairing and rebuilding his GFA-565, it appears to be in impeccable condition and I did my thing, almost exactly the same way as I did my own.
I stripped, cleaned (with ultrasonic) and rebuilt the input board with brand new electrolytic capacitors, IC101 (OP97 op-amp), and variable resistor, replaced all other electrolytic caps in the amp in addition to the bridge rectifier) and then powered up.
Here is the problem:
At powerup, the DC voltage at the speaker terminals is 600mV and immediately starts climbing over the period of a few minutes to 1.7V!
The amp bias sits at 47mV when the variable resistor is set all the way down (when the variable resistor is set to the middle, i get a reading of about 60mV!!)
As soon as I switch off the amp, the DC voltage at the speaker terminals jumps to about 5VDC and climbs (yes, CLIMBS when the amp is powered off) to about 6VDC and then slowly comes back down.
Do you have any idea what is going on? Why is the amp biased so high when I have the resistor turned all the way down? Why is there so much DC offset at the speaker terminals? And why does the offset climb significantly on poweroff?
Thank you gentlemen once again for your amazing patient help!
Hi,
Did you remove the bias trimmers before you cleaned the board?
I guess the first order of business is to check all your work. Pretend that this amp just came in from another shop and do not assume anything at all. Especially when it comes to your own work, people generally do net see their own errors because they assume they already checked something, or they have completed a task when they haven't.
The board may not be totally clean. You need to remove the op amp and possibly other parts to clean the board underneath and also those parts. Make sure the supplies for the op amp are at the proper voltage. Beyond that, standard troubleshooting.
-Chris
Edit: Why are you replacing the bridge rectifiers? Arctic Silver may be electrically conductive. Often the main filter caps are fine too.
Did you remove the bias trimmers before you cleaned the board?
I guess the first order of business is to check all your work. Pretend that this amp just came in from another shop and do not assume anything at all. Especially when it comes to your own work, people generally do net see their own errors because they assume they already checked something, or they have completed a task when they haven't.
The board may not be totally clean. You need to remove the op amp and possibly other parts to clean the board underneath and also those parts. Make sure the supplies for the op amp are at the proper voltage. Beyond that, standard troubleshooting.
-Chris
Edit: Why are you replacing the bridge rectifiers? Arctic Silver may be electrically conductive. Often the main filter caps are fine too.
Also please check the other components for out of spec condition just in case this was an amp who someone tried to repair and has one resister in the wrong spec.
Also check the little diodes to be consistent with others.... these don't go out usually, but if they do can cause the 6v after power off...
Good luck.
Also check the little diodes to be consistent with others.... these don't go out usually, but if they do can cause the 6v after power off...
Good luck.
there are 2 things that can cause the problem you describe. the first and most common is the DC blocking cap in the feedback loop is leaky (has a dc resistance in parallel internally). if a diode is used externally to protect the cap from reverse bias, this could be leaky as well.
the second cause of this (not as often, but i have seen it happen) is one of the transistors in the diff amp is failing.
the second cause of this (not as often, but i have seen it happen) is one of the transistors in the diff amp is failing.
Hi unclejed613,
The time constant makes me think of the DC servo.
Hi Arif,
-Chris
Actually, this is very common when the amp goes DC. One of the pair is forced into reverse breakdown and that actually causes parameter shifts. The part doesn't normally fail outright, but the beta is normally changed. Noise generally increases also.
The time constant makes me think of the DC servo.
Hi Arif,
Oh, sad but true. This causes a lot of grief for the next tech.
Those are precision reference, low noise zeners. They need to be replaced with the same class of part. Normal zeners should not be used, although the use of a normal zener will get the amp back into operation with poorer S/N specs possibly.
-Chris
in my experience, the cap getting leaky causes the slowly changing offset more often than the diff amp itself, though i've seen diff amps do this when one of the pair gets thermally sensitive. if there's a servo in the amp it can behave in the same way if the caps are bad. i've seen the diff amp condition you describe, and noisy is usually an understatement, noisy diff amps tend to create loud pops and frying sounds in the output. i had mentioned in another thread that a NASA engineer found out back in the 60's that transistors begin to get noisy as they reach the end of their life, and NASA uses noise testing to determine when transistors should be replaced.
Hi unclejed613,
Don't forget too, the way transistors are made these days is different to the 60's parts. The early transistors were hermetically sealed with an inert gas. Some even had a getter material to absorb moisture. I have a book from GE around here that details exactly how they tested parts for Nasa and other military applications. They also used leakage currents as a more sensitive indicator for problems. This book was written in the late 50's or mid 60's as I recall. They were still trying to control their production processes back then.
-Chris
Yes, there is a DC servo in this amp.
Well, only when they get really bad. You can catch them before this point by looking a the leakage. A bigger clue in a diff pair is a shifting DC offset. How do you find this in a DC servo type amp you ask? Simple. Watch the output for the DC servo as the amp warms up. You don't need to disconnect anything to do this. A wire tacked to an easily accessed solder pad is the safe way to do this.
I had responded to that comment by saying the noise was due to leakage currents most of the time. You can measure that leakage while causing the part to warm up and cool down.
Don't forget too, the way transistors are made these days is different to the 60's parts. The early transistors were hermetically sealed with an inert gas. Some even had a getter material to absorb moisture. I have a book from GE around here that details exactly how they tested parts for Nasa and other military applications. They also used leakage currents as a more sensitive indicator for problems. This book was written in the late 50's or mid 60's as I recall. They were still trying to control their production processes back then.
-Chris
Thank you all for your very informative posts.
I have found the problem: R121 was malfunctioning.
R121 is a "Resistor, Fusible 5%, 1/4W/10ohms"
When I measured R121 it measured at about 100ohms and then started climbing in resistance and then jumped back down and started climbing again! Something very bad must have happened to it, because the finnicky readings I was getting from it translated directly into massive DC offset issues and too high of a bias.
I replaced the resisitor with the one off of my other amp, plugged in all the wires and it works perfectly. DC offset at the speaker terminals is 0mV and bias is steady at 24mV. Sounds great, the problems I described before are completely gone.
Anatech: I removed the bias trimmer, all electrolytic caps, op-amp IC, cemented wirewound transistor, and pre-driver heatsinks prior to cleansing. After cleansing and a thorough drying cycle I changed (with new components) the trimmer, electrolytic caps and op-amp. I also used a regular thermal compound on the pre-drivers transistors. If you remember from way back in the thread, on my own personal amps, the bridge rectifier failed and presumably fed AC to the filter caps, which had a devastating effect on one of them prior to my ownership of them. I figure since the bridge rectifier and electrolytic filter capacitors have operated in a very hot environment for 20+ years, it wouldn't hurt to replace them all with new units, who knows when they could fail. Isn't the life-span on electrolytic capacitors about 10 years? What about bridge rectifiers, do they fail often?
Arif: I'm glad none of the transistors or diodes were bad on the board, but if one of them was bad, how would you test them? On my DMM I have a tester where you can put the P N and P leads into.
unclejed613: I tested the film capacitors in addition to the DC blocking capacitor (C101) and they all have the right capacitance. I tested the resistance across C101's leads on my good, working amp, and then on the suspect amp and both were identical. It stabilized to 1.05 M-ohms after about 10 seconds.
I have a balanced input stage sitting in front of me for the GFA-565 and there is a lot of gunk all over them. I am fearful to put them back into the amp for fear it ruining all my hard work!
PICTURES!
Balanced Input Stage (top)
Balanced Input Stage (bottom)
I noticed in the service manual that the ICs on the balanced input stage are listed as part number "ADCOM 6A" Do you know what type of IC this is? There are two of them, they are in the 8-pin DIP form.
I have found the problem: R121 was malfunctioning.
R121 is a "Resistor, Fusible 5%, 1/4W/10ohms"
When I measured R121 it measured at about 100ohms and then started climbing in resistance and then jumped back down and started climbing again! Something very bad must have happened to it, because the finnicky readings I was getting from it translated directly into massive DC offset issues and too high of a bias.
I replaced the resisitor with the one off of my other amp, plugged in all the wires and it works perfectly. DC offset at the speaker terminals is 0mV and bias is steady at 24mV. Sounds great, the problems I described before are completely gone.
Anatech: I removed the bias trimmer, all electrolytic caps, op-amp IC, cemented wirewound transistor, and pre-driver heatsinks prior to cleansing. After cleansing and a thorough drying cycle I changed (with new components) the trimmer, electrolytic caps and op-amp. I also used a regular thermal compound on the pre-drivers transistors. If you remember from way back in the thread, on my own personal amps, the bridge rectifier failed and presumably fed AC to the filter caps, which had a devastating effect on one of them prior to my ownership of them. I figure since the bridge rectifier and electrolytic filter capacitors have operated in a very hot environment for 20+ years, it wouldn't hurt to replace them all with new units, who knows when they could fail. Isn't the life-span on electrolytic capacitors about 10 years? What about bridge rectifiers, do they fail often?
Arif: I'm glad none of the transistors or diodes were bad on the board, but if one of them was bad, how would you test them? On my DMM I have a tester where you can put the P N and P leads into.
unclejed613: I tested the film capacitors in addition to the DC blocking capacitor (C101) and they all have the right capacitance. I tested the resistance across C101's leads on my good, working amp, and then on the suspect amp and both were identical. It stabilized to 1.05 M-ohms after about 10 seconds.
I have a balanced input stage sitting in front of me for the GFA-565 and there is a lot of gunk all over them. I am fearful to put them back into the amp for fear it ruining all my hard work!
PICTURES!
Balanced Input Stage (top)
Balanced Input Stage (bottom)
I noticed in the service manual that the ICs on the balanced input stage are listed as part number "ADCOM 6A" Do you know what type of IC this is? There are two of them, they are in the 8-pin DIP form.
bridge rectifiers tend to work well for a long time unless the caps go bad, or something running off them shorts (causing it to overheat a junction and short) or gets hit by a large voltage spike (which is also part of the mechanism which damages the rectifier when the caps go bad). when i used to work on a lot of computer monitors, i could usually tell which power supply caps had dried out because the board would be discolored from the diode that fed that cap overheating. i've developed a "rule of thumb" over the years that if a cap is dried out, the semiconductors in the same circuit have likely been thermally stressed as a result, and if they show signs of having been stressed,(discoloration of the board, the solder flash on the leads or heatsink tab has been oxidizing or melting, or dried out the heatsink grease, or the collector lead pad on the board getting thermal damage) they get replaced.
Hi mjraudio,
That "gunk" could be solder flux. It's hard to tell. The op amps are probably fine.
If you are in doubt about this board, remove the semiconductors, pot and caps. Clean as you have done, then you can use methyl hydrate or lacquer thinner to clean all residue off. You need to remove any parts that could trap fluid or "goo" under them. Also clean the ICs and fixed parts. Remount the sensitive parts after cleaning them carefully with a toothbrush only moist with cleaner. Keep fluids from entering and adjustable parts. Test the board with a hot soldering iron for "that smell". Once you are certain the PCB is clean, remount your parts and clean off your solder flux.
Your open 10 ohm resistor is exactly what I normally test for first, before opening the amp up even. I've gone over that so many times now ....
The bias trimmers I mark location on and reuse them unless there is a problem with them. That way I know the bias should be close to what it needs to be. Bias pots do not go bad nearly as often as it's been suggested.
A failed bridge rectifier is rare in these amps - as unclejed613 mentioned. I would have left it unless the casing showed overheating.
The filter capacitors will last a long, long time unless they have been abused or over heated. I test them by visually inspecting the vent to see if it has ruptured (instant bad cap diagnosis). I then look at the ripple waveform. This is a very good indicator for the health of a filter capacitor, even beyond measuring it. If the leading edge of the charging waveform shows "pips", the capacitor is aging. Large "pips" mean the capacitor is nearing the end of it's useful life. Of course, both caps get replaced.
I'm glad you have regular thermal compound now.
Most half decent meters test semiconductors (diode test) by passing approx. 1 mA through the part and measuring the voltage across it. There are a few ways of measuring beta, that "P N and P leads" thing.
One thing I'd like to see you do.
If you are going to do any work on someone else's equipment, no matter the circumstances, I want to see you invest in some equipment. I also want you to become familiar with parts and electronics in general. So, if you are on the cheap, a used Fluke 87 meter, or a new one like the equivalent 77. The new ones have improved, that's why the drop in model number. You would be very wise to go for an 87 or the new equivalent due to all the additional measurements you can make with them.
I have a real problem with people working on things when they are not qualified to do that work. There is a reason why good technicians have so much experience and training behind them. Most guys have learned so much that the job is generally easy for them. Even they don't understand how much the average person doesn't know.
You need to know how each component works and the differences between component types. You also need to know how to test them, and have the equipment to do so - properly! Much of this stuff you can make yourself, but you have to learn. Otherwise you become more of a problem than a help. This has nothing to do with your intentions.
Working on your own stuff is one thing. Doing work on anyone else's equipment is an entirely different ballgame. You have much more responsibility, both morally and legally.
-Chris
That "gunk" could be solder flux. It's hard to tell. The op amps are probably fine.
If you are in doubt about this board, remove the semiconductors, pot and caps. Clean as you have done, then you can use methyl hydrate or lacquer thinner to clean all residue off. You need to remove any parts that could trap fluid or "goo" under them. Also clean the ICs and fixed parts. Remount the sensitive parts after cleaning them carefully with a toothbrush only moist with cleaner. Keep fluids from entering and adjustable parts. Test the board with a hot soldering iron for "that smell". Once you are certain the PCB is clean, remount your parts and clean off your solder flux.
Your open 10 ohm resistor is exactly what I normally test for first, before opening the amp up even. I've gone over that so many times now ....
The bias trimmers I mark location on and reuse them unless there is a problem with them. That way I know the bias should be close to what it needs to be. Bias pots do not go bad nearly as often as it's been suggested.
A failed bridge rectifier is rare in these amps - as unclejed613 mentioned. I would have left it unless the casing showed overheating.
The filter capacitors will last a long, long time unless they have been abused or over heated. I test them by visually inspecting the vent to see if it has ruptured (instant bad cap diagnosis). I then look at the ripple waveform. This is a very good indicator for the health of a filter capacitor, even beyond measuring it. If the leading edge of the charging waveform shows "pips", the capacitor is aging. Large "pips" mean the capacitor is nearing the end of it's useful life. Of course, both caps get replaced.
I'm glad you have regular thermal compound now.
You need to learn some more about components. You can build some jigs, see http://www.passdiy.com/articles.htm for some ideas on testing parts.
Most half decent meters test semiconductors (diode test) by passing approx. 1 mA through the part and measuring the voltage across it. There are a few ways of measuring beta, that "P N and P leads" thing.
One thing I'd like to see you do.
If you are going to do any work on someone else's equipment, no matter the circumstances, I want to see you invest in some equipment. I also want you to become familiar with parts and electronics in general. So, if you are on the cheap, a used Fluke 87 meter, or a new one like the equivalent 77. The new ones have improved, that's why the drop in model number. You would be very wise to go for an 87 or the new equivalent due to all the additional measurements you can make with them.
I have a real problem with people working on things when they are not qualified to do that work. There is a reason why good technicians have so much experience and training behind them. Most guys have learned so much that the job is generally easy for them. Even they don't understand how much the average person doesn't know.
You need to know how each component works and the differences between component types. You also need to know how to test them, and have the equipment to do so - properly! Much of this stuff you can make yourself, but you have to learn. Otherwise you become more of a problem than a help. This has nothing to do with your intentions.
Working on your own stuff is one thing. Doing work on anyone else's equipment is an entirely different ballgame. You have much more responsibility, both morally and legally.
-Chris
- Status
- Not open for further replies.