4:15 - Those are not technically emitter degeneration resistors. Emitter degeneration is negative feedback and acts on the transconductance (gm) of the transistor. It lowers the gain, improves input linearity, increases the bandwidth of the amplifier. The resistors in this circuit are intended to counter the imbalance in the class B amplifier and prevent huge current to flow between the NPN/PNP transistors.
Furthermore, the stabilization resistor doesn't reduce the damping factor of the amplifier, because the feedback resistor sits directly on the output terminal. So the output voltage is controlled at the output of the amplifier. The resistor between the emitter of the transistor and the output has no influence of the output resistance of the amplifier.
By far one of the best amp restoration channels, great job!, i would definitely send you my JVC-S71 amp to check it over if i lived in the us, too bad i live 10000km away.
Love the Channel....I worked for Crazy Eddies in 84-86...Of all the Audio gear I sold...The service department would send this back to SAE due to a Very High return rate. I never would recommend SAE to anyone.
Glad to see that you found a schematic. I know how much more work it is without one. More than just figuring out the issue, is understanding the design. Amps like this definitely teach you other options in designs.. I would like more amps like this in my shop just to investigate and learn more. Good job 👍
25 volt capacitors where 16 volt capacitors are might have been a main problem here and now fixed. Can't do much more here but check those large Schottky diodes. I wonder if replacing them with silicon carbide types would help? Yes I know they only come in a TO-220 case. Tony your showing a great video here even if you seemed to fail at the end. There is more investigating on this amp to do.
the diodes across the output emitter resistors are feedforward diodes its purpose is cancellation of distortion of large-signal non linearity inherent in the operation of output transistors, especially at heavy loads
Ive seen similar designs, in fact I have seen such amps that are usually class A and move to class AB/B during high power output by pulling the biasing on the driver stage. For example Marantz class A amps sometimes use this. I see them burn out regularly due to bad pots for bias adjustment, diodes who will randomly break down causing DC on the output stage, dried electrolytics etc
(@7:48 or so) - Well, that’s certainly interesting. And I don’t recall ever having seen a diode bypass an emitter resistor before. If your theory is correct, then the Schottky diode acts like a .2 volt zener, holding the emitter of Q17 at a constant voltage, which may help, as it looks like what the transistors are in, is called a cascode arrangement, with Q17 acting as an emitter follower and the signal going through R45 & R43 to the base of Q15, and through R51 & R49 to the base of Q16. But, it seems to me, that if you have a sustained high output (lots of bass or percussion), with the volume turned up, then the diode acts more like a penny in a fuse box, preventing R39 from protecting the transistors, which may explain why they blew. Just my thoughts. 😏
As it turns out, it's not a good idea! I am now anti-schottky, when it comes to the emitter resistor circuit! I will get into this more when I post part 2.
That whole circuit with Cascoded output transistors is very strange to me. I get it that they wanted to lower the Miller effect, possibly because the out transistors are slow with high Cbo. Anyway, complicated piece of vintage piece where they try to overcome one issues that creates other problems.
I don't like those experimental designs, like fooling around with shottky diodes in an output stage. A diode is a switching device with a certain delay and capacity and it never is perfectly linear. In this case, it is a recipe for unnecessary complications, which may lead to serious tail chasing.
Tony, I see some here think this was a failure, what I see is you have revived the AMP to where further in depth diagnostics is required to see what other issues arises. Like you have mentioned cleaning the pots and possibly making the final adjustments. Since the specifications on the amp has somewhat been changed, It is rather interesting the document mentions do not plug and unplug other components while the unit is on. I wonder if the users manual says that in BIG BOLD RED Letters. MOST people do not read the users manual, I wonder if this amp went POOOF early on when someone unplug and plugged in another component and it needed to be serviced.
Respect the work! Not sure if you familiar with the SAE 2600. Original filter caps were 9800 uf 120V DC. Do you have an opinion on what would be a good or great replacement? Thx!
If I remember correctly 4 of the 6 output transistors were darlington, the other 2 were regular bipolar. It was a long time ago though and I might be wrong.
I don't know, but I would be tempted to put a simple RC passive LPF on the input with a -3db point at 80-100khz. I really don't see any significant audio quality advantage for a 200khz input bandwidth. No music recording is going to have anything that far up or even close. The only thing there besides RFI would be bias leakage from a vintage analog tape deck in record monitoring. That would probably destroy this amplifier. And note the amplifier circuit would still have the 200khz equivalent slew rate. The LPF would just ensure the amp never sees that coming in.
I agree. See Bob Cordell's book: Designing Audio Power Amplifiers (page 87, section 4.2). A poor design from SAE not to have such an input filter. I bet many of these amplifiers failed because of the need to power off before connecting a source.
It is not about reproducing 200k. The advantage of 200k are things like reproducing fast transients and complex music passages. The 200k allows the amp to trace the complex input signal more faithfully.
@@jked7463 That's correct, but here is also a practical limit like 80-100K in this case. Also music or any natural sound cannot have fast rise times requiring even close to a 200khz bandwidth due to our 15psi atmosphere. This does not mean square wave testing of audio components is not needed. It is to evaluate circuit stability. But in the application of audio components, there will never be a transient that requires a 200khz bandwidth. Now good test equipment must have 5x the measured bandwidth, but that is mainly needed outside of audio electronics.
Hi Tony, love your videos. I have seen you add a single power resistor/relay soft start, powered by B+ rail...I am wondering why you do not instead use NTC’s, which are designed specifically to handle the extreme initial peak power of the limiter? How do you know your chosen power resistor can handle that peak power? Also, should the relay coil circuit be protected with its own low value fuse?
its very interesting the way they made the power supply in this one it be cool if you could explain on more the way two transformers work together i know u said thier not dual separate power spplys im curious how much supply voltage dose each channel require ? what are the capacitance of those main caps ? teacher i need to know lol ! take care tony!
Great content as usual. I have a Denon receiver made in 2003. Should I recap it? If so, should I concentrate on recapping some capacitors over others? I don't see any leaked or otherwise damaged caps.
Have one from around this time, and didn't notice problems with the caps. Someone blew the surround channels. This one has 'preamp out', EXT in plugs. Works well with DSP crossover.
@ 24:28 That, should be a "common sense" practice to follow on any high power output amp. The transients you may generate from plugging in cables while an amp is on can definitely stress, if not damage components, and /or, god forbid, any speakers that may be connected to the output, and a good habit to follow would be to depower the amp before connecting input cables.
They make a hell of a lot of specialty tools that allow you to get into all kinds of places easily. They aren't cheap, but if you find yourself saying, "How in the hell am I going to get to that part to fix it?" often, it's time to look around for something to do the job. In the long run, that specialty tool will save you a lot of time, which is why they make them, of course.
Do you mean those 220uf caps on the power transistor board? I think the answer is how much does a decent quality 220uf/25v cap cost? It's just not worth the risk trusting the old parts when you already have the amp totally apart like that. Now I do agree with you on the main filter caps. Especially those old computer grade caps. If they measure good, leave them alone. Those caps were specifically designed for high AC ripple current and usually last 50 years or more.
Why do vintage guitar amps use capacitor values such as 20uf or 10uf in their power filtering? Is it because, unlike hi-fi, lower headroom and sag isn't always a bad thing? What would be the result of changing a 20uf to a modern capacitor such as a 220uf?
@@Spentelectrons Well I'm not an expert but the first filter cap after the rectifier needs not to exceed the maximum capacity that the rectifier can handle. Probably better explained on google. The second cap after the choke can be a lot bigger but maybe the amp's designers wanted the distortion resulting from this. Also back in the day big electrolytics were expensive and in (for example) 1940s radios 8uF was seen as normal. I have a1930s radio with two 4uF caps for smoothing after the rectifier.
Other thing, being voltage amplifier device with spkrs connected via transformer, it doesn't suck current the way solid state does. BTW, the capacitors used in valve equipment have directivity. The outer foil acts like screen and should be on ground side and lower impedance side if I remember correctly. Paul Carlson posted good lecture about on YT - very knowledgeable in tube equipment.
On the top board close to the front of the unit there are four white wires soldered (at about the 18:52 minute mark). Am I seeing a poor or broken solder connection on the second from the right?
update to Murphy: If anything shows up on Tony's bench, it will go wrong.... or...even with 50,000 people in the ballpark, the bird will crap on my head!
4:15 - Those are not technically emitter degeneration resistors. Emitter degeneration is negative feedback and acts on the transconductance (gm) of the transistor. It lowers the gain, improves input linearity, increases the bandwidth of the amplifier. The resistors in this circuit are intended to counter the imbalance in the class B amplifier and prevent huge current to flow between the NPN/PNP transistors.
Thank you for the correction. I really admire the work you do on your channel! I wish I had half of your knowledge! Total respect!
@@xraytonyb Thank you. You have great insight into the operation of these equipment. I enjoy your work very much.
Furthermore, the stabilization resistor doesn't reduce the damping factor of the amplifier, because the feedback resistor sits directly on the output terminal. So the output voltage is controlled at the output of the amplifier. The resistor between the emitter of the transistor and the output has no influence of the output resistance of the amplifier.
You made my week end I got up from my bench working on a tube amp then watch you fixing one awesome
By far one of the best amp restoration channels, great job!, i would definitely send you my JVC-S71 amp to check it over if i lived in the us, too bad i live 10000km away.
Love the Channel....I worked for Crazy Eddies in 84-86...Of all the Audio gear I sold...The service department would send this back to SAE due to a Very High return rate. I never would recommend SAE to anyone.
If you glance at the title of this video with the wrong glasses on, it appears to read: SAE Mark III Amplifier - Panic -
I was that soldier...
Glad to see that you found a schematic. I know how much more work it is without one. More than just figuring out the issue, is understanding the design. Amps like this definitely teach you other options in designs.. I would like more amps like this in my shop just to investigate and learn more. Good job 👍
Great video. Would like to see adjustment made with the Simpson analog
Digging this series.👍
25 volt capacitors where 16 volt capacitors are might have been a main problem here and now fixed. Can't do much more here but check those large Schottky diodes.
I wonder if replacing them with silicon carbide types would help? Yes I know they only come in a TO-220 case. Tony your showing a great video here even if you
seemed to fail at the end. There is more investigating on this amp to do.
the diodes across the output emitter resistors are feedforward diodes its purpose is cancellation of distortion of large-signal non linearity inherent in the operation of output transistors, especially at heavy loads
Interesting, thanks
Ive seen similar designs, in fact I have seen such amps that are usually class A and move to class AB/B during high power output by pulling the biasing on the driver stage. For example Marantz class A amps sometimes use this. I see them burn out regularly due to bad pots for bias adjustment, diodes who will randomly break down causing DC on the output stage, dried electrolytics etc
Great series. Very interesting.
That thing is a beast.
(@7:48 or so) - Well, that’s certainly interesting. And I don’t recall ever having seen a diode bypass an emitter resistor before. If your theory is correct, then the Schottky diode acts like a .2 volt zener, holding the emitter of Q17 at a constant voltage, which may help, as it looks like what the transistors are in, is called a cascode arrangement, with Q17 acting as an emitter follower and the signal going through R45 & R43 to the base of Q15, and through R51 & R49 to the base of Q16. But, it seems to me, that if you have a sustained high output (lots of bass or percussion), with the volume turned up, then the diode acts more like a penny in a fuse box, preventing R39 from protecting the transistors, which may explain why they blew. Just my thoughts. 😏
As it turns out, it's not a good idea! I am now anti-schottky, when it comes to the emitter resistor circuit! I will get into this more when I post part 2.
That whole circuit with Cascoded output transistors is very strange to me. I get it that they wanted to lower the Miller effect, possibly because the out transistors are slow with high Cbo. Anyway, complicated piece of vintage piece where they try to overcome one issues that creates other problems.
I would install RF bypass capacitors across the inputs for protection.
I don't like those experimental designs, like fooling around with shottky diodes in an output stage. A diode is a switching device with a certain delay and capacity and it never is perfectly linear. In this case, it is a recipe for unnecessary complications, which may lead to serious tail chasing.
Excellent, but keep the paper quiet when you are explaining :). Thanks.
exactly the same comment I came to post
Good timing, just finished part 1b :)
Another excellent vid!
Tony, I see some here think this was a failure, what I see is you have revived the AMP to where further in depth diagnostics is required to see what other issues arises. Like you have mentioned cleaning the pots and possibly making the final adjustments. Since the specifications on the amp has somewhat been changed, It is rather interesting the document mentions do not plug and unplug other components while the unit is on. I wonder if the users manual says that in BIG BOLD RED Letters. MOST people do not read the users manual, I wonder if this amp went POOOF early on when someone unplug and plugged in another component and it needed to be serviced.
Respect the work! Not sure if you familiar with the SAE 2600. Original filter caps were 9800 uf 120V DC. Do you have an opinion on what would be a good or great replacement? Thx!
I really hope all of those orange discs are good
Wow Tony you are fast. I am loving this SAE series. What a great Saturday this is. Thank you Tony!
If I remember correctly 4 of the 6 output transistors were darlington, the other 2 were regular bipolar. It was a long time ago though and I might be wrong.
You are correct! Thanks!
I don't know, but I would be tempted to put a simple RC passive LPF on the input with a -3db point at 80-100khz. I really don't see any significant audio quality advantage for a 200khz input bandwidth. No music recording is going to have anything that far up or even close. The only thing there besides RFI would be bias leakage from a vintage analog tape deck in record monitoring. That would probably destroy this amplifier. And note the amplifier circuit would still have the 200khz equivalent slew rate. The LPF would just ensure the amp never sees that coming in.
I agree. See Bob Cordell's book: Designing Audio Power Amplifiers (page 87, section 4.2). A poor design from SAE not to have such an input filter. I bet many of these amplifiers failed because of the need to power off before connecting a source.
Another thing I may try!
It is not about reproducing 200k. The advantage of 200k are things like reproducing fast transients and complex music passages. The 200k allows the amp to trace the complex input signal more faithfully.
@@jked7463 That's correct, but here is also a practical limit like 80-100K in this case. Also music or any natural sound cannot have fast rise times requiring even close to a 200khz bandwidth due to our 15psi atmosphere. This does not mean square wave testing of audio components is not needed. It is to evaluate circuit stability. But in the application of audio components, there will never be a transient that requires a 200khz bandwidth. Now good test equipment must have 5x the measured bandwidth, but that is mainly needed outside of audio electronics.
@@andydelle4509 just look at a 10K square wave in an amp that can reach 100K vs. 200k. I'll take an amp capable of 200K any day.
Hi Tony, love your videos.
I have seen you add a single power resistor/relay soft start, powered by B+ rail...I am wondering why you do not instead use NTC’s, which are designed specifically to handle the extreme initial peak power of the limiter? How do you know your chosen power resistor can handle that peak power?
Also, should the relay coil circuit be protected with its own low value fuse?
Would you have a repair on the Eiko st40 amplifier. Love your show. Thanks
I think the disclaimer at the beginning would be more fun if you added audio of the B9 Environmental Control Robot saying "Danger! Danger!"
Nelson Pass uses fast high power diodes in parallel with resistors in his F5 Turbo V3 class A design.
Awesome content! Glad this came up on my feed. New subscriber here.
This video is good for Amplifier
I thought for sure those diodes were resistors. Interesting.
its very interesting the way they made the power supply in this one it be cool if you could explain on more the way two transformers work together i know u said thier not dual separate power spplys im curious how much supply voltage dose each channel require ? what are the capacitance of those main caps ? teacher i need to know lol ! take care tony!
Great content as usual. I have a Denon receiver made in 2003. Should I recap it? If so, should I concentrate on recapping some capacitors over others? I don't see any leaked or otherwise damaged caps.
Have one from around this time, and didn't notice problems with the caps. Someone blew the surround channels. This one has 'preamp out', EXT in plugs. Works well with DSP crossover.
@ 24:28 That, should be a "common sense" practice to follow on any high power output amp. The transients you may generate from plugging in cables while an amp is on can definitely stress, if not damage components, and /or, god forbid, any speakers that may be connected to the output, and a good habit to follow would be to depower the amp before connecting input cables.
They might make a small ratchet set that has a chisel blade that would allow you to adjust those pots without having to remove them, Tony. (Shrug.)
They make a hell of a lot of specialty tools that allow you to get into all kinds of places easily. They aren't cheap, but if you find yourself saying, "How in the hell am I going to get to that part to fix it?" often, it's time to look around for something to do the job. In the long run, that specialty tool will save you a lot of time, which is why they make them, of course.
P.S. You can never, ever have enough tools. Get used to it, dude :D.
The Schotkey diode is the reason the emitter resistor didn’t burn up....
those diodes keep flyback off the outputs
14:28 - It's negative! All the electrons are gonna fall out! :)
Why replace the filter caps were they bad, did you keep them at the same value or go higher, I'm curious to no what they are.
Do you mean those 220uf caps on the power transistor board? I think the answer is how much does a decent quality 220uf/25v cap cost? It's just not worth the risk trusting the old parts when you already have the amp totally apart like that. Now I do agree with you on the main filter caps. Especially those old computer grade caps. If they measure good, leave them alone. Those caps were specifically designed for high AC ripple current and usually last 50 years or more.
@@andydelle4509 no the blue filter caps
@@johnsweda2999 I don't think Tony replaced those?
@@andydelle4509 did he.? I thought they was silver the original ones. Yes he did replace them I don't know why he never answer comments
I'll get into that in the next part
Why do vintage guitar amps use capacitor values such as 20uf or 10uf in their power filtering? Is it because, unlike hi-fi, lower headroom and sag isn't always a bad thing? What would be the result of changing a 20uf to a modern capacitor such as a 220uf?
Are you talking about Valve amps?
@@jonka1 yes
@@Spentelectrons Well I'm not an expert but the first filter cap after the rectifier needs not to exceed the maximum capacity that the rectifier can handle. Probably better explained on google. The second cap after the choke can be a lot bigger but maybe the amp's designers wanted the distortion resulting from this. Also back in the day big electrolytics were expensive and in (for example) 1940s radios 8uF was seen as normal. I have a1930s radio with two 4uF caps for smoothing after the rectifier.
@@jonka1 thank you! That makes sense. Now that you mention it I think I've heard that about the rectifier and first cap. Much appriciated.
Other thing, being voltage amplifier device with spkrs connected via transformer, it doesn't suck current the way solid state does. BTW, the capacitors used in valve equipment have directivity. The outer foil acts like screen and should be on ground side and lower impedance side if I remember correctly.
Paul Carlson posted good lecture about on YT - very knowledgeable in tube equipment.
On the top board close to the front of the unit there are four white wires soldered (at about the 18:52 minute mark). Am I seeing a poor or broken solder connection on the second from the right?
Seen that too. This board, is the bulb board for the VU's. There are 8 bulbs (4 per side).
Murphy's law, if anything can go wrong, it will. By the way , the old guy who trained me said Murp was an optimist.
update to Murphy: If anything shows up on Tony's bench, it will go wrong.... or...even with 50,000 people in the ballpark, the bird will crap on my head!