Thanks for your hard work so far John. Learning a lot from your videos about this insolent beast. Impressive that you have managed to tame the oscillation.
Nice job, my current project is a Pioneer SA7800 amplifier, which is an NSA amplifier that i got for free, but it was in rough shape. For now, the amplifier and the VFD left to fix The power supply fixed The transformer is rewired (due to excessive heat dissipation, the layers of isolation went into dust in my hands, when i started diagnosing it) Preamp and tone controls are checked, and they appears to be ok, i did a shotgun approach if we talking about recapping, due all the caps were bulged, and gave out excessive heat. Even i replaced the main filtering caps, due their series resistance was quite high. (2x 8000uf 50v caps/rail so 4caps.) for now i installed 3x 3300uf Elite 105c caps/rail(so 6 caps), in my country it was the best for the price, and i didnt had too much of money, but im pretty sure i will get something neat caps, i think the Kemet 63V 8200uf will be solid for the given job.
Those two 100pf caps are spoiling your frequency and transient response. Also, they make THD worse. What you are learning is that with old and 'slow' transistors, amp designers got away with bloody murder. Today, with modern and much 'faster' transistors, all the design and build shortcomings of an old amplifier stand out like a sore thumb. It is imperative to have proper twisted wire pairs to reduce wire inductance. Also, consider the signal return path.
Yep old slow outputs made amp design easy! Some say they sound better, others say they havent heard a really good modern amp yet. In the least, the 100pf should be thermally ljnear like silver mica or C0G. Seperate signal returns and zobel network returns are indeed highly recommended.
@Peter Larkin Well, there are some vintage mid 80s amplifiers with about the same complexity that sound phenomenal, where clever designers addressed crossover and switchover distortion in the design. Such amplifiers perform soundwise on par with tube and Class A units. Also, these designers make use of less total negative feedback and use more local negative feedback - for a good reason. Want to hear a good amplifier ? Get yourself a Technics amp and find out !
Great work so far, you are right in what you said about the time you want to spend on a project, if you work for a client it's even more important. If you have a $250 amp, and you have to spend 3 or more days, the amp is virtually lost. I always do talk with the owner about that. It's their choice in the end.
Serious comment: I wonder if the only reason the other amp works is the original transistors have a lower cut off frequency. The replacements have too much gain at the high end.
Maybe before you go to all the trouble of changing the driver stage , you could check the output with the gain back to what it was maybe the miller compensation caps in the driver were sufficient. Other idea is to try and add a miller caps on the outputs to match their input capacitance to the originals?
U.K. Well done that man, have a banana. Boy did you loose me but I'd get lost on a crystal set. Actually I got one to power a speaker quite well. Aerial around 100m.
The PCB is done I'm going to update the transistors Q10, Q12, Q4, Q14, Q8, Q11, Q15, Q13 need to change the traces to accommodate new ones with heat sink part number 634-10ABPE question is possible to use KSA1381 / KSC3503 for them?
It has to be done sometimes! I have refurbed Sony etc. with power ICs. When they fail, i have to redesign the poweramp and tonecontrol/preamp sections because its often passive preamp, or the poweramp feedback loop contains the tone circuit. Ugh what fun.
Hey the comment of stopper resistors (mine) was a suggestion. It’s one of several options and it wasn’t going to work for you. I honestly don’t care. Bob Cordell is a personal friend of mine so you can guess what kind of work I’ve done in my life. At the end of the day, it really doesn’t matter what I think. Cheers!
The stoppers still may be needed in the end. It would be a chore to install them though. I like Bob's assessment of OPS with gain. In essence he says they're not worth the trouble. I hope to hone my skills designing amplifiers, making mistakes and learning to correct them.
Great news Victory!!! good feeling I have a problem maybe you could lend insight to. Replaced blown output transistors 2SD424 2SB554 on Yamaha cr2020 quite a few years back with ECG replacements.....amp works again but with less output on the ECG side any suggestions on what to do? Would replace again if it would return the level to match the original side but have read that some of todays semis are questionable and not what their labeled . Any help would be awesome
10 years ago I have built a class D based on some International Rectifier and if I remember correctly you need to match output inductor to the impedance of your speaker, to keep response flat. But even if you don't, the response does not change that much. And only very close to 20kHz. I also don't think any 250W class D could be worse than this museum Exponate. Especially after "improving" it.
I have found some paperwork related to this design, based on IRS2092S. "IRAUDAMP5 REV 3.3" page 6, Fig.4 shows the frequency response vs load. Not so bad in my opinion.
Nice work tackling the dreaded oscillation problem. I can see how your experience paid off.
Thanks for your hard work so far John. Learning a lot from your videos about this insolent beast. Impressive that you have managed to tame the oscillation.
This is an excellent value information thank you for posting this I might clone the PCB for purpose of keeping this vintage design alive
Nice job, my current project is a Pioneer SA7800 amplifier, which is an NSA amplifier that i got for free, but it was in rough shape.
For now, the amplifier and the VFD left to fix
The power supply fixed
The transformer is rewired (due to excessive heat dissipation, the layers of isolation went into dust in my hands, when i started diagnosing it)
Preamp and tone controls are checked, and they appears to be ok, i did a shotgun approach if we talking about recapping, due all the caps were bulged, and gave out excessive heat.
Even i replaced the main filtering caps, due their series resistance was quite high. (2x 8000uf 50v caps/rail so 4caps.) for now i installed 3x 3300uf Elite 105c caps/rail(so 6 caps), in my country it was the best for the price, and i didnt had too much of money, but im pretty sure i will get something neat caps, i think the Kemet 63V 8200uf will be solid for the given job.
John in my opinion you are a miracle worker. You saved something great!!
Those two 100pf caps are spoiling your frequency and transient response. Also, they make THD worse.
What you are learning is that with old and 'slow' transistors, amp designers got away with bloody murder.
Today, with modern and much 'faster' transistors, all the design and build shortcomings of an old amplifier stand out like a sore thumb. It is imperative to have proper twisted wire pairs to reduce wire inductance. Also, consider the signal return path.
Yep old slow outputs made amp design easy! Some say they sound better, others say they havent heard a really good modern amp yet.
In the least, the 100pf should be thermally ljnear like silver mica or C0G. Seperate signal returns and zobel network returns are indeed highly recommended.
@Peter Larkin Well, there are some vintage mid 80s amplifiers with about the same complexity that sound phenomenal, where clever designers addressed crossover and switchover distortion in the design. Such amplifiers perform soundwise on par with tube and Class A units.
Also, these designers make use of less total negative feedback and use more local negative feedback - for a good reason.
Want to hear a good amplifier ? Get yourself a Technics amp and find out !
Great work so far, you are right in what you said about the time you want to spend on a project, if you work for a client it's even more important. If you have a $250 amp, and you have to spend 3 or more days, the amp is virtually lost. I always do talk with the owner about that. It's their choice in the end.
Serious comment: I wonder if the only reason the other amp works is the original transistors have a lower cut off frequency. The replacements have too much gain at the high end.
This was already established in the last video
Fascinating and informative....cheers.
Maybe before you go to all the trouble of changing the driver stage , you could check the output with the gain back to what it was maybe the miller compensation caps in the driver were sufficient. Other idea is to try and add a miller caps on the outputs to match their input capacitance to the originals?
U.K. Well done that man, have a banana. Boy did you loose me but I'd get lost on a crystal set. Actually I got one to power a speaker quite well. Aerial around 100m.
Nice work around. Beats not being used due to oscillation issues.
Will be interesting to see if you pay in terms of SNR for raising that input section voltages.
I don't see why it would plus the new transistors have a low current noise figure and are common in new high performance amp designs.
The PCB is done I'm going to update the transistors Q10, Q12, Q4, Q14, Q8, Q11, Q15, Q13 need to change the traces to accommodate new ones with heat sink part number 634-10ABPE question is possible to use KSA1381 / KSC3503 for them?
Well, the fix is kinda whole new amp you build... But its working
I'd rebuild the other one to match !
It has to be done sometimes! I have refurbed Sony etc. with power ICs. When they fail, i have to redesign the poweramp and tonecontrol/preamp sections because its often passive preamp, or the poweramp feedback loop contains the tone circuit. Ugh what fun.
Hey the comment of stopper resistors (mine) was a suggestion. It’s one of several options and it wasn’t going to work for you. I honestly don’t care. Bob Cordell is a personal friend of mine so you can guess what kind of work I’ve done in my life. At the end of the day, it really doesn’t matter what I think. Cheers!
The stoppers still may be needed in the end. It would be a chore to install them though. I like Bob's assessment of OPS with gain. In essence he says they're not worth the trouble. I hope to hone my skills designing amplifiers, making mistakes and learning to correct them.
Great news Victory!!! good feeling
I have a problem maybe you could lend insight to. Replaced blown output transistors 2SD424 2SB554 on Yamaha cr2020 quite a few years back with ECG replacements.....amp works again but with less output on the ECG side any suggestions on what to do? Would replace again if it would return the level to match the original side but have read that some of todays semis are questionable and not what their labeled . Any help would be awesome
I'd check signal levels going into the power amp and compare with the good channel.
Thank you i will have a look and see.
Throw the guts out and just put a class D board in there. Fixed it!
@@dexterperucho90 higher frequencies, what are you a bat?
10 years ago I have built a class D based on some International Rectifier and if I remember correctly you need to match output inductor to the impedance of your speaker, to keep response flat. But even if you don't, the response does not change that much. And only very close to 20kHz. I also don't think any 250W class D could be worse than this museum Exponate. Especially after "improving" it.
As someone stated on the last video, this was made by a very early now defunkt computer company. It will be collectable in future.
I have found some paperwork related to this design, based on IRS2092S. "IRAUDAMP5 REV 3.3" page 6, Fig.4 shows the frequency response vs load. Not so bad in my opinion.