THANKS, Tombo.
Tomorrow will be fine.
I am proceeding with the fourth and hopefully I can finish this one without doing something stupid.
Good night, sweet dreams of quiet DC and take care,
Tomorrow will be fine.
I am proceeding with the fourth and hopefully I can finish this one without doing something stupid.
Good night, sweet dreams of quiet DC and take care,
They will gain even more than HF range amps... prepare yourself & hold onto your knickers... see here... how low and uniform that impedance is.. beautiful!
Power supply - 99% of the final audio quality... the whopping 1% that remains - everything else.
Measured voltages indicate that opamp supply, voltage reference and regulation are very likely OK. Setting of trimmer potentiometer, as you have it, sets requested output voltage to 15V!
If you have accidentally connected power amplifier with wrong bias (much more than 3A) there is dangerously high dissipation. With 43V input – 15V output, there would be 28V across regulator. Multiply voltage and current for dissipation → results with > 84W, probably from 100 – 200 W.
That would cook power MOSFET as we can’t remove that much heat with standard thermal interface from the small TO-220 case. Silpad thermal resistance is close to 1°C/W. Alumina (ceramic) pads + good thermal paste are better, but not necessary.
Even if output voltage was set right, with very high bias, like 10A, there is a short dangerous period while output voltage slowly rises, and dissipation is high. Regulator reaches full output in 3 – 4 seconds time.
So, measure resistance between MOSFET pins 2-3 to check weather it is OK, as a first step.
Before connecting amplifier to the next regulator, adjust output voltage close to expected/required.
Mine worked OK directly (well after the correction of the heatsink), there is nothing unclear with either PCB or BOM. Just minor imperfections like the potentiometer and the wrong cap size.
Good afternoon, Tombo,
I had turned down the trimmer just to see if anything would happen - when I took those voltage measurements there was no way to set the voltage. I did not think to attempt to move it back to where it was when the thing worked.
D10 was shorted - I replaced that - and after doing that it was clear that the mosfet was destroyed. Could a shorted D10 make the mosfet test inconclusive? I would not know.
The test load gets plenty warm with the ruined mosfet. I guess this is to be expected?
I need to order a few. I might have one laying around. I proceeded with #4 and hope to have that working soon. I guess I could have got it done last night but the thought of a couple of cocktails and listening in mono (always much better than nothing) was irresistible.
There is no doubt my idiocy of not connecting the load resistor would have asked Herculean tasks of R25. I will assume that the mosfet and D10 saved the rest of R25 by failing.
THANKS, thanks, and thanks again for this regulator and your intimate knowledge of it.
I had turned down the trimmer just to see if anything would happen - when I took those voltage measurements there was no way to set the voltage. I did not think to attempt to move it back to where it was when the thing worked.
D10 was shorted - I replaced that - and after doing that it was clear that the mosfet was destroyed. Could a shorted D10 make the mosfet test inconclusive? I would not know.
The test load gets plenty warm with the ruined mosfet. I guess this is to be expected?
I need to order a few. I might have one laying around. I proceeded with #4 and hope to have that working soon. I guess I could have got it done last night but the thought of a couple of cocktails and listening in mono (always much better than nothing) was irresistible.
There is no doubt my idiocy of not connecting the load resistor would have asked Herculean tasks of R25. I will assume that the mosfet and D10 saved the rest of R25 by failing.
THANKS, thanks, and thanks again for this regulator and your intimate knowledge of it.
Yeah, it is not straightforward to measure having more damaged parts connected together.
Anyway, it was the second failure case: very high load current (SIT can pull a lot without batting an eyelash) during slow power-up phase, overheating and destroying the MOSFET. When output voltage has reached nominal value, there is no problem even with extreme load currents.
Design is very robust and burning a MOSFET probably didn't make any additional damage, except for D10 in some cases.
Suppose one don't like to handle transformers - what do you say about feeding a pair of these boards for a +/- supply but fed by 2 pcs of SMSP DC units?
//
//
No problems at all. As said here:
https://www.diyaudio.com/community/...ilter-and-super-regulator.420933/post-7892529
https://www.diyaudio.com/community/...ilter-and-super-regulator.420933/post-7892529
Number four works from the start. But then so did number three ... Won't forget that wire again.
Now to wait for more devices from MOUSER.
Thanks, again, Tombo
Now to wait for more devices from MOUSER.
Thanks, again, Tombo
And now for a real puzzler.
Number four is working fine so I placed the tweeter amplifier back where it goes; the two amplifiers are stacked with a generous distance between them.
This working amplifier was placed on a table until I was (finally) through with the bass amplifier (with number four R25).
Voltages are where they were before being put on the table.
When the amplifier is turned there is about one second of proper volume coming from the driver and then it dims to being just loud enough to hear it.
Same thing when you turn it off, there is a moment of proper volume (the level is low while I am doing this) before it drops off to nothing.
Voltages for the problem channel are:
2 - 20.4
3 - 16.3
4 - 16.3
5 - 38.5
6 - 3.67
Looked at the other tweeter amp and the voltages are about the same.
The most interesting aspect is that the THF51 is not getting hot. You can touch it with your finger after it has been on for an hour.
Does it sound like a failed THF51? Is there any way R25 could stop delivering the current required? I know when the mosfet is blown you still get warmth on the test load.
Any ideas would be appreciated. Hope all of you trust that if I did something stupid I would admit to it!
Number four is working fine so I placed the tweeter amplifier back where it goes; the two amplifiers are stacked with a generous distance between them.
This working amplifier was placed on a table until I was (finally) through with the bass amplifier (with number four R25).
Voltages are where they were before being put on the table.
When the amplifier is turned there is about one second of proper volume coming from the driver and then it dims to being just loud enough to hear it.
Same thing when you turn it off, there is a moment of proper volume (the level is low while I am doing this) before it drops off to nothing.
Voltages for the problem channel are:
2 - 20.4
3 - 16.3
4 - 16.3
5 - 38.5
6 - 3.67
Looked at the other tweeter amp and the voltages are about the same.
The most interesting aspect is that the THF51 is not getting hot. You can touch it with your finger after it has been on for an hour.
Does it sound like a failed THF51? Is there any way R25 could stop delivering the current required? I know when the mosfet is blown you still get warmth on the test load.
Any ideas would be appreciated. Hope all of you trust that if I did something stupid I would admit to it!
Voltages seems right and regulator is set to 36V output. If output voltage is stable (wait 10 seconds after power on, before checking) then supply works as intended. BTW, what is input voltage? Some 2V drop from input to output is more than enough.
If amplifier doesn’t pull current (staying cold) while being supplied proper volatge, then there is a problem with amplifier. Is that the same channel that you forgot to connect biasing and that pulled very large current, burning supply MOSFET? If so, it could be something in amplifier is damaged, SIT, choke, power resistor …..?
It’s late here and I was just about to end my music listening session and shut down laptop. So, we can continue tomorrow.
If amplifier doesn’t pull current (staying cold) while being supplied proper volatge, then there is a problem with amplifier. Is that the same channel that you forgot to connect biasing and that pulled very large current, burning supply MOSFET? If so, it could be something in amplifier is damaged, SIT, choke, power resistor …..?
It’s late here and I was just about to end my music listening session and shut down laptop. So, we can continue tomorrow.
I knew it was getting late.
No this amplifier was working fine until I put it aside to get to the woofer amplifier which resides beneath it.
It sat on a table for three days while I worked on that amplifier - the amplifier I left the wire unconnected,. It is working fine.
I have measured all of the few components and they are are what they should be - even though it is not possible to get an accurate reading of the choke but it is what it has been when I measured it. The input load and the gate resistor are what they should be.
2 volts drop.
And then I realized I had not looked at the speaker connections. Sure enough a loose wire on the output. Shows that problems can be an advantage. That connection was ready to go. I would have never expected that to do what it did but then I know I have lots to learn.
DAMNIT. Sorry to have bothered you. Three working well.
Thanks for your patience, Tombo
No this amplifier was working fine until I put it aside to get to the woofer amplifier which resides beneath it.
It sat on a table for three days while I worked on that amplifier - the amplifier I left the wire unconnected,. It is working fine.
I have measured all of the few components and they are are what they should be - even though it is not possible to get an accurate reading of the choke but it is what it has been when I measured it. The input load and the gate resistor are what they should be.
2 volts drop.
And then I realized I had not looked at the speaker connections. Sure enough a loose wire on the output. Shows that problems can be an advantage. That connection was ready to go. I would have never expected that to do what it did but then I know I have lots to learn.
DAMNIT. Sorry to have bothered you. Three working well.
Thanks for your patience, Tombo
Only one remaining.
Hopefully, your amplifiers will be capable to reveal something new from the Topping D90III discrete. 🙂
They already are.
Going to replace the switcher in those with SALAS shunts. All built and ready to go but then R21 got my attention and then R25 and that seemed like the most important thing to do.
Needing so many DACs I cannot begin to afford the the fancy ones. Never heard one of them so I have no basis of comparison.
Even my phonograph goes through the digital stuff which is interesting since - even with all of that - it is clear what LPs do better than files.
What baffles me is how superior the LP is in the mid and low bass where you would presume digital would be superior. But then I think about how FIR filters are HUGE in the low frequencies so there must be a correlation there.
Have a good week and take care,
Going to replace the switcher in those with SALAS shunts. All built and ready to go but then R21 got my attention and then R25 and that seemed like the most important thing to do.
Needing so many DACs I cannot begin to afford the the fancy ones. Never heard one of them so I have no basis of comparison.
Even my phonograph goes through the digital stuff which is interesting since - even with all of that - it is clear what LPs do better than files.
What baffles me is how superior the LP is in the mid and low bass where you would presume digital would be superior. But then I think about how FIR filters are HUGE in the low frequencies so there must be a correlation there.
Have a good week and take care,
For several reasons, GLED voltage reference is not suitable for this design and it would neither provide better noise performance. Q5 is not a CCS but part of shunt regulated supply for opamp. Circuit is essentially the same as R21 version and design choices were explained there:
https://www.diyaudio.com/community/...ower-supply-r21-ps-module.376003/post-6759485
This evening I will finally get the last amplifier back into the system.
I did check R25 the third to make sure all it needed was to have the mosfet replaced and it passed.
Extreme_boky is correct - again listening to a single speaker (with mono signal) while getting the amplifier fettled - with both high and low fed by R25 supplied amps the sound can almost spread across the room.
If only Tombo could figure out a way to get the THF51 to reach its best quicker than 45 minutes! What happens when these things get warmed(/hotted?) up is a gift from Providence. And so is R25 by way of Tombo. I am going to try delaying the use of the fans for about fifteen minutes and hear if that quickens the time needed for best performance.
Once heard one cannot return from whence one came ...
I did check R25 the third to make sure all it needed was to have the mosfet replaced and it passed.
Extreme_boky is correct - again listening to a single speaker (with mono signal) while getting the amplifier fettled - with both high and low fed by R25 supplied amps the sound can almost spread across the room.
If only Tombo could figure out a way to get the THF51 to reach its best quicker than 45 minutes! What happens when these things get warmed(/hotted?) up is a gift from Providence. And so is R25 by way of Tombo. I am going to try delaying the use of the fans for about fifteen minutes and hear if that quickens the time needed for best performance.
Once heard one cannot return from whence one came ...
As is with every class A amplifier, with temperature change bias current changes as well, and for most circuits higher operating temperature means higher bias current. Every change in bias current affects distortion and harmonics ratio. With class A, higher bias current always equals less distortion and better harmonics profile (more H2 and less of higher harmonics). So, your amplifier simply reaches optimum bias at a certain temperature.
But, if you also have a class A preamplifier, it may be that sound change after warm up is coming from there?
Very true. But, after some time, however the sound is good, it becomes new norm and you start to wonder why it doesn’t/could it sound better. 😆