For what it's worth, I finally did some listening this past weekend with the SCG driving the TDV output stage. Just one channel, but I'm very happy with the sound. I'm no Stereophile writer but I would describe it as both detailed and non-fatiguing. It definitely sounded, to my ears, "tubey".
From a measurements perspective, once I had tweaked the SCG by removing the buffer and also adjusting the Vd to ~37V (this is what I settled on with my 60.4V supply) I measured the distortion of the whole amplifier at 1W into 8 ohms and it had dropped to 0.094%. My prior measurement at the same output was 0.25% - but this isn't a controlled experiment, I took that previous measurement with the amplifer output stage on my bench supply and prior to adjusting the SCG. Either way, I'm happy.
Now I'm anxious to get going on the enclosure for this project so I can give it a proper listen.
Attachments
The rebuilt preamp build.
@metaphile
With power supply at 68V and Vds at 34V and output JFETs desoldered, it really does what it’s supposed to do.
THD will rise as input level increases, but in practice, when you have a power amp after it, input level will be much much lower than 2V and THD will indeed be much much lower.
I still need to clean the bench area, put a knob on it and set it up for listening with proper speakers. I am sure it’ll be greatly great. Dialed Vds down to 34V for extra tubiness, which puts it at half the supply, as originally intended.
Papa says it sounds pretty good when distortion rises linearly with level as long as it stays below 1% or so. This preamp does exactly that. It’s H2 dominant and starts low and increases linearly with a ton of headroom (clean swing). Hang a follower on it and you’re golden.
I approve this message.
Now, I am about to test it with a class D amp because all my power amps are either class D or tube.
(I cannot believe I posted this in the Pass Labs forum. Will I make it through the day? Should I request witness protection?)
Now, I am about to test it with a class D amp because all my power amps are either class D or tube.
(I cannot believe I posted this in the Pass Labs forum. Will I make it through the day? Should I request witness protection?)
The headphones amp build is the best headphones amp I have heard with ALL of my headphones. I mean 8 ohm to 600 ohm and low efficiency to high efficiency…is a good tests assortment.
Omicron is excellent with most of my headphones, but it’s very very low distortion and can be fatiguing because I end up listening too loud without realizing it.
Omicron is excellent with most of my headphones, but it’s very very low distortion and can be fatiguing because I end up listening too loud without realizing it.
I think all of us here would trust your ears, ElArte.
I am looking forward to Rahul telling us more about his lateral FET amplifier.
I am looking forward to Rahul telling us more about his lateral FET amplifier.
That R25 is amazing. A bit cramped in places, but only because I had to use more heat sinks than most builders.
As for the big power amps, there are so many and so many variations, I really don’t know where to start.
I think I bought boards for the USSA5 at some point.
And I know of Nelson’s many amps, ZM, Rahul’s TDV (I don’t have VFETs), Honey Badger, etc.
If you are not worried about your electric bill, I highly recommend rmthatcher's Classic Aleph Amplifier. The Aleph 60 and Aleph 2 sound soooooooo good.
Now I think I remember I bought an Aleph Jzm kit at some point.
I’ll have to check that out.
I’ll have to check that out.
Gee, why are chassis’ 5 times the price of electronic parts that took humanity thousands of years to discover!?
That always takes me by surprise.
That always takes me by surprise.
The report I promised is delayed a bit. I am chasing a noise issue that is pot-sensitive (highest in the middle position) and only when output RCAs are connected to an amp.
As a test, I put a cheat plug on the amp, but it had no effect, which eliminates the possibility of a ground loop introduced by the amp.
When the amp is not connected, there is no noise at all no matter the pot position.
I don’t think it has anything to do with the removal of the output buffer, but, alas, I didn’t test that when it was still in place.
As a test, I put a cheat plug on the amp, but it had no effect, which eliminates the possibility of a ground loop introduced by the amp.
When the amp is not connected, there is no noise at all no matter the pot position.
I don’t think it has anything to do with the removal of the output buffer, but, alas, I didn’t test that when it was still in place.
Less than 5 inches in both cases.
But I may have a pernicious ground loop.
More than 3 devices and I get caught every time.
But I may have a pernicious ground loop.
More than 3 devices and I get caught every time.
Finally got my left channel boards in place - these are buffer-less
A noise that had been driving me nuts is now gone. Must have been my wiring. Silence is wonderful.
Left channel without buffer - right channel with. So far I do not detect much difference at all. Only one evenings listening - not enough to really hear it.
The left channel woofer has less gain than the right - since all of the FETs were almost exactly the same for all four I cannot explain a gain difference.
I can only measure what comes out of my speakers so the distortion numbers are more of the room than the electronics but there seems to be very little difference in second versus third order distortions.
Could it be that the buffer in the right channel affects the gain - or the ability to drive the THF-51 from 500 hz downwards? That seems unlikely but I would not know. But there is about 8 dB difference. I am using an FE2022 at half gain in front of the bass SCG - I have not checked to see if that is the problem. Need to switch them between channels to be sure, I need to add the EQ for both speakers is the same - started out with both at the same level until it was obvious the difference. Highest regards for REW!
There is one thing for sure - the selection process for the FETs - knowing what to select for - makes a big difference. It is the same sound, only more of the good. Worth the trouble.
A confession - when building these I wanted them to be the same as the previous channels. One gets locked into a mindset - turns out I had used the wrong output - for most the input to the output capacitor - I am using the input cap of the amplifier as the output cap - instead of a jumper I have a stiff wire in the correct hole - which with the second pair was the incorrect hole.
I set up to set the voltage on the one with the buffer FETs pins ones cut - and all seems fine - I have no idea I have done this at this point. I connect the amplifier and this muffled mess of a sound. You can imagine the disappointment. This board did allow me to set the voltage to 39.5 which was half of my B+.
I hooked up the second board which had no buffer components (FETs & resistors) and the voltage between the pin that was actually connected to nothing and ground went sky high. I thought I had blown up those FETs.
Eventually it occurred to me to look at where I was taking the output and I see it is wrong in both cases. I am, at least, consistently wrong. So with taking the reading from the correct output with the clipped buffer FETs the voltage is within a few tenths of a volt where it was supposed to be which does mystify me.
For the second board, with the correct output everything was fine and shortly afterwards I had a stereo system again. EQ followed.
But the fact that the clipped FET board allowed setting the voltage to almost exactly where it needed to be ... Does this surprise you, Rahul?
Sorry to go on so long.
A noise that had been driving me nuts is now gone. Must have been my wiring. Silence is wonderful.
Left channel without buffer - right channel with. So far I do not detect much difference at all. Only one evenings listening - not enough to really hear it.
The left channel woofer has less gain than the right - since all of the FETs were almost exactly the same for all four I cannot explain a gain difference.
I can only measure what comes out of my speakers so the distortion numbers are more of the room than the electronics but there seems to be very little difference in second versus third order distortions.
Could it be that the buffer in the right channel affects the gain - or the ability to drive the THF-51 from 500 hz downwards? That seems unlikely but I would not know. But there is about 8 dB difference. I am using an FE2022 at half gain in front of the bass SCG - I have not checked to see if that is the problem. Need to switch them between channels to be sure, I need to add the EQ for both speakers is the same - started out with both at the same level until it was obvious the difference. Highest regards for REW!
There is one thing for sure - the selection process for the FETs - knowing what to select for - makes a big difference. It is the same sound, only more of the good. Worth the trouble.
A confession - when building these I wanted them to be the same as the previous channels. One gets locked into a mindset - turns out I had used the wrong output - for most the input to the output capacitor - I am using the input cap of the amplifier as the output cap - instead of a jumper I have a stiff wire in the correct hole - which with the second pair was the incorrect hole.
I set up to set the voltage on the one with the buffer FETs pins ones cut - and all seems fine - I have no idea I have done this at this point. I connect the amplifier and this muffled mess of a sound. You can imagine the disappointment. This board did allow me to set the voltage to 39.5 which was half of my B+.
I hooked up the second board which had no buffer components (FETs & resistors) and the voltage between the pin that was actually connected to nothing and ground went sky high. I thought I had blown up those FETs.
Eventually it occurred to me to look at where I was taking the output and I see it is wrong in both cases. I am, at least, consistently wrong. So with taking the reading from the correct output with the clipped buffer FETs the voltage is within a few tenths of a volt where it was supposed to be which does mystify me.
For the second board, with the correct output everything was fine and shortly afterwards I had a stereo system again. EQ followed.
But the fact that the clipped FET board allowed setting the voltage to almost exactly where it needed to be ... Does this surprise you, Rahul?
Sorry to go on so long.
I put the preamp build back on the bench to sort out the noise issue.
I don’t think the gain is reduced.
I’ll reread your post tonight to make sure I understood everything.
So, it seems the input impedance (unknown) of the amplifier doesn’t agree with the output impedance of the preamp. In this very specific case.
There is no ground loop. The pot is not at fault either. All was quiet on the bench, with headphones connected to the RCA output jacks, even though FR was clearly off, because no headphones buffer in between.
I’ll try with another amplifier, of known input impedance.
There is no ground loop. The pot is not at fault either. All was quiet on the bench, with headphones connected to the RCA output jacks, even though FR was clearly off, because no headphones buffer in between.
I’ll try with another amplifier, of known input impedance.