I've got a pair of 600 ohm Sennheiser DT990's. If I can get a pair of 300 ohm I'll give them a try and report back. But on these 600 ohm it sounds superb.
What surprises me is that with gain stage into CF into headphones it's absolutely quiet. No hint of noise. But same setup into amp and speakers the noise is more prominent. I would have thought it would be the opposite. CF stage into amp and speakers is very quiet. Literally have to press my ear up to the cone.
What surprises me is that with gain stage into CF into headphones it's absolutely quiet. No hint of noise. But same setup into amp and speakers the noise is more prominent. I would have thought it would be the opposite. CF stage into amp and speakers is very quiet. Literally have to press my ear up to the cone.
Last edited:
Beyerdynamic 990's maybe? Amps add much further voltage gain when the headphones use what's in the preamp only. Gain mode drops a lot of dB at 600 Ohm load. Also 600 Ohm is more insensitive to hum fields than a several kΩ amp feed. But the main issue should be excessive system voltage gain. Moderate the gain by lifting the cathode bypass cap to see where the situation goes on amps. A low intensity ground loop between preamp and amp could be at play as well.
Anchan would it be possible to see a drawing of how your amp is switchable between gain an cf modes? I'm in the process of building this as well and like the idea of both modes being available. Are you simply powering both versions all the time with an input selector switch? Any other switching to make it more versatile? Do you shut down the mode you're not using, etc?
Oh yes, Beyerdynamic 990 that's right. Mixed it up.
What you said about the gain structure makes total sense. The power amp amplifying by another 26db or so makes sense to cause a problem.
I actually built this preamp for another amp (yet to be built) that has a gain of only 12db so it will be perfect for that.
I kinda need the gain as is for headphone application so will leave the bypass cap in place, but CF mode with volume control works great with my current amp. Couldn't be happier with the result, and glad I built this flexible version.
What you said about the gain structure makes total sense. The power amp amplifying by another 26db or so makes sense to cause a problem.
I actually built this preamp for another amp (yet to be built) that has a gain of only 12db so it will be perfect for that.
I kinda need the gain as is for headphone application so will leave the bypass cap in place, but CF mode with volume control works great with my current amp. Couldn't be happier with the result, and glad I built this flexible version.
For 26dB gain power amp on even moderately sensitive speakers I would recommend the CF mode only anyway.
Interesting read for anybody not familiar: What is gain structure?
When you drive headphones directly from gain mode without the aid of the CF mode remember it loses a very big chunk of its gain potential by the output impedance and loading situation alone.
Interesting read for anybody not familiar: What is gain structure?
When you drive headphones directly from gain mode without the aid of the CF mode remember it loses a very big chunk of its gain potential by the output impedance and loading situation alone.
I didn't actually make a drawing but might be able to get to one later. But basically I just put a switch after the RCA input that then goes to each version.
At the output of the gain version, signal goes to another switch- one way to the input of the CF stage, and in the other direction just out to an RCA output. So you can choose just to take the gain stage straight out or reroute to the input of the buffer stage. The CF stage you can use alone as well, as it's output just goes out to the RCA's. But in my case, I have two sets of CF outputs. One for line level stuff and other through some 90uF caps to the headphone jack. Actually I think it easier to just make a drawing, bit hopefully this makes sense.
Currently I leave both stages on all the time, even if I'm just using one. But I don't see any reason why you can't just remove one set of tubes if find yourself set on using one mode most of the time for a given setup. With a regulated power supply such as maida for example, the output voltage remains constant. If you don't regulate the power supply, the voltqge will be rise higher with the lesser load.
At the output of the gain version, signal goes to another switch- one way to the input of the CF stage, and in the other direction just out to an RCA output. So you can choose just to take the gain stage straight out or reroute to the input of the buffer stage. The CF stage you can use alone as well, as it's output just goes out to the RCA's. But in my case, I have two sets of CF outputs. One for line level stuff and other through some 90uF caps to the headphone jack. Actually I think it easier to just make a drawing, bit hopefully this makes sense.
Currently I leave both stages on all the time, even if I'm just using one. But I don't see any reason why you can't just remove one set of tubes if find yourself set on using one mode most of the time for a given setup. With a regulated power supply such as maida for example, the output voltage remains constant. If you don't regulate the power supply, the voltqge will be rise higher with the lesser load.
When the amp will be ready remember that gain mode of this preamp is phase inverting. You can restore that by swapping + with - cable feeds at the loudspeaker terminals.
Thanks, sounds simple, I kinda though this was how you set it up. I figure its easier to machine one chassis with 4 tube holes rather than two chassis altogether, and it will all take up less space.
I figure it should work fine too. Thanks for the encouragement. I scored a few 61P1 and even some 61P1-EV a while back. I haven't used them for anything, but they seem too perfect for a Salas 6V6 Preamp. 
I've been doing some simulatin'. I have a specific application problem because I use powered speakers with 10k ohm input impedance. However, if I take the original circuit (as in post 1), increase B+ to 325V and reduce Rk to only 100 ohms (and raise Ck to 1500uF to compensate), I get Vp = 145V and Ip = 36mA through the 6V6-triode and THD drops like a stone! It's even really low into a 10k load if used that way.
That puts Pdiss for the 6V6-triode at 5W, which is not a problem. I have some Mills 10k 7W wirewound resistors to use for plate loads, and some decent 4.7uF 250V caps for the outputs.
How sensitive to power supply noise is this contraption? A Maida reg capable of 75mA output will take some serious heat sinks. I have a fairly honkin' filter choke I can use for a choke-input supply (12H 100R). Maybe I can find a 750VCT power transformer good for 100mA or something like that. 375*1.3 = 337.5V, then add another stage of RC decoupling to get to the 325V B+ for each channel. That would probably make a very quiet supply, although a really big one.
--
I've been doing some simulatin'. I have a specific application problem because I use powered speakers with 10k ohm input impedance. However, if I take the original circuit (as in post 1), increase B+ to 325V and reduce Rk to only 100 ohms (and raise Ck to 1500uF to compensate), I get Vp = 145V and Ip = 36mA through the 6V6-triode and THD drops like a stone! It's even really low into a 10k load if used that way.
That puts Pdiss for the 6V6-triode at 5W, which is not a problem. I have some Mills 10k 7W wirewound resistors to use for plate loads, and some decent 4.7uF 250V caps for the outputs.
How sensitive to power supply noise is this contraption? A Maida reg capable of 75mA output will take some serious heat sinks. I have a fairly honkin' filter choke I can use for a choke-input supply (12H 100R). Maybe I can find a 750VCT power transformer good for 100mA or something like that. 375*1.3 = 337.5V, then add another stage of RC decoupling to get to the 325V B+ for each channel. That would probably make a very quiet supply, although a really big one.
--
I aim for 1.6Hz - so for me, it's needed.
I use 1000u/330R on the cathodes of the input tube in my power amplifiers...
Rongon: If you want 325V - use a 120V isolation transformer and a Delon doubler.
Here's an example - replace the 24R resistor with your choke, and use smaller capacitors - This was designed for a power amplifier...
The N-68X from Triad is a cheap option for the transformer.
N-68X Triad Magnetics | Mouser Canada
I use 1000u/330R on the cathodes of the input tube in my power amplifiers...
Rongon: If you want 325V - use a 120V isolation transformer and a Delon doubler.
Here's an example - replace the 24R resistor with your choke, and use smaller capacitors - This was designed for a power amplifier...
The N-68X from Triad is a cheap option for the transformer.
N-68X Triad Magnetics | Mouser Canada
Attachments
Last edited:
If I have an amplifier that is -3dB at 1.6Hz, (-1dB at 3.2Hz),and a signal source that is -3dB at 1.6Hz (-1dB at 3.2Hz) . . .
Then do I also need a loudspeaker that is -3dB at 1.6Hz (-1dB at 3.2Hz)?
That way, my whole playback system is -3dB at 3.2Hz.
I can feel that.
A recording of a pipe organ playing two different 32 foot pipe notes that beat at the frequency difference (the wall is vibrating at 3.2Hz).
Then do I also need a loudspeaker that is -3dB at 1.6Hz (-1dB at 3.2Hz)?
That way, my whole playback system is -3dB at 3.2Hz.
I can feel that.
A recording of a pipe organ playing two different 32 foot pipe notes that beat at the frequency difference (the wall is vibrating at 3.2Hz).
Last edited:
First off, the Triad N-68X is a great idea, and I just happen to have one! I figure I could build the psu and audio circuit in two separate units, since that psu is gonna be hefty. Thanks for the idea.
Re: poles...
The original circuit in this thread has Rk of 680R and Ck of 1000uF. That's a -3dB pole at 0.23Hz! So I don't think 1500uF in parallel with 100R is excessive. Now, whether or not it's necessary is a good question.
As I understand it, the conventional wisdom is to have your -3dB pole a decade below the lowest frequency of interest. If the lowest frequency of interest is 20Hz, then by that standard you'll want the -3dB down point to be 2Hz. Ck of 680uF with Rk of 100R gives you -3dB at 2.34Hz.
I suppose you could call 50Hz the lowest frequency of interest and aim for -3dB at 5Hz. With Rk = 100R that would be Ck = 330uF for -3dB at 4.8Hz.
The downside of having too large a value of Ck would be the possibility of blocking on overload, but when is a line stage going to get pushed into clipping?
Is there a reason we want the line stage to act as a high pass filter?
--
Re: poles...
The original circuit in this thread has Rk of 680R and Ck of 1000uF. That's a -3dB pole at 0.23Hz! So I don't think 1500uF in parallel with 100R is excessive. Now, whether or not it's necessary is a good question.
As I understand it, the conventional wisdom is to have your -3dB pole a decade below the lowest frequency of interest. If the lowest frequency of interest is 20Hz, then by that standard you'll want the -3dB down point to be 2Hz. Ck of 680uF with Rk of 100R gives you -3dB at 2.34Hz.
I suppose you could call 50Hz the lowest frequency of interest and aim for -3dB at 5Hz. With Rk = 100R that would be Ck = 330uF for -3dB at 4.8Hz.
The downside of having too large a value of Ck would be the possibility of blocking on overload, but when is a line stage going to get pushed into clipping?
Is there a reason we want the line stage to act as a high pass filter?
--
If your phono preamp does not have a high pass 'warp filter', and you have a warped record, then yes, you do want your line stage to have a high pass filter.
Many amplifiers, and most loudspeakers do not respond very well to large 3Hz record warp signals.
And they do not respond very well to the 6Hz recorded cannon sound on the definitive Telarc recording of the 1812 overture.
But in that case, you might just want the maximum sound coming from your speakers, even if it is distorted, just to get as much
of the effect as possible.
I have been to a live performance of the 1812 Overture, and they used a Howitzer for the cannon.
Wow!
"All Generalizations Have Exceptions" - Me
Many amplifiers, and most loudspeakers do not respond very well to large 3Hz record warp signals.
And they do not respond very well to the 6Hz recorded cannon sound on the definitive Telarc recording of the 1812 overture.
But in that case, you might just want the maximum sound coming from your speakers, even if it is distorted, just to get as much
of the effect as possible.
I have been to a live performance of the 1812 Overture, and they used a Howitzer for the cannon.
Wow!
"All Generalizations Have Exceptions" - Me
Last edited:
That is a true fact, yes indeed.
My phono preamp introduces a high pass filter, a few dB down at 10Hz. I've worked pretty hard to get the cartridge/tonearm resonance to a place where small warps don't cause big rumbles. I don't believe I have a problem in that department, fortunately.
Another issue is OPT saturation on high level subsonic signals in tube amps. My solution would be to put the HPF at the input of the power amp, if one is needed. If the power amp is a zero NFB type of thing and has RC coupling, the HPF could be implemented in the interstage coupling time constant.
There are lots of ways to DIY what one needs. Even a defeatable HPF at the line stage input, right? Actually, a switchable 'Late Night Filter' for apartment dwellers would be a pretty nice idea. -3dB at 20Hz?
I also listen to a lot of (gasp!) digital source material, so the record warp problem isn't there for that.
"Everything is a compromise between this and that." - Me
My phono preamp introduces a high pass filter, a few dB down at 10Hz. I've worked pretty hard to get the cartridge/tonearm resonance to a place where small warps don't cause big rumbles. I don't believe I have a problem in that department, fortunately.
Another issue is OPT saturation on high level subsonic signals in tube amps. My solution would be to put the HPF at the input of the power amp, if one is needed. If the power amp is a zero NFB type of thing and has RC coupling, the HPF could be implemented in the interstage coupling time constant.
There are lots of ways to DIY what one needs. Even a defeatable HPF at the line stage input, right? Actually, a switchable 'Late Night Filter' for apartment dwellers would be a pretty nice idea. -3dB at 20Hz?
I also listen to a lot of (gasp!) digital source material, so the record warp problem isn't there for that.
"Everything is a compromise between this and that." - Me
metal 6v6's
Enjoying this amp a lot, and also exploring and experimenting a bit.
Are metal can 6v6's like in the pic below less microphonic, especially in the gain stage? I would think so, but I don't quite understand the mechanism in causing microphony just yet.
Enjoying this amp a lot, and also exploring and experimenting a bit.
Are metal can 6v6's like in the pic below less microphonic, especially in the gain stage? I would think so, but I don't quite understand the mechanism in causing microphony just yet.
Attachments
Bigger elements and bottle of power tubes vs Noval preamp tubes makes them prone to mechanical feedback when not used as intended i.e. using them for line level instead of power output as we heretically do here. Thus the elastic mounting and hand picking measures we took in gain mode.
The output transformer in a power amp divides their gain by its ratio so they behave well.
Metal ones I haven't had any to test in this preamp so I can't give you a reliable opinion.
The output transformer in a power amp divides their gain by its ratio so they behave well.
Metal ones I haven't had any to test in this preamp so I can't give you a reliable opinion.