The pot value compared to the impedance of the windings is tricky, and you are not necessarily going to balance the biasing but more likely to make an asymmetry between the MOSFETs operation. Guessing, increase the even-order harmonics?
All these things are open to experimentation though, and there is no reason not to try.
I have my own way of doing things, but even I do change and adapt (even if it is painful!).
For example, I have moved to using triode/bipolar hybrid followers in my line-drivers. In the process of getting the parts together to make another one, this time possibly using ECC99 rather than 7N7 (6sn7) tubes.
Thanks.
Dear Susan,
I’m flabbergasted by your sheer bottomless creativity with only one goal in mind, producing the best possible sound while not listening to those who seem to know better.
Amplifiers, Loudspeakers an even a 3D printed Cart!
It’s like the Italian cuisine, the quality of the components is what counts, not the complexity.
I just came across your thread that’s going back to 2004, and not yet been able to read the whole thread I wondered how many people have built your amp, or is it still in the talking about stadium.
Kind Regards and chapeau,
Hans
I’m flabbergasted by your sheer bottomless creativity with only one goal in mind, producing the best possible sound while not listening to those who seem to know better.
Amplifiers, Loudspeakers an even a 3D printed Cart!
It’s like the Italian cuisine, the quality of the components is what counts, not the complexity.
I just came across your thread that’s going back to 2004, and not yet been able to read the whole thread I wondered how many people have built your amp, or is it still in the talking about stadium.
Kind Regards and chapeau,
Hans
Here is a quick pic of setting up to do some basic ECC99 testing last weekend.
The meter on the right is reading mA, most of that is through the bipolar-transistors not the tubes.
Input transformer is a Sowter 9063-E, output transformer is a quad-filar wound on an EI-96 core with approx 1.4 ohm DC resistance per winding.
The meter on the right is reading mA, most of that is through the bipolar-transistors not the tubes.
Input transformer is a Sowter 9063-E, output transformer is a quad-filar wound on an EI-96 core with approx 1.4 ohm DC resistance per winding.
Dear Hans,
Thank you for your kind words.
A number of people, on this thread and from my website, have experimented or built versions of my amps.
... maybe a dozen all told?
I must admit I am not sure of the numbers, as you say this thread has been running for a long time!
Thanks.
Best,
Susan.
Many years ago, in the early 1980's a friend used this circuit to produce real power from a 12V automotive power supply. I think he was using Germanium transistors to keep the saturation down. His application was a PA speaker in the car grill, so fidelity was not critical, but the local feedback of a follower drastically reduces the distortion that a common emitter/source circuit exhibit. Cathode follower tube amps have been done too, but perhaps the driver requirements meant they were not popular. I repaired a bullhorn that used a common emitter transformer output which sounded so bad that I added a bit of feedback, making it sound much better, ie better frequency response and lower distortion, which you might not think matters to a bullhorn, but it does.
I would put some resistance between the input transformer and the Zeners if not directly on the gates. The popularity of this thread reminds me of what "Papa" figured out long ago.
I would put some resistance between the input transformer and the Zeners if not directly on the gates. The popularity of this thread reminds me of what "Papa" figured out long ago.
@steveu Interesting. If it was germanium, then PNP shown with a negative power-rail?
I have done a direct cathode-follower version of the line-driver using 6C33C triodes, in push-pull and two channels.
With 4 tubes (2 per channel), that's 160 watts just to get the heaters glowing properly. The good thing is that the a/c heater supply can be rectified and with all four in series is just right for the "HT" anode voltage.
I shall properly rebuild this at some point, it is however more suitable as a "Winter" audio component.
I have done a direct cathode-follower version of the line-driver using 6C33C triodes, in push-pull and two channels.
With 4 tubes (2 per channel), that's 160 watts just to get the heaters glowing properly. The good thing is that the a/c heater supply can be rectified and with all four in series is just right for the "HT" anode voltage.
I shall properly rebuild this at some point, it is however more suitable as a "Winter" audio component.
Hi, yes these days I solder the zeners right across the MOSFET leads; I generally use 12V parts.
The input transformers' secondaries have a fair amount of DC resistance in their windings, so my thought is that the external gate-resistors may not always be needed - unlike when being driven in a switching-circuit from a gate-driver.
This is dependent on device type, and is one of the reasons one really needs a 200MHz bandwidth oscilloscope, as if the MOSFET takes off I have seen oscillations up to 180MHz.
I'm currently building this amp with Sowter transformers, feeding the input with a DAC having an balanced output impedance of 100 ohm. The problem I have is a steep bass-cut and I attribute it to the imput impedance of 100 ohm being too high for the input transformer, wired in series. Parallel is of course even worse. Prototype was with a mains toroid and the square wave looked much better to almost perfect. Of course I want to use the Sowters in the final project. I can flatten the 1kHz square wave with a 120 ohm Rterm (not 120k) but then almost all gain is lost.
Can a circuit be build so that I can keep the balanced outputs of the DAC? I'm aware of Susan's line driver but that circuit is for unbalanced to balanced IIRC. Perhaps this can be modified? I must admit that I'm really bad with math and creating circuits so if anyone can point me to something practical that would be much appreciated.
Hugo
Can a circuit be build so that I can keep the balanced outputs of the DAC? I'm aware of Susan's line driver but that circuit is for unbalanced to balanced IIRC. Perhaps this can be modified? I must admit that I'm really bad with math and creating circuits so if anyone can point me to something practical that would be much appreciated.
Hugo
@Netlist - it is not usual to use a square-wave to test in this fashion.
A toroid will have a huge inductance compared to the Sowter, so is great for very low-bass and as such will give much flatter results for 1kHz square-waves.
Driving the power-stage directly needs something like a headphone output.
The DAC output at 100 ohms should work for sine-waves / music, but there will be some amplitude loss.
A toroid will have a huge inductance compared to the Sowter, so is great for very low-bass and as such will give much flatter results for 1kHz square-waves.
Driving the power-stage directly needs something like a headphone output.
The DAC output at 100 ohms should work for sine-waves / music, but there will be some amplitude loss.
Susan, thank you.
Indeed, square waves are ugly, at 10k I get a perfect triangle wave.
But how did you manage to make these almost perfect square waves?
http://www.susan-parker.co.uk/zeus-test-square-1khz.htm
Answer: With a line driver.
I looked at your test line driver:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.html
and I assume the AD815 alone (without the SSM2142) would do the trick if I keep the balanced DAC output.
Unfortunately, AD815 is no longer available and I can't find a suitable replacement.
The transformer line driver is a bit out of budget..
I will test with a headphone output but that can't be a permanent solution.
I'm trying to make a circuit as in
https://www.diyaudio.com/community/threads/zero-feedback-impedance-amplifiers.42259/post-7384047
with balanced inputs.
Hugo
Indeed, square waves are ugly, at 10k I get a perfect triangle wave.
But how did you manage to make these almost perfect square waves?
http://www.susan-parker.co.uk/zeus-test-square-1khz.htm
Answer: With a line driver.
I looked at your test line driver:
http://www.audiophonics.com/audiophonics-zeus-line-driver-ad815-1.html
and I assume the AD815 alone (without the SSM2142) would do the trick if I keep the balanced DAC output.
Unfortunately, AD815 is no longer available and I can't find a suitable replacement.
The transformer line driver is a bit out of budget..
I will test with a headphone output but that can't be a permanent solution.
I'm trying to make a circuit as in
https://www.diyaudio.com/community/threads/zero-feedback-impedance-amplifiers.42259/post-7384047
with balanced inputs.
Hugo
@Netlist - you could try the THS6012 as an alternative.
or
TPA6120A2 looks promising to get going with.
https://www.ti.com/lit/ds/symlink/tpa6120a2.pdf
or
TPA6120A2 looks promising to get going with.
https://www.ti.com/lit/ds/symlink/tpa6120a2.pdf
Oh, I hadn't seen that.
Susan, you are such a great help, even after 6.847 days since this thread started.
I suppose many questions have been answered before but wading trough 18.75 years of data isn't easily done.
As for the output transformer, am I correct:
4:1 1-3-5-7
2-4-6-8
is 4 ohm
and
2:1 1-3
2-6
4-7
6-8
is 8 ohm
And for the primary side I assume this is the correct schematic, independent of 4 or 8 ohm:
And last but not least, I drew a schematic to switch between 4 & 8 ohm.
I think it works but when looking at your amplifier it seems you used a simpler switch, DPDT?
But I couldn't find how to draw it without 4PDT
Hugo
Susan, you are such a great help, even after 6.847 days since this thread started.
I suppose many questions have been answered before but wading trough 18.75 years of data isn't easily done.
As for the output transformer, am I correct:
4:1 1-3-5-7
2-4-6-8
is 4 ohm
and
2:1 1-3
2-6
4-7
6-8
is 8 ohm
And for the primary side I assume this is the correct schematic, independent of 4 or 8 ohm:
And last but not least, I drew a schematic to switch between 4 & 8 ohm.
I think it works but when looking at your amplifier it seems you used a simpler switch, DPDT?
But I couldn't find how to draw it without 4PDT
Hugo
I would like to build this amp.
Is there a parts list?
I need a detailed list so I know what to order.
Would be nice. I am not into electronics a schematic how to wire would be nice, too.
I started a thread about current driven chipamps
https://www.diyaudio.com/community/threads/chip-amp-modification-to-current-drive.389985/
I am an expert for diy aluminized fullrange loudspeaker cones.
If you search for my posts easy to find.
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7380591
https://www.diyaudio.com/community/threads/how-to-make-a-whizzer-cone.398063/post-7321153
https://www.diyaudio.com/community/threads/acoustic-research-ar-7.156828/post-7177968
Is there a parts list?
I need a detailed list so I know what to order.
Would be nice. I am not into electronics a schematic how to wire would be nice, too.
I started a thread about current driven chipamps
https://www.diyaudio.com/community/threads/chip-amp-modification-to-current-drive.389985/
I am an expert for diy aluminized fullrange loudspeaker cones.
If you search for my posts easy to find.
https://www.diyaudio.com/community/threads/full-range-speaker-photo-gallery.65061/post-7380591
https://www.diyaudio.com/community/threads/how-to-make-a-whizzer-cone.398063/post-7321153
https://www.diyaudio.com/community/threads/acoustic-research-ar-7.156828/post-7177968
Last edited: