"Mind, you can get virtually zero distortion from even a cheap linear pot if you use it as the feedback divider in an inverting amplifier. Clearly doing the same with a logarithmic pot is a waste of time. Doing so with an expensive logarithmic pot is a great way of discovering just how many relays the same amount of moolah can buy."
I don't know, what you make out of it?
I don't know, what you make out of it?
First he states that the esoteric TKD clearly distorted. Then he mentions that a cheap linear potentiometer performs better (zero distortion) than the logarithmic version when used as the feedback divider in an inverting amplifier.
Then he mentions an expensive TKD logarithmic potentiometer used as the feedback divider in an inverting amplifier is bettered by an also cheaper relay controlled volume control. In other words even the expensive logarithmic TKD performed less good than expected/desired/both the cheap linear and logarithmic potentiometer in the same circuit. It caused distortion.
Cheap linear potentiometer in feedback circuit zero distortion, cheap logarithmic potentiometer waste of time. Expensive TKD even worse. Relay controlled most worthwhile with cheap linear potentiometer in feedback as a second.
Many potentiometers are phobic to DC voltages so maybe...
Then he mentions an expensive TKD logarithmic potentiometer used as the feedback divider in an inverting amplifier is bettered by an also cheaper relay controlled volume control. In other words even the expensive logarithmic TKD performed less good than expected/desired/both the cheap linear and logarithmic potentiometer in the same circuit. It caused distortion.
Cheap linear potentiometer in feedback circuit zero distortion, cheap logarithmic potentiometer waste of time. Expensive TKD even worse. Relay controlled most worthwhile with cheap linear potentiometer in feedback as a second.
Many potentiometers are phobic to DC voltages so maybe...
Last edited:
So, for a 10k load, nothing less than about 4.7uF? Funny, that's what I usually use. Got a bunch of 4.7uF 400V metalized polyester caps surplus, just for that purpose. Also got some 10uF 450V metalized polypropylene, but those are HUGE.
With normal modern sources low voltages are there and a 5 mm pitch 4.7 µf 50V is enough. However...
Tube devices in general are best not loaded with the lowish 🙂 10 kOhm. It seems more optimal to choose a higher value potentiometer when using wimpy drivers. Secondly in fact most if not nearly all modern sources have better driving capability than the average inserted tube output stage. This somewhat defeats the benefits of the added tube output stage.
Thirdly most devices already have output capacitors when needed. A series chain of 2 coupling capacitors is wonderful for the audio boutiques but not beneficial for the final results. So there is no need for a 4.7 µf 450V input cap or any input coupling cap at all as well designed tube devices would include an output cap already.
Tube devices in general are best not loaded with the lowish 🙂 10 kOhm. It seems more optimal to choose a higher value potentiometer when using wimpy drivers. Secondly in fact most if not nearly all modern sources have better driving capability than the average inserted tube output stage. This somewhat defeats the benefits of the added tube output stage.
Thirdly most devices already have output capacitors when needed. A series chain of 2 coupling capacitors is wonderful for the audio boutiques but not beneficial for the final results. So there is no need for a 4.7 µf 450V input cap or any input coupling cap at all as well designed tube devices would include an output cap already.
Last edited:
Tube devices can have MOSFET source followers on their outputs and then they have good drive capability (because now they're hybrid, if you like).
This thread is about a 6V6-triode line stage, which sources roughly 20mA of Ia, with a Zout of about 2k ohms, open loop. Granted, that is only adequate to drive a 10k ohm load with a meter of interconnect cable, not optimal, but it's not completely wimpy. Of course if you wrap 6dB of NFB around that 6V6-triode line stage, the Zout will be about halved, to 1000 ohms. That's now getting close to 'optimal', or at the very least, 'quite adequate'.
If you're DIYing it, then you can know in advance what loads your source component will be driving. But if you don't know whether the end user will have a class D amp, or a tube amp, or a pair of powered speakers, then it's probably best to be prepared for all of the above. I prefer to have the option to hook a source up to a tube amp or to powered speakers, or even sometimes to an old Hafler power amp I have (claimed 47k ohms input impedance).
1/2piRC
Do you know what note is the flat-13th on a B-flat dominant seventh chord?
This thread is about a 6V6-triode line stage, which sources roughly 20mA of Ia, with a Zout of about 2k ohms, open loop. Granted, that is only adequate to drive a 10k ohm load with a meter of interconnect cable, not optimal, but it's not completely wimpy. Of course if you wrap 6dB of NFB around that 6V6-triode line stage, the Zout will be about halved, to 1000 ohms. That's now getting close to 'optimal', or at the very least, 'quite adequate'.
If you're DIYing it, then you can know in advance what loads your source component will be driving. But if you don't know whether the end user will have a class D amp, or a tube amp, or a pair of powered speakers, then it's probably best to be prepared for all of the above. I prefer to have the option to hook a source up to a tube amp or to powered speakers, or even sometimes to an old Hafler power amp I have (claimed 47k ohms input impedance).
1/2piRC
Do you know what note is the flat-13th on a B-flat dominant seventh chord?
Last edited:
They perfectly can but I haven't seen one in the wild yet. Also we are talking about the potentiometer at the input of the preamp aren't we? So the average volume control side as seen by the sources (you mentioned a DAC, those are DC coupled in many cases...) that laugh out loud at 10 kOhm load and 2V rms as the standard. Adding a tube output stage with worse driving capability than the sources themselves and a then necessary highish value output coupling capacitor only to drive a wrongly chosen too low 10 kOhm potentiometer and the 6V6 preamp with also not very high driving capability maybe is twice (thrice?) the inconvenience. Just saying...
Thanks for explaining the thread subject!
Thanks for explaining the thread subject!
Last edited:
https://file-examples.com/index.php/sample-audio-files/sample-wav-download/
I use above source file for recording through FB (test FB) and Non FB (test.wav)
Input
wavefile=d:\test\file_example_WAV_1MG.wav
Output
.wave "d:\test\test.wav" 16 44.1k v(out2)
I hear not much difference between FB ver. and original but I hear non FB ver is more sparkling or alive (should not be there?), try if you hear the same.
There are many samples instrument wave file here: https://freewavesamples.com/
I use above source file for recording through FB (test FB) and Non FB (test.wav)
Input
wavefile=d:\test\file_example_WAV_1MG.wav
Output
.wave "d:\test\test.wav" 16 44.1k v(out2)
I hear not much difference between FB ver. and original but I hear non FB ver is more sparkling or alive (should not be there?), try if you hear the same.
There are many samples instrument wave file here: https://freewavesamples.com/
Attachments
Yes, listen on real thing with you ears! I build OTL amp for last 10 years, I never listen to a wave file until now 😉 I wish I knew this earlier but is very subject, it maybe worse like you say. But I have hard time to get as good sound wave with all versions of 6v6 preamp, it's time you pull out your solidering iron, I can' help you any more.
Last edited:
Hello Salas, I spotted the long and interesting thread and I was thinking to try my hand to this building.
The reason is that I grabbed a very good globe AZ1, of early 40's, and I was looking for a good preamp (whith low-moderate gain) to put it in. I was wondering if this could be an adequate project...the alternative I considered is the Millet preamp or a modified version with the 1626 of Darling Amp. Thank you
The reason is that I grabbed a very good globe AZ1, of early 40's, and I was looking for a good preamp (whith low-moderate gain) to put it in. I was wondering if this could be an adequate project...the alternative I considered is the Millet preamp or a modified version with the 1626 of Darling Amp. Thank you
Hello, AZ1 is a very nice rectifier but its borderline capable of keeping up with an SSHV2 shunt reg set at 60mA constant current to support two channels of this preamp.
Alternatively you can combine it with an HV capacitance multiplier and it will do well. Like the one I posted in #3,885.
Alternatively you can combine it with an HV capacitance multiplier and it will do well. Like the one I posted in #3,885.
Yes, in effect I had this doubt that the AZ1 could barely fit to the project due to its specs. Anyway I consider what you reccommend, combining it to the cap. multiplier instead of the SSHV2.
I was able to lash together a test setup for the PCBs today. It's a good thing I went to test as I found some issues with the 9 pin boards- somehow when I converted it over from the octal (7pin dual pattern) version a couple traces didn't get reconnected- the cathode pin and anode pin from the socket of the right channel didn't connect to any components. Easy fix with jumpers and on to testing.
I was able to test the original style vanilla PCB with 6AQ5s installed, and get some measurements of DC voltages.
5K (2x10K) anode load, 680R cathode resistor for the following numbers.
300 volt supply:
Left channel- cathode 13.26 volts, plate 201 volts
Right channel- cathode 13.44 volts, plate 200 volts
150 volt supply:
Left channel- cathode 6.57 volts, plate 101.2 volts
Right channel- cathode 6.47 volts, plate 101.8 volts
Additionally, I was able to do some testing with the 9 pin version of the PCB, using the Russian 6P43P-E and some RCA Britain 6CW5/EL86. Since these guys have good triode curves at lowered voltages I tested both at 150, 225, and 300 volt supplies for comparison purposes of where they bias up.
Next up is the numbers for the low voltage/low impedance loadout on the 9 pin variant of the PCB. This may be primarily a fun exercise for me but I'm honestly surprised at how close the 6P43P-E and 6CW5 come to each other triode connected here.
2.5K (2x5K) anode load, 180R cathode resistor for each set of the following numbers. (I think I should increase the cathode resistor a hundred ohms or so)
150 volt supply, 6P43P-E:
Left channel- cathode 4.8 volts, plate 81.3 volts
Right channel- cathode 4.8 volts, plate 81.7 volts
150 volt supply, 6CW5/EL86:
Left channel- cathode 4.9 volts, plate 80.7 volts
Right channel- cathode 5 volts, plate 78.4 volts
225 volt supply, 6P43P-E:
Left channel- cathode 7.56 volts, plate 116.6 volts
Right channel- cathode 7.51 volts, plate 116.8 volts
300 volt supply, 6P43P-E:
Left channel- cathode 10.32 volts, plate 154.6 volts
Right channel- cathode 10.31 volts, plate 153.4 volts
300 volt supply, 6CW5/EL86:
Left channel- cathode 10.15 volts, plate 154 volts
Right channel- cathode 10.52 volts, plate 148 volts
That's all for now, tomorrow the plan is listening tests and some simple gain measurements.
I was able to test the original style vanilla PCB with 6AQ5s installed, and get some measurements of DC voltages.
5K (2x10K) anode load, 680R cathode resistor for the following numbers.
300 volt supply:
Left channel- cathode 13.26 volts, plate 201 volts
Right channel- cathode 13.44 volts, plate 200 volts
150 volt supply:
Left channel- cathode 6.57 volts, plate 101.2 volts
Right channel- cathode 6.47 volts, plate 101.8 volts
Additionally, I was able to do some testing with the 9 pin version of the PCB, using the Russian 6P43P-E and some RCA Britain 6CW5/EL86. Since these guys have good triode curves at lowered voltages I tested both at 150, 225, and 300 volt supplies for comparison purposes of where they bias up.
Next up is the numbers for the low voltage/low impedance loadout on the 9 pin variant of the PCB. This may be primarily a fun exercise for me but I'm honestly surprised at how close the 6P43P-E and 6CW5 come to each other triode connected here.
2.5K (2x5K) anode load, 180R cathode resistor for each set of the following numbers. (I think I should increase the cathode resistor a hundred ohms or so)
150 volt supply, 6P43P-E:
Left channel- cathode 4.8 volts, plate 81.3 volts
Right channel- cathode 4.8 volts, plate 81.7 volts
150 volt supply, 6CW5/EL86:
Left channel- cathode 4.9 volts, plate 80.7 volts
Right channel- cathode 5 volts, plate 78.4 volts
225 volt supply, 6P43P-E:
Left channel- cathode 7.56 volts, plate 116.6 volts
Right channel- cathode 7.51 volts, plate 116.8 volts
300 volt supply, 6P43P-E:
Left channel- cathode 10.32 volts, plate 154.6 volts
Right channel- cathode 10.31 volts, plate 153.4 volts
300 volt supply, 6CW5/EL86:
Left channel- cathode 10.15 volts, plate 154 volts
Right channel- cathode 10.52 volts, plate 148 volts
That's all for now, tomorrow the plan is listening tests and some simple gain measurements.
Last edited:
It's what we've all been waiting for -- a shootout between 6P43P-triode and 6V6-triode!
And good to see that EL86 and 6P43P voltages come out so close.
And good to see that EL86 and 6P43P voltages come out so close.
I'm getting about 10 watts out of a push pull pair of trioded 6P43P.
Now I just need to figure out a good way to do THD measurements with my workshop PC. Then things will really get interesting 🙂
