Hi,
In order to match impedances of tube source, attenuator and tube power amplifier to 1:10 rule I just need explenation how source "sees" input impedance. Which of this two calculations are correct:
1. Z_pot = 10 x Z_out and Z_in=10 x Z_pot (difficult to achieve)
or
2. Z_pot || Z_in = 10 x Z_out (possible to achieve)
In attached diagram C1 and C2 represents capacity of interconnect cables. C1 is very short cable, C2 is around 30cm. I don't know if they can be ignored in this calculation.
Z_out = 4K5
Z_in= from 86K (20khz) to 330K (<5khz)
In order to match impedances of tube source, attenuator and tube power amplifier to 1:10 rule I just need explenation how source "sees" input impedance. Which of this two calculations are correct:
1. Z_pot = 10 x Z_out and Z_in=10 x Z_pot (difficult to achieve)
or
2. Z_pot || Z_in = 10 x Z_out (possible to achieve)
In attached diagram C1 and C2 represents capacity of interconnect cables. C1 is very short cable, C2 is around 30cm. I don't know if they can be ignored in this calculation.
Z_out = 4K5
Z_in= from 86K (20khz) to 330K (<5khz)
Attachments
Last edited:
The way this is normally described you want 1:10:100 between Zout, Zpot and Zin (so Zpot is 1/10th of Zin and not vice-versa as you describe). In an opamp circuit this would typically look like Zout = 100R, Zpot = 10k and Zin = 100k, but of course tube impedance values are different and the very low Zout is easier to achieve with integrated or solid state that tubes..
There is a slight error with 30 cm cabling after the volume control. Without buffer it would be best to have it in the amplifier with short internal connections.
Input filtering before the potentiometer is lacking.
Nisbeth is right. An optimized situation would be normal sources that can drive 10 kOhm (they practically can all do that), have a 10 kOhm volume control for lowest noise in the tube amplifier. Ticks quite some boxes.
Input filtering before the potentiometer is lacking.
Nisbeth is right. An optimized situation would be normal sources that can drive 10 kOhm (they practically can all do that), have a 10 kOhm volume control for lowest noise in the tube amplifier. Ticks quite some boxes.
Last edited:
Then buffers seem appropriate. Especially with very high impedance potentiometers you may otherwise have built a variable filter. Driving a high impedance with also a relatively high impedance is a system choice.
System choices often require custom solutions.
It wouldn’t hurt to check both impedances and gain structure and see if things can be optimized.
System choices often require custom solutions.
It wouldn’t hurt to check both impedances and gain structure and see if things can be optimized.
Last edited:
Power amp is SET ec8010+kt88. Ec8010's Rg=330k which gives Z 330K from 20hz-5khz, then goes down to 85K at 20khz.
Volume control is stepped attenuator with series resistors. Values can be any i need. Now is 24K.
Source is DAC 9018 with output stage 6n8s as common cathode. Ra=10k, unbypassed Rk=820. Z_out around 6.2k. When bypass rk, zout is 4k.
System plays very very nice but i want max. quality possible.
I'm considering to change 6n8s with el34 in triode mode, CCS load, rk bypassed and get zout around 1200ohms. Make stepped att. 10k and this will be close to 1:10:100 ratio.
Volume control is stepped attenuator with series resistors. Values can be any i need. Now is 24K.
Source is DAC 9018 with output stage 6n8s as common cathode. Ra=10k, unbypassed Rk=820. Z_out around 6.2k. When bypass rk, zout is 4k.
System plays very very nice but i want max. quality possible.
I'm considering to change 6n8s with el34 in triode mode, CCS load, rk bypassed and get zout around 1200ohms. Make stepped att. 10k and this will be close to 1:10:100 ratio.
If having the volume control at the source of the power amplifier is the best option, could there be a cunning solution with a remote control pot in each one?
Impedance matching means that Zin = Zout. That's for maximum power transfer and is used in RF.
For audio (and many other applications) it is recommended that the input impedance is at least 10x the output impedance. That is to prevent the input impedance from loading down the source resulting in the highest voltage transfer. The optimum would be Zout = 0 Ω and Zin = infinite, but that doesn't exist in reality, though with enough feedback one can get pretty close.
The factor of 10 is just engineering speak for "significantly greater than". There's no magic to this number.
An unbuffered volume pot has the worst case output impedance at an attenuation of 6 dB. At this point Zout = Zpot/4. If you want to prevent the cable and power amp from loading the volume pot, you need to add a buffer on the output of the volume pot.
Tom
For audio (and many other applications) it is recommended that the input impedance is at least 10x the output impedance. That is to prevent the input impedance from loading down the source resulting in the highest voltage transfer. The optimum would be Zout = 0 Ω and Zin = infinite, but that doesn't exist in reality, though with enough feedback one can get pretty close.
The factor of 10 is just engineering speak for "significantly greater than". There's no magic to this number.
An unbuffered volume pot has the worst case output impedance at an attenuation of 6 dB. At this point Zout = Zpot/4. If you want to prevent the cable and power amp from loading the volume pot, you need to add a buffer on the output of the volume pot.
Tom