Generally I see that often less is more. More tubes may be added if there are good reasons, but only then. So I thought that between the passive attenuator and the input stage of the valve amplifier there may be some circuits, filters or what ever. It would be tempting idea to leave out buffer amplifier stages. In that case it would be beneficial if the input impedance of the first amplifier stage was very high. So, it would be reasonable and simple to add a bootstrap circuit for the input triode. But, are there reasons to avoid a bootstrap circuit?
Buffers aren't just fitted for impedance matching they can amplify a low output device to say-2 volts RMS or more .
Bootstraps are usually fitted to solid state (BJT ) devices not valve triodes with very high natural input impedance's .
Very high input impedance has its problems ,the liability to RF/noise /radiation of equipment etc to pick up by the amplifier .
Bootstraps are usually fitted to solid state (BJT ) devices not valve triodes with very high natural input impedance's .
Very high input impedance has its problems ,the liability to RF/noise /radiation of equipment etc to pick up by the amplifier .
Nikolas Ojala,
I like your tag line:
"High-order harmonics are more offensive than low."
Tell that to the lead flautist in a major classical orchestra.
Then, tell that to a hard rock electric guitarist who drives his guitar amp into hard clipping.
Just saying . . .
I like your tag line:
"High-order harmonics are more offensive than low."
Tell that to the lead flautist in a major classical orchestra.
Then, tell that to a hard rock electric guitarist who drives his guitar amp into hard clipping.
Just saying . . .
That is my point.
For both the Performer; and for the Listener.
A famous quote from a famous composer/band leader/producer/performer:
In response to "What kind of music do you like?"
. . . His answer: "Good Music" - Duke Ellington
For both the Performer; and for the Listener.
A famous quote from a famous composer/band leader/producer/performer:
In response to "What kind of music do you like?"
. . . His answer: "Good Music" - Duke Ellington
A small triode grid, naked, is about 100Meg effective resistance. It may "require" a 1Meg max DC leak, but this can be cheated; and 1Meg is high enough for nearly any audio purpose (condenser mike head the obvious exception).
Harmonics of musical instrument vs distortion. Not the same thing.
You are talking about a desired feature of a musical instrument.
You are talking about a desired feature of a musical instrument.
The whole point of musical enjoyment is to listen to what your ears prefer not some guru pointing out what you SHOULD listen to .
We are ( not yet ) programmed to accept what another likes to listen to each persons brain is unique in its processing of visual and aural content , advertising hasn't reached that futuristic peak in telling you what to eat and you obey --you eat it --then spit it out because you don't like it then wonder why people are put under so much pressure to conform to set values of taste.
We are ( not yet ) programmed to accept what another likes to listen to each persons brain is unique in its processing of visual and aural content , advertising hasn't reached that futuristic peak in telling you what to eat and you obey --you eat it --then spit it out because you don't like it then wonder why people are put under so much pressure to conform to set values of taste.
Very often, the harmonic content of a flute is very low.
That was my example of a "low distortion music tone".
But there is also a playing mode where its harmonics become quite dominant.
The sound of that music is all up to the Flautist's intent.
Music without any variance can/might be boring.
The Rock Guitarist who clips his amplifier chooses to do so on purpose (or does not know how to prevent that).
The sound of that music tone is very much dominated by the upper harmonics.
What any person's playback system does to either faithfully playback, or to slightly modify, or highly modify either of those musicians recordings is up to the taste of the persons choice.
To each his own.
That was my example of a "low distortion music tone".
But there is also a playing mode where its harmonics become quite dominant.
The sound of that music is all up to the Flautist's intent.
Music without any variance can/might be boring.
The Rock Guitarist who clips his amplifier chooses to do so on purpose (or does not know how to prevent that).
The sound of that music tone is very much dominated by the upper harmonics.
What any person's playback system does to either faithfully playback, or to slightly modify, or highly modify either of those musicians recordings is up to the taste of the persons choice.
To each his own.
Last edited:
Nikolas Ojala,
What is the signal source that requires that it be connected to a 10Meg Ohm input buffer or 10Meg Ohm input stage?
A 1 foot RCA to RCA shielded cable may have 30 pF of capacitance (might be less, but then instead, use a 3 foot shielded cable, which is at least 30pF).
30pF has 531k impedance at 10kHz.
Will that 531k Ohm impedance affect the signal source?
I can hear 10kHz, so the answer to that might be important.
What is the signal source that requires that it be connected to a 10Meg Ohm input buffer or 10Meg Ohm input stage?
A 1 foot RCA to RCA shielded cable may have 30 pF of capacitance (might be less, but then instead, use a 3 foot shielded cable, which is at least 30pF).
30pF has 531k impedance at 10kHz.
Will that 531k Ohm impedance affect the signal source?
I can hear 10kHz, so the answer to that might be important.
Last edited:
Without additional circuits 1 MΩ would be just fine
I see it this way.
My CD player's user manual states that to achieve minimum distortion the input impedance of the receiving device (amplifier) should not be less than 50 kΩ. So I thought that let that be 100 kΩ then. That 100 kΩ may be a stepped attenuator or a potentiometer, but the resistance is the same. Between the attenuator and the first amplifier stage (probably a triode) there may be some passive circuits, such as balance control circuit or an adjustable loudness filter. There should be some useful playground between the attenuator's 100 kΩ and the input impedance of the first triode stage. What ever passive circuits there would be, they could be added later. The design could be modular but completely in one unit.
(a) simple
Attenuator 100 kΩ
(short wires)
Triode input 10 MΩ
(b) one passive circuit
Attenuator 100 kΩ
Circuit 1 MΩ
Triode input 10 MΩ
(c) two passive circuits
Attenuator 100 kΩ
Circuit a 464 kΩ
Circuit b 2.15 MΩ
Triode input 10 MΩ
I see it this way.
My CD player's user manual states that to achieve minimum distortion the input impedance of the receiving device (amplifier) should not be less than 50 kΩ. So I thought that let that be 100 kΩ then. That 100 kΩ may be a stepped attenuator or a potentiometer, but the resistance is the same. Between the attenuator and the first amplifier stage (probably a triode) there may be some passive circuits, such as balance control circuit or an adjustable loudness filter. There should be some useful playground between the attenuator's 100 kΩ and the input impedance of the first triode stage. What ever passive circuits there would be, they could be added later. The design could be modular but completely in one unit.
(a) simple
Attenuator 100 kΩ
(short wires)
Triode input 10 MΩ
(b) one passive circuit
Attenuator 100 kΩ
Circuit 1 MΩ
Triode input 10 MΩ
(c) two passive circuits
Attenuator 100 kΩ
Circuit a 464 kΩ
Circuit b 2.15 MΩ
Triode input 10 MΩ
What is the model of your CD player?
Does it use a Vacuum Tube to drive the output connector?
How long of an RCA to RCA shielded cable will you use?
My CD player specifies a minimum load of 10k Ohms.
All of my tube amplifiers have 50k volume controls.
That is so that the input tube's Miller Effect Capacitance does not significantly change the high frequency response, when the volume control is turned through its range.
I do not use balance controls, and I do not use Loudness Contour circuits.
I have not done that for decades.
If I was to do that again, I would put it later in the amp circuitry, not at the input like so many amplifiers used to do.
Amplifiers that I owned decades ago, that did that (both tube and solid state), made the tradeoff of using volume, loudness contour, and balance all in the same circuit.
That affected the mid and high frequency response of the amplifier, depending on the setting of the volume and the balance controls (both with the loudness contour on, and with the loudness contour off).
One un-nerving effect was that as you changed the balance, the sound levels changed (as they should); but the high frequency of one channel rolled off too. That is not a good balance.
I measure the capacitance of the RCA to RCA cables, before I use them in my system. Some do not get used for the various Hi Fi systems I have in the house.
Does it use a Vacuum Tube to drive the output connector?
How long of an RCA to RCA shielded cable will you use?
My CD player specifies a minimum load of 10k Ohms.
All of my tube amplifiers have 50k volume controls.
That is so that the input tube's Miller Effect Capacitance does not significantly change the high frequency response, when the volume control is turned through its range.
I do not use balance controls, and I do not use Loudness Contour circuits.
I have not done that for decades.
If I was to do that again, I would put it later in the amp circuitry, not at the input like so many amplifiers used to do.
Amplifiers that I owned decades ago, that did that (both tube and solid state), made the tradeoff of using volume, loudness contour, and balance all in the same circuit.
That affected the mid and high frequency response of the amplifier, depending on the setting of the volume and the balance controls (both with the loudness contour on, and with the loudness contour off).
One un-nerving effect was that as you changed the balance, the sound levels changed (as they should); but the high frequency of one channel rolled off too. That is not a good balance.
I measure the capacitance of the RCA to RCA cables, before I use them in my system. Some do not get used for the various Hi Fi systems I have in the house.
Last edited:
Cables could be something else
Sony CDP-XB930 (no vacuum tubes)
Cables 50 cm RG 58 but later could be something else.
That was a good question about the cables. RG 58 capacitance is about 82 pF/m.
41 pF is about 200 kΩ @ 20 kHz.
Together 41 pF capacitance and 100 kΩ resistance in parallel connection make a 89 kΩ load @ 20 kHz. Should be ok.
Sony CDP-XB930 (no vacuum tubes)
Cables 50 cm RG 58 but later could be something else.
That was a good question about the cables. RG 58 capacitance is about 82 pF/m.
41 pF is about 200 kΩ @ 20 kHz.
Together 41 pF capacitance and 100 kΩ resistance in parallel connection make a 89 kΩ load @ 20 kHz. Should be ok.
Yes, that is what you did not ask. Tube capacitance generally swamps resistances at the top of the audio band.
I'm also stunned anybody suggests a 50k even 10k load on solid-stage sources. Even a 19 cent opamp can show vanishing THD well below 5k, typically 2k, and the 22 cent chip can go 600 Ohms clean. 10k has been a "standard test load" though I doubt anybody follows standard anymore.
I'm also stunned anybody suggests a 50k even 10k load on solid-stage sources. Even a 19 cent opamp can show vanishing THD well below 5k, typically 2k, and the 22 cent chip can go 600 Ohms clean. 10k has been a "standard test load" though I doubt anybody follows standard anymore.
So, practically speaking, the input impedance is limited by capacitance.
I did some calculations. For example ECC83 input capacitance is 1.6 pF which looks like small, but the calculation shows that the (capacitive) impedance is about 5 MΩ @ 20 kHz. That is really not much. Have some gain and bring in the Miller effect: Where is the playground now? It just vanished.
I think I got the answer to my question.
Thanks for your contribution.
I did some calculations. For example ECC83 input capacitance is 1.6 pF which looks like small, but the calculation shows that the (capacitive) impedance is about 5 MΩ @ 20 kHz. That is really not much. Have some gain and bring in the Miller effect: Where is the playground now? It just vanished.
I think I got the answer to my question.
Thanks for your contribution.
It is very common to bootstrap the input if the required grid resistor is less than 1M. It is accomplished by splitting the cathode resistor to hook the grid resistor. Of course an input capacitor is necessary.
So if we add a second cathode resistor, and an input capacitor . . .
We still have to calculate the Miller Effect Capacitance that may be a problem at high frequencies.
Gain is never free.
Tektronix, in the 1960s made a couple of models of vacuum tube probes.
They were cathode followers with bootstrapped rg.
The Miller Effect multiplier was (1 + < 1).
We still have to calculate the Miller Effect Capacitance that may be a problem at high frequencies.
Gain is never free.
Tektronix, in the 1960s made a couple of models of vacuum tube probes.
They were cathode followers with bootstrapped rg.
The Miller Effect multiplier was (1 + < 1).
The lower grid resistor 100k for example instead of 1M shunts, may be more, the input, grid-plate+ grid-cathode capacitors.
Last edited: