How to measure the output impedance of a valve preamp?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
AX tech editor
Joined 2002
Paid Member
Not sure the OP means to straight load a triode stage output with a low value resistor to lower its output impedance, the proper way is to use a voltage divider so tube plate only "sees" the upper branch of the divider.

IF he uses, say, a 100k:1k divider plate will "see" 100k but divider output impedance will be roughly 1k .
Of course, there´s a large attenuation involved.

At least, that´s what he did in his schematic, where tube drives a tone stack and volume control and then a 3k3 resistor.

This is a bad idea because you throw away a lot of gain, so you need lots of gain to throw away which means lots of distortion.

The way to do it is design for lowish Zout, or add a cathode follower. And a power amp with 10k Zin is not a low input impedance, so no need for heroic measures at the sending end. If you feed a 10k Zin from a 5k Zout you lose only 3dB. That's OK.

Jan
 
This is a bad idea because you throw away a lot of gain, so you need lots of gain to throw away which means lots of distortion.
Well, he IS looking for distortion, it´s a Guitar pedal.
And yes, he uses tons of gain precisely to overdrive a tube stage.
After which he´ll have some 90V RMS "clean" signal or some 250Vpp ... in any case he *must* attenuate that signal to drive anything.
The way to do it is design for lowish Zout, or add a cathode follower.
After which he´ll still have the 250Vpp signal which must be attenuated anyway.


Quote:
Ideally I would need an attenuation 100:1while reducing the output impedance at the same time. So, a full voltage divider is a proper way to do that, right? Where would you suggest to place it - before or after the master volume? Thanks.
We are on post #40 by now, lots of water passed under the bridge, please re-confirm what preamp will you use and what load will you drive.
We need input impedance and signal level required.
My point being that probably the Master Volume will be part of the attenuator and not 1M any more.
 
Last edited:
AX tech editor
Joined 2002
Paid Member
Well, he IS looking for distortion, it´s a Guitar pedal.
And yes, he uses tons of gain precisely to overdrive a tube stage.
After which he´ll have some 90V RMS "clean" signal or some 250Vpp ... in any case he *must* attenuate that signal to drive anything.

OK, yes, that'll change things ...
Funny guys, these guitar people ;-)

Jan
 
Funny guys, these guitar people ;-)
Jan
I was interested in electronics (mostly radio) as a child, then in electronics (mostly Hi-Fi) as a teenager and into my twenties. In my twenties I also taught myself to play acoustic guitar, then electric guitar, and finally, married and with a mortgage, I got serious about guitar playing in my forties.

So now I am one of those "guitar people", but with almost a life-long background in "proper" analogue electronics, particularly audio.

It was a heck of an unlearning curve, I can tell you! Almost everything I knew about audio electronics no longer applied where electric guitar was concerned. Low distortion sounds bad. Opamps sound bad. Speakers with a flat frequency response sound bad. Tweeters sound bad. Low output impedance (to the speaker) sounds bad. :D

Because I had the technical background, I could, with time, tell the B.S. from the reality. All the things I listed in the previous paragraph really are engineering and scientific reality when electric guitar amplification is involved - there really are good technical reasons for them!

(For example, opamps sound bad because with guitar, they will be driven into clipping. Because opamps are designed with near-infinite voltage gain, when used for audio amplification, they are always set up with large amounts of negative feedback. And this means that when clipping does occur, it is harsh and very abrupt. So as you turn up the level, an e-guitar through an op-amp goes from too-clean and boring to too-harsh and nasty, with nothing in between!

As has been said before, we really shouldn't think of a guitar amplifier as an audio amplifier. We should think of it as a part of the instrument - it is a signal processor, designed to take an instrument that generates rather boring sounds on its own, and modify the signal so much that what comes out is almost unrelated to what went it.


-Gnobuddy
 
To give some practical example: this is a killer early Marshall Guitar amplifier ,18W, **obviously** overdriven, no need for an analyzer to detect that:
YouTube

Please listen to it first, only then click the link below to see its waveform when overdriven.

Please do not cheat so as not to be prejudiced ;)

https://robrobinette.com/images/Guitar/Overdrive/HeavyOverdriveScope.jpg

also same but somewhat *more* overdriven at bottom:

https://www.paulamps.com/OutputsNoBuzzCap.jpg

another example, here bottom waveform is measured across speaker terminals, top one is measured at a tube *before* the power ones, which is also clipping on its own, go figure:

https://www.paulamps.com/CrossOverSolution.jpg
 
AX tech editor
Joined 2002
Paid Member
I was interested in electronics (mostly radio) as a child, then in electronics (mostly Hi-Fi) as a teenager and into my twenties. In my twenties I also taught myself to play acoustic guitar, then electric guitar, and finally, married and with a mortgage, I got serious about guitar playing in my forties.

So now I am one of those "guitar people", but with almost a life-long background in "proper" analogue electronics, particularly audio.

It was a heck of an unlearning curve, I can tell you! Almost everything I knew about audio electronics no longer applied where electric guitar was concerned. Low distortion sounds bad. Opamps sound bad. Speakers with a flat frequency response sound bad. Tweeters sound bad. Low output impedance (to the speaker) sounds bad. :D

Because I had the technical background, I could, with time, tell the B.S. from the reality. All the things I listed in the previous paragraph really are engineering and scientific reality when electric guitar amplification is involved - there really are good technical reasons for them!

(For example, opamps sound bad because with guitar, they will be driven into clipping. Because opamps are designed with near-infinite voltage gain, when used for audio amplification, they are always set up with large amounts of negative feedback. And this means that when clipping does occur, it is harsh and very abrupt. So as you turn up the level, an e-guitar through an op-amp goes from too-clean and boring to too-harsh and nasty, with nothing in between!

As has been said before, we really shouldn't think of a guitar amplifier as an audio amplifier. We should think of it as a part of the instrument - it is a signal processor, designed to take an instrument that generates rather boring sounds on its own, and modify the signal so much that what comes out is almost unrelated to what went it.


-Gnobuddy

Thanks for the explain. Fascinating!

Jan
 
We are on post #40 by now, lots of water passed under the bridge, please re-confirm what preamp will you use and what load will you drive.
We need input impedance and signal level required.
My point being that probably the Master Volume will be part of the attenuator and not 1M any more.
I would like to make the output compatible with as many devices as possible. So the preamp could be connected to the front of guitar amp, to the fx return input of an amp, line level audio input or daisy chained with another guitar pedals. So I would guess the input impedance of the load would be >10K, the signal level in the range of -10dBv to +4dBu. Please see the schematics of my last two gain stages and the output. As I mentioned, I would like to attenuate the signal, say 100:1, and get a lower output impedance.
So, with the master volume pot being used as a part of the voltage divider, where would you place the rest of the voltage divider circuit - before or after the pot?
 

Attachments

  • tube preamp.png
    tube preamp.png
    21 KB · Views: 64
Last edited:
AX tech editor
Joined 2002
Paid Member
I would like to make the output compatible with as many devices as possible. So the preamp could be connected to the front of guitar amp, to the fx return input of an amp, line level audio input or daisy chained with another guitar pedals. So I would guess the input impedance of the load would be >10K, the signal level in the range of -10dBv to +4dBu. Please see the schematics of my last two gain stages and the output. As I mentioned, I would like to attenuate the signal, say 100:1, and get a lower output impedance.
So, with the master volume pot being used as a part of the voltage divider, where would you place the rest of the voltage divider circuit - before or after the pot?

With this scheme, the Zout depends on the setting of the last volume control.

Remember, Zout is what you see when you look back into the output. With the control all the way down, you see 250k. With the control all the way up, you see 3.3k in parallel with 250k. With the control in mid position, you see 125k plus 125k in parallel with 3.3 is about 132k.

Jan
 
Last edited:

PRR

Member
Joined 2003
Paid Member
I think the switch is not corrected correctly. The label on the switch connection CH2_Vol should be CH1_Vol.
Also, the Zout is dependent on the pot setting...

I assume there is a CH2, not shown in the image I cribbed; and that what is shown is CH1. In any case, that's his problem (flip the switch until something happens).

I clearly noted "Zout 0 to 3K". Did not say "depends on knob setting", no. However for the intended uses, it hardly matters; anything under 10K is "low" in guitar-cord work. (Typical input impedances are far over 100K, over 500K for best sound straight from a guitar.)
 
This is what I suggested when I said "the pot will be part of the attenutor".
With R22=100k you´ll have 10:1 attenuation, good to drive a power amp
With R22=1M you´ll have 100:1 attenuation, which is Guitar/pedal level.

Output impedance will vary from 0 to 10k according to Master Volume setting, which is fine.

Pick desired R22 depending on your needs.
You may even make it switch selectable: "Guitar/Line level"
 

Attachments

  • TubePre 01.jpg
    TubePre 01.jpg
    72.6 KB · Views: 54
I have one suggestion regarding the channel select switch. Currently it's wired in series with the signal. Feeding into a 100k or 1M impedance, even a few pF of stray capacitance across the switch will "leak" signal, so that "offness" will be poor - there may be audible cross-talk from the channel that's supposed to be switched off.

I suggest wiring for shunt (parallel) switching instead:

1) Add one more 100k resistor from the top of the Master Vol pot to the wiper of the channel 2 volume pot.

2) Wire the switch pole (moving contact) of the channel select SPDT switch to ground, one throw to wiper of CH1_VOL, the other throw to the wiper of CH2_VOl.

The output of one channel or the other (from the individual channel volume pot wipers) is now shorted directly to ground. The 100k resistors allow the other channel signal to get through to the master volume pot. Offness will be better this way.


-Gnobuddy
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.