A viewer sent me the parts to build this "color" pre-amp for my YouTube channel. In the past I did build one of the "universal" single tube pre-amps shown on this website, that uses a volume pot on the output, and it seems to work fine. The idea of this two tube version is to have the ability to drive a lower impedance SS amp with it.
This cathode follower one I am now looking at building is found here.
https://www.cascadetubes.com/the-12au7-cathode-follower-color-preamp/
I redrew his schematic to be easier to read, it looks like this.
My question is: what advantage or disadvantage would there be to moving the volume control to the front of the pre-amp like this. Is 100K a good pot to use for this or would a 250K work better?
I understand there are many other pre-amp designs that are more "HiFi" etc, so that's not the question, the viewer specifically ask me to build him one based on this design. I'm just curious about possibly moving the volume control to the front end of this design, and which position would work better (and why).
I'm also wondering if the 1.0uf cap paired with the 10K output resistance is why he is seeing the low frequency roll off starting at 50hz he talks about in the website article, and would something like a 4.7-6.8uf output coupling cap avoid that roll off.
Thanks in advance!
This cathode follower one I am now looking at building is found here.
https://www.cascadetubes.com/the-12au7-cathode-follower-color-preamp/
I redrew his schematic to be easier to read, it looks like this.
My question is: what advantage or disadvantage would there be to moving the volume control to the front of the pre-amp like this. Is 100K a good pot to use for this or would a 250K work better?
I understand there are many other pre-amp designs that are more "HiFi" etc, so that's not the question, the viewer specifically ask me to build him one based on this design. I'm just curious about possibly moving the volume control to the front end of this design, and which position would work better (and why).
I'm also wondering if the 1.0uf cap paired with the 10K output resistance is why he is seeing the low frequency roll off starting at 50hz he talks about in the website article, and would something like a 4.7-6.8uf output coupling cap avoid that roll off.
Thanks in advance!
It is not a good idea to put volume control at the output, as the signal into the amplifier is always high the distortion will be high and constant. A more severe problem is that the output impedance will change depending on the volume setting, so the frequency response will be dependent on the volume setting. "All" professional pre-amplifiers have an amplification stage after the volume control.
Like it is, for me definitely the one with pot at the input.
I suppose the idea behind the volume control at the output is to use a smaller value and decrease its distortion and improve insensitivity to frequency. However to have that pot at the output the preamp stage needs to be really good because it will have to take whatever comes in up to 2V rms at least and amplify it with very low distortion. Also output resistance will depend on the volume setting. It can be done but not with that simple circuit running at 2.4mA, IME. In the end it introduces a bigger problem than the one it's aiming to solve.....
For the second circuit I would change the 10K output with 470K-1 Meg and let the impedance of the power amp set the actual load. I tend to avoid coupling capacitors larger than 1 uF because that is usually the upper limit for lowest loss capacitors. The larger the value the more losses.
I would change the cathode resistor of the common cathode to 1K and the anode resistor to 39K. This should result in about 130-140V anode voltage and 4 mA+ anode current which is normally a more linear spot for the ECC82. DC coupling the cathode follower would only require 33K cathode resistor for about the same anode current, possibly slightly higher. This will get rid of 1 coupling cap and 2 resistors.
The input volume control doesn't have to be 100K at any cost, it could be 47K or even a nice 20K ALPS RK27
I suppose the idea behind the volume control at the output is to use a smaller value and decrease its distortion and improve insensitivity to frequency. However to have that pot at the output the preamp stage needs to be really good because it will have to take whatever comes in up to 2V rms at least and amplify it with very low distortion. Also output resistance will depend on the volume setting. It can be done but not with that simple circuit running at 2.4mA, IME. In the end it introduces a bigger problem than the one it's aiming to solve.....
For the second circuit I would change the 10K output with 470K-1 Meg and let the impedance of the power amp set the actual load. I tend to avoid coupling capacitors larger than 1 uF because that is usually the upper limit for lowest loss capacitors. The larger the value the more losses.
I would change the cathode resistor of the common cathode to 1K and the anode resistor to 39K. This should result in about 130-140V anode voltage and 4 mA+ anode current which is normally a more linear spot for the ECC82. DC coupling the cathode follower would only require 33K cathode resistor for about the same anode current, possibly slightly higher. This will get rid of 1 coupling cap and 2 resistors.
The input volume control doesn't have to be 100K at any cost, it could be 47K or even a nice 20K ALPS RK27
Last edited:
Using a pot of 500k at the output would be totally unworkable, and certain to audibly truncate the HF response.
If the input impedance of the power amp is 20K that's not surprising. In parallel with the 10K it would be 1 dB down around 50Hz....
SS amps have usually lower input impedance around 10-20K and less than 20K would be even worse.
That is why I suggest 1Meg. The 1 Meg resistor is just there as a dummy, basically.
"If the input impedance of the power amp is 20K that's not surprising. In parallel with the 10K it would be 1 dB down around 50Hz....
SS amps have usually lower input impedance around 10-20K and less than 20K would be even worse.
That is why I suggest 1Meg. The 1 Meg resistor is just there as a dummy, basically."
I see what you are saying, and it does make sense to just use something like a 1M on the output. And I think part of the idea of his design for this pre-amp was to not operate in the "best" spot, but to add a little color to the tone for use with an overly sterile sounding SS setup. Thanks for your input!
SS amps have usually lower input impedance around 10-20K and less than 20K would be even worse.
That is why I suggest 1Meg. The 1 Meg resistor is just there as a dummy, basically."
I see what you are saying, and it does make sense to just use something like a 1M on the output. And I think part of the idea of his design for this pre-amp was to not operate in the "best" spot, but to add a little color to the tone for use with an overly sterile sounding SS setup. Thanks for your input!
The Zout of the circuit is around 450 ohm.
At 1 kHz the reactance of 1 uF is 150 ohm that is in series of 450 ohm= 600 ohm total
At 20 Hz we are at 8400 ohm (7950 + 450) of Zout; too much.
So the pot at output must be high value to avoid loss.
With also some problem ( if not good quality) on tracking between channel.
With the volume at the input the results are always better even using a relative low value.
And , in every case the C in out I prefer to be relative high; in this case 4,7 uF, good quality.
Walter
At 1 kHz the reactance of 1 uF is 150 ohm that is in series of 450 ohm= 600 ohm total
At 20 Hz we are at 8400 ohm (7950 + 450) of Zout; too much.
So the pot at output must be high value to avoid loss.
With also some problem ( if not good quality) on tracking between channel.
With the volume at the input the results are always better even using a relative low value.
And , in every case the C in out I prefer to be relative high; in this case 4,7 uF, good quality.
Walter
After thinking though some of the responses and reading the OG webpage on this "Color" design, I think having the full amplification of the signal is intentional as they are looking for "the signal into the amplifier is always high, the distortion will be high and constant" and then bleeding off the extra gain on the output. Distortion that varies with the signal strength is not the goal of this pre-amp. So I think I will build it as is, and maybe play around with the output coupling cap size and see how that impacts the measured frequency response curve. I might toy around with the idea of placing the pot between the stages...
I know for many people super low distortion is the ideal, IMHO if that was what I was looking for, I'd just listen to a high quality SS amp and never mess with tubes 🙂
Thanks for the responses and for helping me clarify in my head what is happening here.
I know for many people super low distortion is the ideal, IMHO if that was what I was looking for, I'd just listen to a high quality SS amp and never mess with tubes 🙂
Thanks for the responses and for helping me clarify in my head what is happening here.
A high value pot on the output will give poor drive capability to the next stage, as well as creating greater roll off in the low pass filter when combined with the capacitance of the interconnect cable.
TANSTAAFL
TANSTAAFL
There's also a third option, to put the volume control in a feedback loop around the whole stage. Boegli popularized this in the early 1960s under the name "anode follower" and it was revived in modern context by B. Putzeys in Linear Audio.
All good fortune,
Chris
All good fortune,
Chris
I disagree on the sterile sound of SS amps. A sterile sounding SS amp is just a bad amp. As there can be bad tube amps.
I have just given back a nice Eamlab Studio 162 that I have borrowed from a friend of mines for a time. There is nothing sterile about its sound. It's actually quite warm sounding amp despite distortion is absent. I could happily live with it and forget about tubes. I don't think the magic of tube power amp is in the distortion but in the different way it drives the speakers. The valve preamp + SS amp (or hybrid power amp with tube voltage amp and SS output stage) are last in my preferences. That's my opinion, of course.
I have just given back a nice Eamlab Studio 162 that I have borrowed from a friend of mines for a time. There is nothing sterile about its sound. It's actually quite warm sounding amp despite distortion is absent. I could happily live with it and forget about tubes. I don't think the magic of tube power amp is in the distortion but in the different way it drives the speakers. The valve preamp + SS amp (or hybrid power amp with tube voltage amp and SS output stage) are last in my preferences. That's my opinion, of course.
Yet another option is to use the volume pot as the grid leak resistor.
Peter Walker used an active volume control in a preamp before the Quad QC22, in the 1950s.
Peter Walker used an active volume control in a preamp before the Quad QC22, in the 1950s.
Here a my Linea
https://milleniumaudiolab.blogspot.com/p/junior-linea.html?m=1
With a minimum of test lab that in most of the cases are missinig
Walter
https://milleniumaudiolab.blogspot.com/p/junior-linea.html?m=1
With a minimum of test lab that in most of the cases are missinig
Walter
A transformer or autoformer volume control on the output can work quite well. That's how I have my preamp/DAC setup, with 01A amplification in parafeed to Bent Audio remote control autoformer outputs.
With VC on the output linestage noise is reduced along with signal; with VC on the input the linestage noise is always 100%.
With VC on the output linestage noise is reduced along with signal; with VC on the input the linestage noise is always 100%.
But with 01A you have a low-medium gain with an Rp not so low so it is noisy itself
In addition the power supply for filaments and HT must be done almost perfect to kill the ripple.
Then you have to match carefully with autoformer about impedance
Have you tested it?
Walter
In addition the power supply for filaments and HT must be done almost perfect to kill the ripple.
Then you have to match carefully with autoformer about impedance
Have you tested it?
Walter
Built it, tested it, use it exclusively. Frequency response is very flat, with a slight bass boost as the volume control changes to lower volumes (-6dB to -24dB shown below).
Here's some shots of the preamp and its power supply. Mu-output cascoded CCS as the plate load. Built it with the 26 tube and then changed over to 01A (preferred by my ears). Rod Coleman filament regulators, 9V battery bias. That's a Buffalo IIISE DAC in the upper right hand corner of the preamp; with a switch on the back and a relay I can go from using it as a tube-output DAC to as a straight multi-input preamp.
I would suggest this pre-amplifier design is called "color" since the output spectrum will change depending on the volume level. The sound will be coloured.
I would certainly expect that at low volume levels, the low frequencies will roll off, no matter how large a coupling capacitor is made to be. At higher volumes, the bass may even be accentuated. You can easily simulate this in pspice or ltspice.
Your solution to move the volume pot to the input is the most natural solution.
Ian
Btw - I always design my outputs with careful consideration of the most likely input of the following stage.
This is what I might do:
The 5K grid stopper on the input seems excessive to me. I would go much lower. 1K is fine.
Then:
Add a grid stopper to the 2nd stage.
Not use a 33uF cathode bye-pass cap. I would go for 100uF - Why? because it will hopefully last longer. When the 33uF loses capacitance and drops to 22uF, you will hear the difference (loss of bass). When the 100uF drops to 80uF, you will not hear the difference.
The output grounding resistor should clearly be a larger value than 10K. I suggest at least 100K here. 220k would be fine.
On the other hand, if they like the way it 'sounds' and 'colours' the music, there is no helping that.
Since it is a 12au7 it is a fun exercise to look at the plate curves. 😉
The 1st valve is running only 2.42 mA. The load line is in the knees of the plate curves. Take a look:
But wait... what is that plate voltage? the plate resistor is 51k so how much voltage is dropped if only 2.42mA is running through it? Answer is about 123V.. This puts the plate at 300V-123V = 177V right? Yes, this should be true! So what is going on with this valve? 😉
I won't go into details, but this is NOT good if the intention of the designer is low distortion and somewhere close to linear response.
What if you want to make it perform a little more linear? Without a total re-design? Hmmmmm. The 12au7 plate curves are hard to work with since they're not very pretty. Plus we are limited by a few things such as the 300V supply too.
Quick (modest) improvement: Without replacing the 51k plate load, try running approx. 3.6mA using a 1.11k om cathode resistor (4V on the cathode) which should result in approx. 117V on the plate:
It's a little more linear and 4V headroom is still very good. Still, its a 12au7 so there is only so much you can do.
Btw. I didn't check the follower circuit.
Ian
Last edited: