I'm putting this in Instruments and Amps for no other reason than I'm designing a guitar amp, but the question is obviously fairly generic. No other forum seemed particularly appropriate for the question, otherwise.
In my first preamp stage, Ra = 220K and Rg of the following stage is 1M. That would normally give me a plate load impedance of 180K. But I have a 220K mixing resistor (I have 2 pre-amp channels) between the 1st stage plate and 2nd stage grid. Do I just add that to the 180K for a 400K plate load impedance?
In my first preamp stage, Ra = 220K and Rg of the following stage is 1M. That would normally give me a plate load impedance of 180K. But I have a 220K mixing resistor (I have 2 pre-amp channels) between the 1st stage plate and 2nd stage grid. Do I just add that to the 180K for a 400K plate load impedance?
If I have understood the circuit you are describing, you would add the mixing resistor in series with the Rg of the following stage, which gives 1220K. Then put that in parallel with Ra, giving an AC load of 186K.
If this is the "classical" musical instrument amp I am thinking of...
The 1 Meg is Volume Controls for mixer...
Since you mention "mixer" I am assuming 2 channels here...
The loading will vary according to where the volume control is set to....
For analysis sake, assume the unused channel to be OFF and turned down to 0 ....
If the volume control is at full 10...then you have 220K // 1Meg // (mixer)....
Then this feeds into your mixer ...which is two 220K resistors in series...making a -6dB voltage divider ... This is 440K ....the middle of the this mixer/divider goes into Hi-Z input grid of 2nd stage valve...
SO you now have 220K // 1Meg // 440K = 128K for the loading on this valve at full volume...
When the volume is at 0 the loading is 220K// 1Meg = 180k
So your loading will vary between 180K to 128K depending on the position of volume control....assuming other unused channel is turned down to 0 ....
This is refered to as Mid-Band Gain... For full frequency domain..you would need a transfer function that included the capacitance of the tubes and coupling cap..
If you are trying to calculate the coupling capacitor value for -3dB roll off...you would need to include the plate resistance of 1st stage... this varies but can use 60K for 12AX7 that is cathode by-passed .......60k // 220k = 47K for your SOURCE resistance..
Then the load would be 1Meg // 440K = 305K ...
Put the Source and Load in series to make 352K ...
If your cap is .022uF then your roll-off will occur at 1/(6.28* RC) = 20.52 Hz ...
This is -3dB @ 45 degrees lag...
The 1 Meg is Volume Controls for mixer...
Since you mention "mixer" I am assuming 2 channels here...
The loading will vary according to where the volume control is set to....
For analysis sake, assume the unused channel to be OFF and turned down to 0 ....
If the volume control is at full 10...then you have 220K // 1Meg // (mixer)....
Then this feeds into your mixer ...which is two 220K resistors in series...making a -6dB voltage divider ... This is 440K ....the middle of the this mixer/divider goes into Hi-Z input grid of 2nd stage valve...
SO you now have 220K // 1Meg // 440K = 128K for the loading on this valve at full volume...
When the volume is at 0 the loading is 220K// 1Meg = 180k
So your loading will vary between 180K to 128K depending on the position of volume control....assuming other unused channel is turned down to 0 ....
This is refered to as Mid-Band Gain... For full frequency domain..you would need a transfer function that included the capacitance of the tubes and coupling cap..
If you are trying to calculate the coupling capacitor value for -3dB roll off...you would need to include the plate resistance of 1st stage... this varies but can use 60K for 12AX7 that is cathode by-passed .......60k // 220k = 47K for your SOURCE resistance..
Then the load would be 1Meg // 440K = 305K ...
Put the Source and Load in series to make 352K ...
If your cap is .022uF then your roll-off will occur at 1/(6.28* RC) = 20.52 Hz ...
This is -3dB @ 45 degrees lag...
Last edited:
Just to clarify, don't you mean RL as in plate load resistor?
ra, with lower case r represents the resistance of the plate itself and depends on the operating point, but for a 12AX7 is in the 60K range.🙂
RL is in parallel with ra as well as Rg of the following stage.
Your description is a little vague and a diagram would be helpful.
A single 220K in series with the next stage 1M Rg would not change the plate load very much because it "sees" it in series with the 1M, but it would form a voltage divider of the signal to the grid of the next stage...220K series 1M shunt. A slight attenuation.
but if you are saying that you have 2 triodes feeding 1 input through two 220K mixing resistors......
then the more significant effect is the voltage divider formed by the two 220K mixers at the input of the next stage, which will attenuate the signal by half [6db] because the signal "sees" ground through the other 220K and the ra and RL of the other tube.
That's the penalty of passive mixing.
Yes, there must be another 220K mixing resistor connecting into the second stage grid (from the other pre-amp channel). So we need to find the resistance to AC ground via that other 220K and put the resulting value in parallel with the 1M Rg of the second stage.
These passive mixers used in guitar amps are "bloody awful" when using both channels simultaneously.... The 1st stage valve gets loaded down to where the intermod gets bad..
I typically assume one channel turned off when doing the analysis and also do this in actual practice as well...
I typically assume one channel turned off when doing the analysis and also do this in actual practice as well...
Sorry, yes, I was a complete idiot. It was perfectly obvious as soon as he said it because the 220K is in series with the 1M. The two channel thing messed me up.
Cerrem, actually the volume is a bit down the road from there. The 1st stage pre-amp feeds a 2nd stage which feeds a cathode follower which then goes to the tone stack and then volume. Which then goes to a 3rd preamp stage, master volume, and then power amp. Hmmm. I might want a gain control in there...
LPMark: Well, yes and no. When I said Ra, I was talking about the Anode resistor which is 220K. The mixing resistor, which I guess is also RL would be 220K as well. Rg of the succeeding stage is the 1M. Sorry for the confusion.
And yes, I now understand that 6db loss (I knew it was there, but I didn't understand why. Thanks for that explanation. It also pointed out a flaw in my schematic. I might want to rethink the 2 channel thing...
I'm definitely more attached to one channel than the other... 🙂
Cerrem, actually the volume is a bit down the road from there. The 1st stage pre-amp feeds a 2nd stage which feeds a cathode follower which then goes to the tone stack and then volume. Which then goes to a 3rd preamp stage, master volume, and then power amp. Hmmm. I might want a gain control in there...
LPMark: Well, yes and no. When I said Ra, I was talking about the Anode resistor which is 220K. The mixing resistor, which I guess is also RL would be 220K as well. Rg of the succeeding stage is the 1M. Sorry for the confusion.
And yes, I now understand that 6db loss (I knew it was there, but I didn't understand why. Thanks for that explanation. It also pointed out a flaw in my schematic. I might want to rethink the 2 channel thing...
I'm definitely more attached to one channel than the other... 🙂
Cerrem, I didn't think of that. I could just totally remove the channel via switch instead of doing passive mixing. It would actually simplify the analysis and more importantly, it might make the amp not suck.
That's correct, as far as the preceding stage is concerned.
To be clear, only the 220K resistor connected to the HT supply is considered RL, ........... the mixing resistor would be treated as a series resistance to the signal path and would have little effect if it weren't for the connection to the other 220K mixing resistor and what's attached to it.
If you end up using only one channel then of course remove the mixing resistor.
If the unused section is on the same tube its easy to run them in parallel,
though some may argue that has no benefit, I feel like it sounds just a little fatter and gives just a tiny bit more gain.
220K RL is pretty high for 12AX7. Trying to squeeze out as much gain and not in a very linear region, and you say the amp sucks.
If you could put up a schematic with values and voltages someone could probably help more.
Ah, I guess people just assumed a 12AX7. I didn't even think to correct them. One channel is a 12AX7, but that one only has a 100K plate resistor. The other is the pentode side of a 6U8. That's the one with the 220K plate resistor.
The schematic is posted in this thread: http://www.diyaudio.com/forums/instruments-amps/273487-cobalt-twin-amp-design.html
I posted this question first, but then decided to post the schematic and start really trying to fix it up. I've already gotten a lot of good comments, so I need to go back and fix some stuff up, so understand that the schematic is a very early draft. Obviously I'm still trying to firm up even some of the basic architecture (like how I want to handle the two channels). And though I have reverb and tremolo in the circuit, they're very low priority for me right now. My plan is to initially build the amp without them and add them after I have a working amp (trying not to make things too complicated). But I added them to the schematic because I want to make sure I can add them in when I'm ready to.
The schematic is posted in this thread: http://www.diyaudio.com/forums/instruments-amps/273487-cobalt-twin-amp-design.html
I posted this question first, but then decided to post the schematic and start really trying to fix it up. I've already gotten a lot of good comments, so I need to go back and fix some stuff up, so understand that the schematic is a very early draft. Obviously I'm still trying to firm up even some of the basic architecture (like how I want to handle the two channels). And though I have reverb and tremolo in the circuit, they're very low priority for me right now. My plan is to initially build the amp without them and add them after I have a working amp (trying not to make things too complicated). But I added them to the schematic because I want to make sure I can add them in when I'm ready to.
- Status
- Not open for further replies.