1. The Rules have been updated regarding posting as a business on TGP. Thread with details here: Thread Here
    Dismiss Notice

PI plate resistor trimmers?

Discussion in 'Amps/Cabs Tech Corner: Amplifier, Cab & Speakers' started by Normster, Nov 17, 2005.

  1. Normster

    Normster Member

    Messages:
    921
    Joined:
    Oct 14, 2004
    Location:
    Sacramento, CA
    Some of the Dumble schematics I've seen include a trimmer between the PI plate resistors. According to info on Aiken's site, "When only one signal input is used (ignoring feedback inputs) R1 is usually made 10% - 20% lower than R2 to compensate the unbalanced gains of the two tube sections and make the two output amplitudes equal."

    So...my question is, does this make an audible difference in the tone? If so, how would I go about adjusting the trimmer if I don't have a scope to measure amplitude? (Plate voltage? Best tone? Sustain?)

    Thanks in advance for any advice.
     
    Roe likes this.
  2. Fuchsaudio

    Fuchsaudio Member

    Messages:
    7,492
    Joined:
    Jan 3, 2002
    Location:
    Bloomfield, New Jersey
    In a perfect world, both tube halfs would have identical gain and balance. In the real world both AC and DC balance on a tube is never the same unless the tube has matched halfs. The trimmer is an audiophile type circuit tweak that HAD carried over from hi-fi amps. It's one of the subtle ways he improved the original Fender Schmitt style phase splitter circuit.

    Okay, so you can set the trimmer for DC balance (measure pin 1 to pin 6 for '0' volts between them), but that rarely is the same point at which AC signal balance occurs.

    BUT if you look at the amp when it's at soft-clipping, you can actually adjust the trimmer and see it make the clipping become equal on both waveform halfs. You can hear this, especially if one side of the wave is hard clipping before the other starts to.

    Now, if you can't hear it it's another story (if you don't believe in Santa, it doesn't mean he isn't there Tele Man), but it does make a (subtle) difference in "feel" and note swell etc. ;)

    Ask ol' dog ears......he sets his by ear, my scope confirmed he was right (I hate that).
     
  3. Normster

    Normster Member

    Messages:
    921
    Joined:
    Oct 14, 2004
    Location:
    Sacramento, CA
    Thanks for the response, guys. I guess the real test is just to add a trimmer to my circuit and see what happens. ;)
     
  4. trdlasvegas

    trdlasvegas Member

    Messages:
    167
    Joined:
    Aug 30, 2005
    Location:
    Las Vegas
    I've always wondered about this. No matter how much I read about the PI section, I still am not satisfied.

    First off the purpose is to deliver both a in-phase and out-of-phase signal to the power tubes right. So you would think in a perfect world these two signals should be EXACTLY the same amplitude but just out of phase from each other.

    So why did Fender and others purposely make these sections not equal? (different value plate resistors)

    Now GT (and others) makes a matched Phase Inverter tube where they measure the gain to be the same between sections of the tube. So isn't this going against what Fender was trying not to do?

    And now your talking about triming the DC plate voltage to be exactly the same. You guys have me all confused.

    Is this a conflict of interest or just different views on how to operate the Phase Inverter section?

    -Tony
     
  5. scottl

    scottl Member

    Messages:
    17,041
    Joined:
    Apr 11, 2003
    Location:
    Cherry Hill, NJ
    Tony,

    The plates need to be different in order to balance the circuit. The NF returns into the second plate of the PI. The balanced PI tube is a good idea IMO since you can set the PI balance with your scope and then just replace the PI tube with another balanced one and probably be close.

    Fwiw, I can ALWAYS hear the difference when I adjust the trimmer in my PI section. As Andy said, I brought 3 amps to his shop, all adjusted by ear, all three were spot on when scoped.......

    .....ole dog ears
     
  6. trdlasvegas

    trdlasvegas Member

    Messages:
    167
    Joined:
    Aug 30, 2005
    Location:
    Las Vegas
    OK now you got me going. Of all the books and articles I read about the PI circuit, no one ever mentioned the NF line. This does make some sense.

    So now I'm searching around for a schematic that uses this PI trimmer you guys are talking about. I can't find any. So are we saying maybe a 100k trim pot between the two plates with the wiper going through another 100k to ground?

    I'm real excited to give this a try.

    -Tony
     
  7. Normster

    Normster Member

    Messages:
    921
    Joined:
    Oct 14, 2004
    Location:
    Sacramento, CA
    It's actually set up as a voltage divider with each leg feeding the input to one plate resistor and the wiper going to B+. Although the schematic shows a 10K trimmer, the few guys I've talked to seem to recommend 20K.
    [​IMG]
     
  8. Fuchsaudio

    Fuchsaudio Member

    Messages:
    7,492
    Joined:
    Jan 3, 2002
    Location:
    Bloomfield, New Jersey
    "in most cases, the out-of-phase output plate resistor will have to be made around 10% smaller than the in-phase output to achieve perfect balance between the two outputs. However, if the second input is to be used, as a second channel input, for example, the resistors should be made equal, as it will only be balanced for one input, and the other input will be unbalanced even more".

    From Me:

    Using two equal plate loads and the trimmer resistors (we use 20-K for maximal range) the resistors can be equal in value. Typically 90-K to 110-K. With a 10-K trimmer you can reduce one side slightly (82-K/100-k) and usually hit balance easily.

    The comment about DC balance was to point out that although you can adjust for DC balance (where the plate voltages are equal) but it does not occur at the same point where AC balance occurs. The pot will allow you to set for AC or DC balance, you want to set for AC balance and not pay any attention to the DC balance. Sorry if it confused anyone.
     
  9. reaiken

    reaiken Member

    Messages:
    1,796
    Joined:
    Jan 4, 2002
    Location:
    Greenwood, SC
    Another point:

    If nearly perfect equal-value voltage swings are desired at the plate, you need to use a current source (actually a current "sink") in place of the "tail" resistor and equal value plate load resistors.

    The next best thing is to increase the value of the tail resistor, but this limits the headroom of the stage. When using both PI inputs, it is preferable to use a larger tail resistor and make both plate resistors equal. This is why Vox type amps that have large tail resistors have equal value plate loads, and Fender and Marshall type phase inverters have small tail resistors and unequal plate load values.

    If you have a smaller value tail resistor, deliberately unbalancing one of the resistors to balance the outputs, one input side will be perfectly balanced, but the other input side will be way out of balance. This is not so important when using global negative feedback and only one of the inputs, but extremely important when using both inputs.

    Randall Aiken
     
  10. trdlasvegas

    trdlasvegas Member

    Messages:
    167
    Joined:
    Aug 30, 2005
    Location:
    Las Vegas
    Well I thought that I was following this fine but Randall you have lost me. I think it just has to do with the terms that you guys are using. Let me see if I understand this.

    Andy, in your post you quoted Randall as saying:
    "However, if the second input is to be used, as a second channel input, for example, the resistors should be made equal, as it will only be balanced for one input, and the other input will be unbalanced even more"

    I'm thinking that you actually mean the second input or the other side of the phase inverter, since I cannot see what differerence the second channel of the amp would make.

    And then Randall, in your post you said:
    "...deliberately unbalancing one of the resistors to balance the outputs, one input side will be perfectly balanced, but the other input side will be way out of balance. This is not so important when using global negative feedback and only one of the inputs, but extremely important when using both inputs."

    So again you must be talking about the two inputs of the phase inverter, right?

    So what I think your saying is that on a Vox amp because there is no negative feedback, one side of the phase inverter is connected to ground and the "tail resistor", which I think must be the resistor from the two cathodes is deliberatly made larger while the plate resistors are matched.
    But on a Fender, one side of the phase inverter is being fed from the negative feedback wire so the tail resistor is made small and the plate resistors are purposly unbalanced.

    So with all that said, where I'm especially confused is Andy says you can get equal value voltage swings at the outputs of the phase inverter on a negative feedback amp by balancing the DC voltage at the plate resistors.

    And Randall your saying you can't do that without having a current "sink" in place of the tail resistor. Which comes to your last statement:
    "If you have a smaller value tail resistor, deliberately unbalancing one of the resistors to balance the outputs, one input side will be perfectly balanced, but the other input side will be way out of balance. This is not so important when using global negative feedback and only one of the inputs, but extremely important when using both inputs."

    Isn't that exactly what we are trying to do. Use both side of the phase inverter, one side for the signal, one side for the negative feedback. Keep the tail resistor as small as possible to give us the maximum voltage swing between plate and cathode. And finnally balance the phase inverter outputs to exactly the same but out of phase with each other?

    -Tony

    Andy, Normster, and Randall,
    I really appreciate your time in this discussion, as I have been eagerly trying to understand this phase inverter section for quite awhile now. I've learned more in these few posts than all the textbooks I've read on the subject.
     
  11. reaiken

    reaiken Member

    Messages:
    1,796
    Joined:
    Jan 4, 2002
    Location:
    Greenwood, SC
    Yes, Tony, when I say "second input" I mean the second grid of the phase inverter, which can be used as a feedback input or as a second channel input.

    Ideally, you would want balanced outputs whether or not you are using the PI as a feedback summing junction or as a second input. However, as previously mentioned, you run into the problem of reduced headroom if you make the tail resistor large enough to act as a "pseudo-current source". This is more of a problem wihen driving EL34s than EL84s, because they require a much larger voltage swing at the grid to drive them to full output (especially in a 100W amp).

    Feedback has a way of balancing itself out, if you apply enough of it. It will twist itself up and generate a strange enough looking waveform to make the output appear as distortionless as possble, provided the output stage is not in clipping and there is enough forward gain and slew rate to accomplish this.

    Having said that, there is generally not that much feedback in a typical guitar amp, and the output stages are usually driven to clipping, so we want to make the PI as balanced as possible, yet still accomodate the increased headroom requirements of the EL34 drive.

    This is accomplished by utilizing the "third" input of the long-tail pair - the bottom of the "tail". If you look at a standard Marshall or Fender PI, you'll see that the negative feedback is run through a voltage divider (typically the 100K resistor/5K presence pot on a Marshall), and then coupled into the second input through a DC blocking capacitor. You will also notice that this signal is also injected into the tail, which adds in phase with both outputs. Since the feedback signal is 180 degrees out of phase with the main input plate signal, this added signal makes the first plate signal smaller, while adding to the other plate signal, making it larger. This effectively helps balance the outputs while using a smaller tail resistor, but the main function is to better balance both outputs when driven from the feedback side.

    You could just ground the tail resistor and run the junction of the 100K/5K pot to the second input blocking capacitor, but you would find the outputs to be not very well balanced (the balance will be particularly bad when driving the feedback input, less so when driving the main input) unless you used a large tail resistor, which would reduce your headroom.

    This is all discussed in this paper:

    http://www.aikenamps.com/LongTailPair.htm

    Randall Aiken
     
    Roe likes this.

Share This Page