Questions, coupling Bassman pre to 6AK6 parallel

Discussion in 'Shock Brother's DIY Amps' started by mountainhick, Oct 9, 2021.

  1. mountainhick

    mountainhick Tele-Meister

    Age:
    64
    Posts:
    211
    Joined:
    May 2, 2021
    Location:
    Rocky Mountains
    The preamp is unaltered 6G6B bassman. (I may play with tone, coupling and cathode values depending on what I get from the stock design after trying it)

    The power section is a take off from the "Littlerock", 2 6AK6 in parallel.

    Questions:

    Connecting preamp to power amp- What to do for each of these:

    1-The 470K resistor is from the 6G6B, a mixing resistor that had a partner from the "normal" preamp channel. I am only using the bass channel. Grid stoppers are added for the power tubes, I am assuming as demixers to isolate grids*. Are all of these necessary?

    *(the Littlerock designer extolled the virtues of the power tubes not being coupled so it is more tolerant of non matched tubes)

    2- Coupling cap following the 470K is
    - .0005uf on one PI grid, 0.1uf on the other on the Bassman,
    - .022uf on the Littlerock
    Seems .022 is a reasonable choice, but I am not certain.

    3-(Side question, is the imbalance for the bassman PI to create more wave asymmetry for harmonics??)


    Cathode values:

    4-standard values for single ended, single tube 6AK6 is 560 ohm, and 22uf cap. The Littlerock designer used 510ohm and 100uf. Am I correct that this will allow more bass? Perhaps get muddy?

    5-The grid leak resistor covers both power tubes. It is 220K for the Littlerock. In other SE single tube designs this value varies from a 1meg volume pot to a 1meg pot followed by 260K, to 470 ohm, all to ground, and some are bypassed with a cap. I don't understand the possible trade offs here. Can someone fill me in?

    EDIT: Screen resistors are 1K

    amp-sch-lowres.jpg
     
    Last edited: Oct 9, 2021
  2. andrewRneumann

    andrewRneumann Tele-Afflicted

    Posts:
    1,067
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    The 470K resistor behaves as a voltage divider (attenuator) with the grid leak resistance of the following stage. I would guess you would need to set this value to get the right mix of power amp distortion before the preamp distorts. If I was you, I would put it after the coupling cap and probably use a 1M linear pot wired as a variable resistor for testing. I would think goal is to get the power amp to distort before the preamp or else you will lose a lot of the SE juiciness.

    The .022 coupling cap is a reasonable guess, but again, set yourself up for the inevitable desire to experiment. Start with a small value soldered in and additional capacitance temporarily clipped in parallel.

    Imbalanced plate resistors in LTP phase inverters were originally selected to balance the signal properly. But in reality, the values might cause more imbalance than balance if they are just blindly copied from our circuit to the next. If the designer is going for balance, plate load resistors should be selected based on a specific analysis on that circuit.

    I would keep the cathode bypass caps large and address muddiness earlier in the circuit. OMMV

    The grid leak resistor should be at most half the data sheet maximum for cathode bias. You have two tubes “leaking” so the resistance needs to be half.
     
    mountainhick likes this.
  3. mountainhick

    mountainhick Tele-Meister

    Age:
    64
    Posts:
    211
    Joined:
    May 2, 2021
    Location:
    Rocky Mountains
    Hmm, I thought the sequence of mixing resistor before coupling cap would negate the voltage divider aspect of that resistor with the grid leak resistor. Rather, in the diagram above you have a CR filter with the coupling cap and grid leak resistor, and the 470K is just a lone attenuator.

    My bad regarding what I at first saw as two coupling caps in the bassman's circuit. (see pic below) I see I misinterpreted the circuit's transition to PI here. The .0005uf is the coupling cap, but the 0.1uf is not a coupler from the preamp, rather functioning in the feedback/presence circuit. Still there is a CR filter after the mixing resistors by virtue of the coupling cap and the 1meg resistor onward into the network. I don't see a simple voltage divider in this, rather a circuitous arrangement with the presence network, then ground. I am quite lost in this circuit leading to the PI grids, but that's pretty moot to what I need to do with the SE circuit at hand.

    So It seems that the voltage divider you refer to isn't actually implemented as such in either circuit, but it could be a valuable tool in adjusting input signal to the power tubes.

    Unless, Is a CR arrangement also a voltage divider?

    Should i perhaps look at this as the 470K just a redundancy as a mixer, and I should start over regarding that voltage divider you speak of?

    So in the end of all that mental gyration, adding different value resistors after the coupling cap before the grid leak resistor just drops grid input voltage that much more?


    6G6B schematic picking up from the 470K. It's other channel's 470K mixer is out of the picture.

    6G6B-pre-PI.jpg

    Agree on that plan!


    Got it, thanks. Again i misread that aspect of the circuit. It isn't simply the .0005 vs .1uf, those caps are integrated into the grid and feedback network. The issue is moot for the parallel pair anyway, was just curious.


    OK. I'll start there. The original schematic i started from for a single tube SE 6AK6 design has the 25uf... it also had a 5" speaker. Am i correct that perhaps that 25uf was appropriate for the smaller speaker with lesser bass handling?


    This is a bit beyond me. i understand the halving aspect, but don't know how to find or calculate the "maximum for cathode bias". It is not referred to as such on the data sheet. I'll just go with the 220K in the Littlerock schematic.



    Any thoughts about grid stopper values? The Littlerock has them at 3.9K each. How would adjusting this value change things?


    Updated schematic with values added:

    amp-sch-lowres-update1.jpg
     
    Last edited: Oct 10, 2021
  4. andrewRneumann

    andrewRneumann Tele-Afflicted

    Posts:
    1,067
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    Whew... lots of questions and good ones too. I will try my best.

    When a resistor is in series with a capacitor, they act in concert to form an impedance. The 470K impedes current at all frequencies and the capacitor impedes current dependent on frequency. Their sum can be considered the "upper leg" of a voltage divider with the grid leak resistor.

    Yeah, don't worry about the PI. The so-called mixing resistor is only required because two channels are joined up. Without those mixing resistors one channel would send signal current out through the tube of the other channel and not build up very much voltage there where they join.

    As an aside--this is why I prefer to not think of a signal as flowing as one smooth river from input jack to power tube plate. All interconnected pieces of the circuit "vibrate" together under signal--there isn't really a direction to it. When you start to think in terms of current instead of voltage, this stuff all of sudden makes more sense too.

    There is a voltage divider behind every corner of a circuit. It just takes practice to see them.

    YOU JUST SAW ONE! :D Yes it is a frequency dependent voltage divider.

    Yes. Put a resistor behind that coupling cap and adjust it so you get just the right amount of signal hitting the grid of the power tubes. The coupling cap still functions as a CR filter and may need to be adjusted. If you like high gain sounds, usually a downward adjustment is required to cut bass and reduce blocking distortion. If you like clean, warm sounds, a large coupling cap with do.

    Yes, I think you are getting the idea quite nicely.

    I'm not sure on this one. Punt to smarter people. :lol:

    I just looked up the Tung-Sol data sheet and it specifies 500K maximum for a single tube cathode biased. Half that for two tubes is 250K. Your small of amount of grid stopper resistance counts too, so 220K looks pretty much by the book.

    My ear tells me this: very little. Others have their opinions. The way I see it, with a pentode, this value won't affect tone nearly a lick. Just need enough resistance to keep nasty oscillations away. 1K or 3K9 are probably both equally suitable.
     
    mountainhick likes this.
  5. mountainhick

    mountainhick Tele-Meister

    Age:
    64
    Posts:
    211
    Joined:
    May 2, 2021
    Location:
    Rocky Mountains
    @andrewRneumann, Thank you! I get it.


    mountainhick said:
    Any thoughts about grid stopper values? The Littlerock has them at 3.9K each. How would adjusting this value change things?​

    Impossible for me too to interpret how much difference there would be at the speaker and pushing the power tubes with a 3.9 fold difference. Guess I'll just go with the 3.9 K for now.
     
  6. dan40

    dan40 Friend of Leo's

    Posts:
    2,365
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    I doubt your ear could hear any difference between a 1k and 3.9k grid stopper. I have gone as high as 10k on some of my builds and still don't believe I could hear any difference in sound.
     
    mountainhick likes this.
  7. andrewRneumann

    andrewRneumann Tele-Afflicted

    Posts:
    1,067
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    Tacking on to this thought. This resistor would be sort of like permanent master volume setting. It may turn out you need no resistance to get the drive you want out of the power tubes. Or you may want a master volume control. In this cases replace the resistor and the grid leak with a 250/220KA potentiometer.

    If you start with a high value, you can easily clip in lower values or short the resistor completely for no attenuation. If you start with a low value, or no resistor at all, you are forced to unsolder something to experiment. I always recommend a test pot in these situations. In this case you might just leave the test pot in as a master volume.
     
    mountainhick likes this.
  8. mountainhick

    mountainhick Tele-Meister

    Age:
    64
    Posts:
    211
    Joined:
    May 2, 2021
    Location:
    Rocky Mountains
    Master volume is potentially appealing. A number of schematics use a 1meg in similar position. Is the 220-250 value based on the grid leak value for that reason? I'm guessing that that just erases the 470K function altogether except it is accounted for in the sweep of the pot.

    EDIT: This what you mean? 6AK6-6G6B-update2mv.jpg
     
    Last edited: Oct 12, 2021
  9. andrewRneumann

    andrewRneumann Tele-Afflicted

    Posts:
    1,067
    Joined:
    Mar 22, 2020
    Location:
    Cincinnati, OH, USA
    Yes.

    Yes.
     
    mountainhick likes this.
  10. mountainhick

    mountainhick Tele-Meister

    Age:
    64
    Posts:
    211
    Joined:
    May 2, 2021
    Location:
    Rocky Mountains
    Revisiting this, would there be any problem adding effects send return between master vol and power amp grid stoppers?
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.