NFB with Cathodyne

Discussion in 'Amp Tech Center' started by _Steve, Jul 29, 2020.

  1. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    20
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Hi,

    I've nearly finished what will be only my 3rd build, so still very much learning.

    It's a 6G3 brownface Deluxe built from a Hammond AO-29 chassis and transformers.

    The only deviation I made from the 6G3 schematic was to go for a cathodyne inverter rather than the LTP. The reason I did this is that Im usually a Marshall guy and I wanted this amp to be the polar opposite, clean-only, smooth, mellow-toned amp. Going with a cathodyne also means I'll have an extra triode to experiment with later. I lifted the cathodyne spec directly from the 5E3.

    The 6G3 has its NFB going into the input side of the LTP inverter. I've have read Rob Robinettes 5E3 NFB mod but that connects the NFB to a prior gain stage.

    So my question is can link the NFB directly to the input-side of the cathodyne rather than a gain stage?

    Many thanks
    Steve
     
  2. elpico

    elpico Tele-Afflicted

    Posts:
    1,017
    Joined:
    Sep 14, 2011
    Location:
    Vancouver BC
    The answer is yes, but a cathodyne has less headroom than a Marshall style ltp and a 5e3 is famous for being a loose, wooly, crunch machine not a "clean only, smooth" amp so something got a bit mixed up here.

    With a high enough supply voltage to the cathodyne you could make it stay super clean, but it takes a lot less to do the same job with an ltp.
     
    Wally, Nickfl and D'tar like this.
  3. Jon Snell

    Jon Snell Tele-Meister

    Posts:
    394
    Joined:
    Aug 31, 2015
    Location:
    Jurassic Coast. UK.
    To correct the phase, you can reverse the output transformer secondary polarity if you insert the Nfb to the wrong place and incur Positive feedback.
     
  4. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    20
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Thanks for the replies. So would I just inject it directly into the grid of the cathodyne?

    deluxe_5e3_schem (2).gif
     
  5. Mexitele Blues

    Mexitele Blues Tele-Afflicted

    Age:
    42
    Posts:
    1,057
    Joined:
    Mar 7, 2018
    Location:
    Westminster, CO
    The Princeton Reverb injects the NFB at the cathode of the final preamp stage before the cathodyne.
     
    Wally likes this.
  6. Jon Snell

    Jon Snell Tele-Meister

    Posts:
    394
    Joined:
    Aug 31, 2015
    Location:
    Jurassic Coast. UK.
    No. Doing that will destroy the bias on the cathodyne. A 0.1uF 400v capacitor will need to be in series with the Nfb resistor feed to block the DC.
    If you want Nfb place a 180R resistor at the ground end of the 25uF and 1k5 pair and feed your Nfb at that point. (Left hand side of 12AX7)
     
    robrob and Mexitele Blues like this.
  7. D'tar

    D'tar Friend of Leo's

    Posts:
    2,712
    Joined:
    Jan 11, 2013
    Location:
    WNY
    upload_2020-7-30_10-36-39.png
     
  8. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    20
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    OK thanks. Im going to accept it's not possible however for my own learning can someone please elaborate on why it cant inject straight into the grid of the cathodyne?

    It seems to already be shielded from DC on the preamp side with a coupling cap, and the NFB signal itself should be AC only right?
     
  9. Wally

    Wally Telefied Ad Free Member

    Posts:
    35,999
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    As noted above, what you are building is the opposite of what your goals are...except for the NFB. A straight 6G3 would be a cleaner amp, and it could be made cleaner very simply. Change the power supply resistors for the preamp to 100K. Run a lower gain input stage tube...5751 or 12AY7. Decrease the value of that 56K NFB resistor....and put a pot in series with it to warm things up if wanted.
    urn a 12AT7 in the LTP phase inverter.
    So...this is what you are building...more or less..??

    https://schematicheaven.net/fenderamps/vibrolux_5e11_schem.pdf
     
  10. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    20
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    I think I probably could have explained my goals a little better :) The 6G3 is really just a starting point for further experimentation and learning. I heard that the Tremolo is outstanding and it is a very easy build. The reason decided to substitute in a cathodyne was because I will only be running it clean and therefore do not need to worry about getting a good distorted tone. My understanding is that the LTP was added to mainly to address this. Going with the cathodyne gives me: 1) an extra triode stage freed up for experimenting with and b) experience with a piece of topology that is different from a Marshall

    Yes that 5E11 Vibrolux looks pretty much exactly what I am building, other than having grid stoppers on the 6V6s and some different values in the bias section. I didnt know this amp existed - my knowledge of Fenders is very limited. I've literally played on 2 in my entire lifetime :)
     
    Wally likes this.
  11. Wally

    Wally Telefied Ad Free Member

    Posts:
    35,999
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Steve, ime, you have chosen the wrong amp if you want to get away from ‘Marshall’ sounds. Ommv....but the 6G3 and 6G2 amps are rather firmly situated in the tweed world despite being part of the 6G amp family. This means that they are very lively clean amps and ime they RIP when pushed....sort of like a low volume big Marshall....or a low volume 5F6A from which the Marshalls were derived....when seriously pushed.
    The more complex tone stacks that follow the first gain stage in the BF amps gets away from that tweed influence. A bias vary trem could be implemented into say an AB763 Deluxe non-verb...and that would be a much cleaner machine, imho.
    .that said, I would look t9 that 5E11 for guidance on the NFB loop.
     
  12. Lowerleftcoast

    Lowerleftcoast Tele-Afflicted

    Posts:
    1,406
    Joined:
    Dec 30, 2019
    Location:
    california
    Do a search for Rob Robinette 5E3 mods. There is a good description for NFB on the 5E3.

    If I am reading you right, I think you are searching for that 5E3 wooly clean. You may be able to hang onto it with higher voltage but even if you max the Hammond out, the B+ may not be higher than a stock 5E3. The AO-29 schematic B+ is only 320 with the 5U4. (Maybe a little higher with higher wall voltage.) (Might be just the ticket for a 5E11.)

    Also you may want to bias it a little cooler than the over 100% crowd. It will keep it from distorting so quickly but there may be trade offs affecting the cleans.

    I am not impressed with the OT of the Hammond AO-29. You might be better served with something/anything else. OMMV.
     
  13. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    20
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Thanks for the info and opinions. Im very relaxed about the 6G3 not being the best amp or the one 'im looking for'. I just want to get it across the line so that I have a benchmark and then I can plan where to go from there - I really know next to nothing about Fender/American sounding amps. Its very much the start of a journey and im hoping to use it as a foundation to learn more about dynamics and harmonics as I add/change bits of it. I've built it P2P using individual tagboards for each section so I can easily swap out chunks. I also have 3 triodes spare to experiment with. But of course I have to get it over the line first, which Im hoping will be today!

    Im going to follow the Rob Robinette method of injecting the NFB into the cathode of the previous stage, but its still bugging me that I don't understand why going into the grid of the cathodyne wouldn't work. Its not that I don't believe it, i'd just like to understand it more.
     
  14. Wally

    Wally Telefied Ad Free Member

    Posts:
    35,999
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Being that this build and the 5E11 that is essentially what Steve is going for are fixed biased, I would definitely go for less 100% of max plate dissipation. 55-65% maybe...to the pleasure of one’s ears.
     
  15. Wally

    Wally Telefied Ad Free Member

    Posts:
    35,999
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Tag board is not PTP....no matter, though.
     
  16. Lowerleftcoast

    Lowerleftcoast Tele-Afflicted

    Posts:
    1,406
    Joined:
    Dec 30, 2019
    Location:
    california
    My mistake. I thought he was adopting the cathode bias of the 5E3. He is still with the 6G3/5E11 fixed bias.

    Carry on.:cool:
     
    Wally likes this.
  17. printer2

    printer2 Poster Extraordinaire

    Posts:
    6,740
    Joined:
    May 24, 2010
    Location:
    Canada
    NFB can be injected into the Cathodyne. I do not know how effective it is but Peavey did it for Their Classic 20.

    [​IMG]
     
  18. timfred

    timfred TDPRI Member

    Posts:
    93
    Joined:
    Mar 14, 2016
    Location:
    Northern CA
    In most NFB circuits, the interesting ratio is the FB resistor divided by the resistor to ground at the injection point - the voltage divider that sets the right level of NFB to inject. That's one part of the gain of the loop. (It's not that simple, also depends on the tap off the OT and the rest of the circuit, etc., but as a rough guide it works.) So in a normal cathode injection of NFB, that ratio is like 20:1 - 80:1.

    If you wanted to inject into the grid ... maybe that would work? But you would need a 10-80 Megaohm NFB resistor (PFB?) because your voltage divider is now has a 1M resistor as the bottom leg. I'm not sure what resistors that large, in combination with the capacitances and reactances of the rest of the circuit, would do to the tone.
     
  19. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    20
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Thanks for posting! Theres a lot going on there that i need to digest. Im going to try both methods to compare and hopefully post back the results as I think its an interesting topic. Obviously ill need to swap phases between the two.

    I've finished the build and am half way through the startup checks and all the voltages look good except the bias (its maxing out at around -7V). As soon as I work my way through that I should be up and running! Excited!
     
  20. _Steve

    _Steve TDPRI Member

    Age:
    40
    Posts:
    20
    Joined:
    Jul 14, 2020
    Location:
    Los Angeles
    Thanks for the explanation - it makes sense! The Peavy circuit above seems to be running around 3M on the top of the top of the divider but it also seems to be bypassing and shunting a bunch of frequencies and im not sure what the overall attenuation effect of that would be - still learning as I go!
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.