1. Win a Broadcaster or one of 3 Teles! The annual Supporting Member Giveaway is on. To enter Click Here. To see all the prizes and full details Click Here. To view the thread about the giveaway Click Here.

Bias circuit from main B+ - question and sanity check!

Discussion in 'Shock Brother's DIY Amps' started by buddy_coles, Apr 23, 2021.

  1. buddy_coles

    buddy_coles TDPRI Member

    Posts:
    43
    Joined:
    Dec 16, 2014
    Location:
    Midlands, UK
    I need to get a bias supply for 6L6s in push-pull, and my transformer does not have a dedicated winding for this. To the best of my knowledge, I would need roughly 50-65 negative VDC to bias the valves, which is not hard to achieve using a simple voltage divider.

    For this build I also want to add a Lar-Mar style master volume control (based on Rob Robinette's website) which as I understand would bleed some of the signal from the phase inverter to ground.

    I was also looking at Merlin Blencowe's site where he suggests a couple of mods to improve reliability and 'idiot-proofing' and I just wanted to check I haven't missed anything obvious! Here's a sketch of the circuit I think will work. Assuming half wave rectification at about 90% efficiency, I reckon 350VAC would turn into about:

    350 / 2 * sqrt(2) * 0.90 = 223VDC approx. although there should be plenty of adjustment on the pot for the final bias voltage.

    Question 1: Are there any obvious flaws in this circuit?

    Question 2: Does anyone see an issue with adding the Lar-Mar master volume mod (also pasted below) to this circuit? I believe the 39uF cap on the left would allow the amplified guitar signal to pass to ground without any issue, but wanted to check with others more experienced than I am!

    Many thanks for any comments or input,

    Sid

    upload_2021-4-23_14-21-18.png
    My proposed bias circuit, should give about -27 to -95 VDC, not considering the 100K emergency resistor.
    upload_2021-4-23_14-26-23.png
    Lar-Mar master volume from Rob Robinette's site
     
  2. Ten Over

    Ten Over Tele-Holic

    Posts:
    771
    Joined:
    May 13, 2015
    Location:
    Central California
    The 47K resistor in the bias supply needs to be 1W minimum.

    The bias range seems excessively large to me. Maybe a 10K pot with a 15K idiot resistor for an expected range of -41V to -69V.

    The bias capacitors don't need to be that large, unless you're going for really low ripple voltage. The large discharge path for the second capacitor significantly reduces the ripple voltage.

    Here is another take on the same idea.
    Bias Table 20 Example PNG.png

    The master volume is fine as drawn.
     
    buddy_coles likes this.
  3. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    90
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    If you are trying have a bias to allow something with large neg bias it makes sense but if that is not the case, perhaps use the Marshall 2204 bias circuit. I am able to put any tube in my amp with the range I get from it, assuming close reflected impedance. And even with that in mind the bias point and plate voltage can change the impedance range so, it doesn't have to be perfect.
     
  4. buddy_coles

    buddy_coles TDPRI Member

    Posts:
    43
    Joined:
    Dec 16, 2014
    Location:
    Midlands, UK
    Thanks both - really appreciate your advice and help. It's no bother to swap a few resistor values, and the Marshall style looks no less elegant than Fender ☺
    For the caps they are higher value just because that's what I happen to have!! I've probably got some lower value ones kicking around so will have a look.

    Thanks again, Sid
     
  5. buddy_coles

    buddy_coles TDPRI Member

    Posts:
    43
    Joined:
    Dec 16, 2014
    Location:
    Midlands, UK
    Ok, another (possibly stupid) question following on from the bias supply design:

    For the tremolo jump start on an AB763 circuit, is the connection to the bias supply ONLY intended to provide a big negative grid voltage wrt cathode, to ensure 100% cut-off when tremolo is not in use?
    Or in other words, is the voltage particularly sensitive? If I've understood it correctly then I would expect that the -55VDC could be +/- 30% and still work fine. But I am not completely sure I have correctly understood it!!!

    Thanks in advance

    Sid
     
  6. Ten Over

    Ten Over Tele-Holic

    Posts:
    771
    Joined:
    May 13, 2015
    Location:
    Central California
    I'm sure the jump start effect was just a happy coincidence. The negative voltage on the grids of the LFO triode and the neon driver triode puts both triodes firmly in cutoff. If you merely stop the LFO triode and leave the neon driver triode's grid at zero volts, sometimes the neon bulb remains lit. The constant light on the photocell reduces its resistance so that the tremolo intensity control acts like a volume control for the vibrato channel.

    I'm not sure exactly how negative the neon driver's grid has to be in order to starve the neon lamp. The LFO is off at the same time so that the negative grid voltage doesn't have to fight that signal. All it takes is enough negative grid voltage to make the neon lamp go off. Fender uses -7V on their '68 Custom Deluxe.
     
    XTRXTR, buddy_coles and robrob like this.
  7. XTRXTR

    XTRXTR TDPRI Member

    Posts:
    90
    Joined:
    Jul 13, 2020
    Location:
    Northern Colorado
    @Ten Over and others know way more than me about that. I always understood it is there to maintain the tremolo from stopping and to start it when first powered on. Like a spark plug. And the tremolo oscillator resistor values are large to keep it from pulling any current from the bias supply. Similar to a voltage divider on the HT to raise the heater center tap reference.

    I'm still learning.
     
  8. buddy_coles

    buddy_coles TDPRI Member

    Posts:
    43
    Joined:
    Dec 16, 2014
    Location:
    Midlands, UK
    I hope the day never comes when I stop learning!

    Really grateful to you and to Ten_Over for the input here. I'll do a bit more circuit studying later on and hope I'll get my brain around it.
     
    XTRXTR likes this.
  9. Pick_n_Strum

    Pick_n_Strum Tele-Meister

    Age:
    31
    Posts:
    157
    Joined:
    Sep 28, 2016
    Location:
    Lino Lakes, MN
    Great thread OP! I've been reading up on bias supplies myself - I have a question for a couple of you:

    @Ten Over: Is that 27k acting as an idiot resistor? I see the wiper and outer lug are going directly to ground which is obviously different than the Blencowe design. I'm personally someone who needs that idiot proofing but curious to know more if it's still implemented in this design.

    Looks like the Marshall design posted by @XTRXTR is similar to the one posted by Ten Over. Curious if that one also includes idiot proofing.
     
  10. buddy_coles

    buddy_coles TDPRI Member

    Posts:
    43
    Joined:
    Dec 16, 2014
    Location:
    Midlands, UK
    Sorry to answer when the question wasn't directed specifically to me, but yes the 27k resistor makes the circuit idiot-proof :)
    I try to think of it in terms of: if the pot is turned to give hottest bias (lowest negative voltage), what is present between the 'output' to the grids and ground?
    In the Blencowe example the 10k resistor prevents the grid supply from going directly to ground; in the Marshall example (and Ten_Over's example) the 27k resistor is always present between grids and ground.
    The circuits differ mainly in where the pot goes, but all adjustable bias supplies I've seen are essentially a voltage divider i.e. two resistors with a tap between them. In the examples shown there is also some low pass filtering for smoother DC.
    I am leaning more toward the Ten_Over / Marshall style now I ponder it, though I also need it to be compatible with the Fender type tremolo circuit.

    I'm 90% there with my complete schematic so will post that in my Tweed-o-Verb thread as soon as it's done.

    Thanks again for all the info in this thread - it's been a huge help for me and given me a lot more confidence regarding the available options and tweaks.
     
    Pick_n_Strum likes this.
  11. Ten Over

    Ten Over Tele-Holic

    Posts:
    771
    Joined:
    May 13, 2015
    Location:
    Central California
    The 27K resistor puts a limit on how hot the bias can be just as the 10K resistor does in the Valve Wizard design.

    The two bias circuit styles under discussion here are very common in guitar amps. Both styles prevent the end user (idiot) from setting the bias hot enough to red plate the output tubes, at least when properly designed.
    Bias Basic Styles PNG.png
    Style 1 is the Valve Wizard one with the bells and whistles stripped away. Style 2 is the version that I posted, also stripped down. Style 1 keeps the ratio of the load resistance to the dropping resistor constant and varies the bias voltage with a potentiometer. Style 2 changes the ratio of the load resistance to the dropping resistor in order to vary the bias voltage.

    One of the problems with Style 1 is that the output tubes lose their bias and red plate if the wiper fails. To prevent this, a large resistor is placed between the top lug and the wiper so that the bias voltage defaults to its coldest setting in the event of a wiper failure. The resistor is large so that it doesn't interfere with the potentiometer. There is very little DC current in the power tube grid circuit so that there is essentially no voltage drop across that large protection resistor.

    I have had the wiper of a potentiometer find an alternate path to ground before, which created a ground loop and a hum. If the wiper finds an alternate path to ground in Style 1, it will make the bias hotter and possibly red plate the output tubes in an extreme case.

    If the wiper fails in Style 2, the bias voltage defaults to its colder setting. If the wiper shorts to ground, so what? The wiper is already connected to ground. This is why the pot is connected to ground instead of having the idiot resistor connected to ground.
     
    Pick_n_Strum, XTRXTR and buddy_coles like this.
  12. buddy_coles

    buddy_coles TDPRI Member

    Posts:
    43
    Joined:
    Dec 16, 2014
    Location:
    Midlands, UK
    Thanks again Ten_Over - that is a really clear explanation. I have gone for your version of the Marshall type bias as I think it's neater than the Valve Wizard one. Mainly because it incorporates the possibility of wiper failure into the design, rather than requiring an additional failsafe resistor. I've now plotted out my circuit diagram for the approximation of a Vibroverb AB763 that I'm building - if anyone is interested to have a peep it's here :)
    I think I'm now clear on the tremolo operation, master volume mod, and how both of those relate to the bias supply, so again a big thanks to everyone in this thread for the wealth of information and knowledge that you have provided. What did we ever do without the internet???
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.