Voltage Increase w Bias Increase in Vintage Fender Amp

Discussion in 'Amp Tech Center' started by ocduff, Aug 22, 2019.

  1. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    Okay I’m away from home but a conundrum kind of struck me on vacation and I’m trying to wrap my head around this for when I get home.

    1961 Fender Super 6G4-A - bias cap has been replaced. Original diode. Filter caps replaced. Original Yellow Astron caps in phase inverter section (I suspect this my be issue) - but let’s leave these facts out of equation for the moment.

    My experience has always been - if the bias is increased, the plate voltage decreases.

    Leaving bias resistor the same and not “adjusting” bias, I swapped in three matched NOS Tung Sol 5881 tube sets when I got the amp. Basic Results:

    Tube Set 1 - 20ma / 450 volts
    Tube Set 2 - 30ma / 460 volts
    Tube Set 3 - 35ma / 470 volts

    Tube bias is stable and I have not checked the Astrons for leakage as the amp was serviced 2 years ago with what seems like a competent tech so I didn’t second guess him as he replaced anything bad.

    So my question is - if I put in a tube Set that drew more bias current, why is my plate voltage going up. My experience has always been opposite but mostly my experience has been with cathode biased amps.

    My inclination is as bias current is increasing I’m getting more DC leakage.

    Lastly I don’t recall the negative bias results but I recall they changed as I expected them to (ie, -ma with every tube swap). But I just don’t have those numbers committed to memory.

    My intention of course is to get a realistic voltage of 450v or so, and have my nice vintage tubes biased at 50-60%.

    My memory already isn’t what it used to be - I find I have to learn the same lessons over and over again if not confronted with them regularly. My impression was that something didn’t seem quite right here.

    Thanks for replies.
     
  2. peteb

    peteb Friend of Leo's

    Posts:
    4,042
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    It can’t be, it’s in violation of ohms law.


    The B+ has its max, everything takes away from it. The plate voltage is the B+ minus the voltage drop across the OT primary. More current, more voltage drop.



    How are you measuring current? Measuring plate voltage is simple. Measuring current is not. You can track down the discrepancy. Measure the B+, measure the voltage drop on the OT and the plate voltage. Calculate the current thru the OT. It has to work out.
     
    ocduff likes this.
  3. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    Thanks for reply. I’ll try several biasing methods and see if my results differ but I’m suspecting these two Astrons are toast and leaking DC.

    Or the old diode is letting something pass.

    No, it doesn’t compute and an odd result - just didn’t know if this raised some red flag for a “usual suspect”.

    Always something new to learn and in this case I’m learning my gorgeous minty amp’s Astrons probably all have to go at least where lots of voltage is present.
     

    Attached Files:

  4. Wally

    Wally Telefied Ad Free Member

    Posts:
    32,512
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Ocduff, different tubes will act differently in a circuit and cause different voltages to happen. I am going to think that if you had an adjustment pot in that bias circuit, things would act as you expect with any one pair of those tubes...as current draw increases the plate voltage will decrease and vice-versa. If you put in one of the other pairs, everything resets according to that pair’s operating parameters.
     
    Andy B, aerhed, RLee77 and 1 other person like this.
  5. peteb

    peteb Friend of Leo's

    Posts:
    4,042
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    I was thinking that too Wally. Changing the bias on a set of tubes would get the expected results, and changing tubes is different than a bias change, it is a reset.



    But, these three tube sets display hot, medium, and cooler bias characteristics.



    It should not matter if a tube draws more current because it has a low internal resistance, the cathode resistor is reduced, or the fixed DC bias voltage gets reduced. All of these will result in increased current flow and an increased voltage drop across the OT and a lower plate voltage.




    I would like to see these three numbers for all three tube sets, or at least two tube sets:



    B+ voltage

    Plate voltage

    B+ minus the plate voltage (voltage drop across OT primary)




    That’s only two numbers.
     
    ocduff likes this.
  6. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    Perhaps that’s what’s going on as well. It seemed uncanny that three sets I happen to have would ALL increase voltage and bias readings.

    Question: I know techs all have preferred methods of measuring bias and opinions on which is best.

    I use the 1-ohm cathode resistor method because I find it quick, safe, and accurate. Simple.

    The transformer shunt method is nice but with the amount I fiddle with amps, sooner or later it’s going to bite me in the butt, and so it’s my personal choice not to use it. Part of that is, when I only bias an amp every year or so, I don’t to want to forget something and get hurt or burn up equipment. End of story there. Mind’s made up there.

    However, the transformer resistance method is perhaps the easiest - you don’t have to even install resistors. However, it is a method I never got accurate results with. Because I believe you MUST have a digital multimeter that has decimal places when measuring V otherwise you get very vague results. There is NO setting on my DMM that I can find or am aware of that will get me decimals when taking voltages. I’ve spent time on this, and these are not cheapy basic DMM (nor are they top of the line).

    Example (and these are just fake numbers I’m pulling out of my memories that are 5 years old so they may seem ridiculous)

    Plate Voltage Tube 1: 448v
    Plate Voltage Tube 2: 449v
    Measure OT CT voltage: 450v

    Resistance of OT side 1: 250 ohms
    Side 2: 270 ohms

    2v/250ohm=12.5ma
    1v/270ohm=27ma

    See what I’m saying? I KNOW plate voltages are usually the same on power tubes - but often a DMM will waffle between 1 number and the other. There is no: 448.27v, so thus there is no 1.73v to divide by.

    And these are tubes I KNOW are very closely matched. Period. So that’s out of the equation.

    I’ve sought out DMM’s that show decimal places but they often don’t when they arrive. The answer to buy a Fluke can’t apply to everyone who just wants to adjust the bias on their tubes and be good for 10 years. That’s not an economical solution, but perhaps is the only answer, and so is the reason I use the cathode resistor method.

    But it’s relegated to my ONLY method as I can’t seem to get good result from any other method. I cannot be the only one who faces this issue, and so I suppose I’m saying this is a caveat to that method - if your DMM doesn’t show decimals, or you can’t make the damn thing show decimals even though it “can”, then you have a method that’s ultimately not always accurate.

    Not sure if that’s a rant or a question. But I know I’m going to want to use this resistance method to figure out my problem and maybe it’s time to pony up for some professional equipment - a Bias King Tester or similar just be most helpful.
     
  7. peteb

    peteb Friend of Leo's

    Posts:
    4,042
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    Agreed, the measurement requires reading at least one decimal point.



    Aren’t you measuring real small voltages across the 1 ohm resistors to calculate current that way?
     
    ocduff likes this.
  8. Wally

    Wally Telefied Ad Free Member

    Posts:
    32,512
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Millivolts.....if the meter cannot read millivolts across those 1 ohm resistors, then things get a bit vague, imho. Reading as volts makes it a bit less precise???.
    Fwiw, ocduff, a bias probe tool is in essence doing what you have been doing...reading MV across a 1ohm resistor. The only advantage I see with such a tools is if it reads plate voltage as well. Then, if the bias is adjustable from outside the chassis, one does not have to pull the chassis to set the bias while understanding exactly what is going on..current draw, plate voltage, and hence plate dissipation.
     
    ocduff likes this.
  9. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    Yes but my DMM I believe only gives 3 numbers regardless of whether there’s a decimal. IE I just get 0.27 when reading DC MV.

    If I’m reading higher voltages like 35.8v it will add a decimal place. But getting into 100v or above, I’m out of room.
     
  10. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    I’ll try resistance method and see what I come up with for voltage drop across OT primaries with different tube sets.

    I just remember thinking with the first set I tried, reading 450v and 20ma, that the other sets I had (which I knew often had a higher bias when all else stayed same) were going to give me spot on voltages and bias.

    No dice.

    Again, would leaky Astrons on phase inverter and negative bias throw off weird numbers?

    Only thing I could think of was that as current draw increased, so would the amount of leaky DC, thus throwing off the numbers - literally as if a bucket was overflowing back into the well the water is being drawn from. Or something.

    I would assume this may affect voltage drop across OT as well.
     
    peteb likes this.
  11. Wally

    Wally Telefied Ad Free Member

    Posts:
    32,512
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Put a bias adjustment pot in that circuit and see what happens with the voltage as you increase current draw for one pair. Then, install another of these pairs and do the same thing.....and compare the beginnings of each pair to the beginning of each pair at the same bias voltage, right? Or...if you have another amp to try these tubes in, see if you see the same thing...or something different.
     
    ocduff likes this.
  12. peteb

    peteb Friend of Leo's

    Posts:
    4,042
    Joined:
    Apr 25, 2003
    Location:
    Cascadia

    I would get some good numbers using the OT resistance method using two place accuracy from your DMM.




    Then ask the question above, if the numbers are still off.



    Adding an adjustable bias is always a good option. You have three sets of tubes running at three different operating points. You should probably be able to find one pair that gets you close to the bias you want.
     
    ocduff likes this.
  13. Wally

    Wally Telefied Ad Free Member

    Posts:
    32,512
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    One can sit and wonder about methods and meters all day long. Until one does some basic experimentation so that things can be accurately compared, one does not know what is going on concerning the original question...why do those three pairs yield the numbers they do????
    Maybe I see this as so simple because I use the shunt method and these types of experiments can be done very quickly?? Of course, measuring the 1 ohm resistors should be just as quick. I have never used the OT resistance method and never will. It has been to be no more accurate than the shunt method...within 1% of each other, iirc.
     
    ocduff likes this.
  14. RLee77

    RLee77 Friend of Leo's

    Posts:
    4,293
    Joined:
    May 15, 2016
    Location:
    Nevada
    The OT drop is somewhat insignificant in this case, as a 10ma difference in current only creates 2.5v change across the OT winding (using a winding R of 250 ohms). So that’s only 25% of the 10v change in Vp recorded.

    It’s certainly not what you’d expect, to see an increase there. I suspect either meter inaccuracies, or varying incoming ac line voltage, or the amp warming up and causing the power supply to slowly increase its total output. I would put my money on the latter.
    @ocduff : Next time you do it, get the B+ voltage as well in each case.
     
  15. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    I’m away from home and will have it sorted once it’s in front of me. When I used to use the shunt method I did not find it extremely accurate either - chalk it up to user error but I did not always get same results across methods. Others have noted this too.
    So, since then, I have stuck with cathode resistor method as at least it was consistent in giving me readings that I could work with, and I trusted what I was working with within 5-10%. Good enough for my purposes.
     
  16. Wally

    Wally Telefied Ad Free Member

    Posts:
    32,512
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    I have doing this for over a quarter of a century. I can measure current draw with the shunt method all day long and get consistent results. Comparisons have shown it will give results within 1 percentage point of the other two.
    I apologize for posting 8n this thread. Play music.
     
    ocduff likes this.
  17. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    I’ve found 10v discrepancy in line voltages throughout the day.

    One thing never mentioned in biasing is how long the amp has been on. There must be a 15% discrepancy between just turned on and an hour of warm up. So I don’t know what to make of that in a conversation on biasing.

    So this mayyyyy factor in - again I need to focus on this.

    FYI - I’m not the bias police BUT when using vintage glass I like to know what’s going on. With new tubes, bias them to red plate. Who cares, they will develop mechanical noises shortly if they didn’t already come from the factory that way.

    Lastly, the tube sets in question are Tung Sol 5881 and Jan 6L6WGB and there are those who believe the former is a 23 watt tube and the latter a 26, and those who believe they are both 23 watt tubes.

    Usually in a clone I’d add a bias pot and a dropping resistor to get voltages and bias where I want them - I wish someone would market brackets for real bias pots to mount on PT bolts. I don’t even totally trust bias pots after having one fail and suddenly have frigid bias and extremely high voltage.

    But unless you all know of a good source for those little screwdriver tip sized bias pots of a high quality I’m okay with swapping resistors as needed.
     
  18. ocduff

    ocduff Tele-Meister

    Posts:
    102
    Joined:
    Nov 17, 2003
    Location:
    USA
    Wally - didn’t mean to suggest there was anything wrong with shunt method.

    Been playing music through amp for 5 months and only now really want to dial it in.

    And, I’m on vacation, should’ve brought some amps and tools with me as I’m getting punchy.
     
    Luthier Vandros likes this.
  19. dan40

    dan40 Tele-Afflicted Silver Supporter

    Posts:
    1,271
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    Be sure the battery in your meter is not getting low. A weak battery can give higher than expected voltage readings.
     
    Luthier Vandros and ocduff like this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.