Harvard 5f10 bias Q

[owen]

Having trouble biasing hotter on Harvard 5f10 clone: hammond 290HAX, tried two NOS rectifiers, one JJ 5y3s, and two modern 6v6gt sets

50v bias tap -> stock 6k8 resistor -> 1N4007 diode ->50kL trim -> 27k resistor -> ground.

with available bias from -48v (380v plates, 3mA bias) to -36v (356v plates, 12mA bias)

can’t get the 6v6gts above 12mA. sluckey’s advice elsewhere was keep the stock 6k8 resistor before the 50k trim and 24k (i used 27k, will try a 22k) resistor, and several people have made this same circuit work beautifully.

any suggestions on where to focus?

*UPDATE: swapped 27k resistor for a net 16k resistor after the trim pot, can’t get above 16mA on v3, 12mA v4

[jordan86]

Yeah I would start with a lower value on your 27K resistor. Like 15-22k range. You could parallel a 50k or 100k in with the 27k pretty easily just to experiment, without doing a full solder job.

[owen]

Yeah I would start with a lower value on your 27K resistor. Like 15-22k range. You could parallel a 50k or 100k in with the 27k pretty easily just to experiment, without doing a full solder job.

thank you. i was going to try a 68k paralleled but there’s no room, have to unsolder anyway

what threw me a little is the OT primary resistances. higher values than i’ve built with before.

[owen]

update: with 16.1k total resistance to ground (2 parallel resistors =16k plus the pot fully bottomed out)

i am only getting 16mA on v3 and 12mA on v4, odd.

[sluckey]

What voltage is on pin 5?

[owen]

What voltage is on pin 5?

presently, -27.6 which is about 1v under bias voltage.

[tubeswell]

update: with 16.1k total resistance to ground (2 parallel resistors =16k plus the pot fully bottomed out)

i am only getting 16mA on v3 and 12mA on v4, odd.

The bias supply is a couple of resistors in series, with the voltage input at the source end of the 6k8, and the voltage output at the junction of the 6k8 and the 56k (which makes a plain ole voltage divider. The 25uF cap is there to shunt noise out of the voltage divider output). So if you want a specific output voltage, you need to know the input voltage (at the source). Formula for Vout is R2/(R1+R2), where R2 is the 56k and R1 is the 6k8 (on the Fender Schematic). The Source Vin is AC, so to get an equivalent DC Vin, it’s the source VAC x 1.4


The adjustment range is just making R2 into an adjustable resistance with a pot (wired as a variable resistor) and another (tail) resistance in series (to give a minimum Vout when the pot is cut). You add another 10-50uF cap across the tail resistance to shunt noise out.


Experiment away!

[glass54]

Owen
What is your voltage range, pin 5, when you go min to max on Bias pot?
....and by the way, the bias voltage is only a guide. Change output tube brand and you will probably find a variation of 10% from previous bias setting  and while I'm at it V3 and V4 Quiescent currents are not likely to be exactly the same, but close.
As Tubeswell said "experiment"
Regards
Mirek

[owen]

Owen
What is your voltage range, pin 5, when you go min to max on Bias pot?
....and by the way, the bias voltage is only a guide. Change output tube brand and you will probably find a variation of 10% from previous bias setting  and while I'm at it V3 and V4 Quiescent currents are not likely to be exactly the same, but close.
As Tubeswell said "experiment"
Regards
Mirek

currently at 22k total resistance  -49.2 to -31.5

with the 16k it was around -46 to -27.6

and now with 13k after the diode i’m around 20-21mA

i understand it’s a guideline, but changing the bias circuit so extremely far off spec isn’t something i’ve run into yet without even being close to the ballpark of the other 12ish watt fixed bias 6v6 amps i’ve built.

[glass54]

Thanks Owen.
That's crazy. The range is not theoretically correct 
Are you sure its a 50k Trimpot for bias?
Power off, measure the pot resistance (remove one side if needed).
I would also take the Output tubes out, power up and recheck the range of bias adjustment (but confirm 50k pot first).
Bias circuit is just a simple voltage divider (to quote Tubeswell "plain ole divider") with a very light load (output tube grids) therefore only a problem if you have excess current drawn somewhere 
Regards
Mirek

[owen]

The bias supply is a couple of resistors in series, with the voltage input at the source end of the 6k8, and the voltage output at the junction of the 6k8 and the 56k (which makes a plain ole voltage divider. The 25uF cap is there to shunt noise out of the voltage divider output). So if you want a specific output voltage, you need to know the input voltage (at the source). Formula for Vout is R2/(R1+R2), where R2 is the 56k and R1 is the 6k8 (on the Fender Schematic). The Source Vin is AC, so to get an equivalent DC Vin, it’s the source VAC x 1.4


The adjustment range is just making R2 into an adjustable resistance with a pot (wired as a variable resistor) and another (tail) resistance in series (to give a minimum Vout when the pot is cut). You add another 10-50uF cap across the tail resistance to shunt noise out.


Experiment away!

thank you for the math and response! I understand what’s happening in the bias circuit itself, I was just a bit dismayed after reading a dozen threads or so from others building harvards that I’d have to change this much to hit a good sounding (and performing) target.

even with sets of old RCA 6v6 that perform well in my brown princeton to 3 matched modern sets.

[owen]

Thanks Owen.
That's crazy. The range is not theoretically correct 
Are you sure its a 50k Trimpot for bias?
Power off, measure the pot resistance (remove one side if needed).
I would also take the Output tubes out, power up and recheck the range of bias adjustment (but confirm 50k pot first).
Bias circuit is just a simple voltage divider (to quote Tubeswell "plain ole divider") with a very light load (output tube grids) therefore only a problem if you have excess current drawn somewhere 
Regards
Mirek

this is exactly what’s bothering me. I measured the pot without power and no tubes before i even installed it, checks out to 50k. I measured the bias pot plus the resistor after it, and at full, it was about 77k, at ‘zero’ on the pot, 27k. I even tried two other 50k trim pots just in case.

i’ve done continuity checks everywhere, measured every component critically, and this is the one thing that doesn’t quite add up.

thanks for the suggestions. I’ll put a fresh set of eyes on it tomorrow evening and try again.

[sluckey]

with the 16k it was around -46 to -27.6

That's a good range for 6V6s. My voltage on pin 5 is -36v to give me 25mA cathode current. Plate voltage is 388V so idle dissipation is 9.7W (about 70%).

I'm using 1Ω cathode resistors to measure tube current. How are you measuring the current through the tubes?

[owen]

with the 16k it was around -46 to -27.6

That's a good range for 6V6s. My voltage on pin 5 is -36v to give me 25mA cathode current. Plate voltage is 388V so idle dissipation is 9.7W (about 70%).

I'm using 1Ω cathode resistors to measure tube current. How are you measuring the current through the tubes?

i barely get 20ma at -26v and 340v at the plates.

I also am using 1Ω cathode resistors.

[PRR]

Screen G2 voltage is important also. What do you have?

[owen]

Screen G2 voltage is important also. What do you have?

with 22kohm total bias resistance after the diode, -32.96 bias voltage

v4: 3) 347.1v
      4) 347.4v
      5) -31.9v
      8) 13.6mV

v3: 3) 346.8v
      4) 347.2v
      5) -31.9v
      8) 13.7mV

[owen]

Thanks Owen.
That's crazy. The range is not theoretically correct 
Are you sure its a 50k Trimpot for bias?
Power off, measure the pot resistance (remove one side if needed).
I would also take the Output tubes out, power up and recheck the range of bias adjustment (but confirm 50k pot first).
Bias circuit is just a simple voltage divider (to quote Tubeswell "plain ole divider") with a very light load (output tube grids) therefore only a problem if you have excess current drawn somewhere 
Regards
Mirek

Hi Mirek, with output tubes out, I get a the same range plus about .2 to .3v per socket pin five. 20 volt range on the trim pot. currently with 15k resistance. stumped. 

while testing the OT, the resistance difference between the OT primaries and center tap is a bit larger than I’m used to. 60 ohms (about 80ohms out of spec to the data sheet on one primary)

I have another hammond 8.5k 8ohm OT here that tests much closer to spec, with a difference of only 12 ohms. should at least help the power tubes bias a little closer to one another.

[glass54]

I'd be surprised if your Output Tx was a problem.

currently at 22k total resistance  -49.2 to -31.5
with the 16k it was around -46 to -27.6
and now with 13k after the diode i’m around 20-21mA

Leo's standard 5F10 had HT=305V and Bias=-27V and NOTICE........Voltmeter Values shown +/-20%
Maybe try 10k from Trimpot to ground and check because with your 13k you are close to your target.
You might also like to supply us with a good resolution photo of bias circuit  a picture is as good as 999 words.

[owen]

Leo's standard 5F10 had HT=305V and Bias=-27V and NOTICE........Voltmeter Values shown +/-20%
Maybe try 10k from Trimpot to ground and check because with your 13k you are close to your target.
You might also like to supply us with a good resolution photo of bias circuit  a picture is as good as 999 words.

compared to Sluckey’s schematic, i was getting close to target voltages fine, incl plate and bias -v within range, but the current for some reason couldn’t get anywhere near high enough.

Here’s a photo of my original bias scheme with 27k. I’ve since reversed the pot with the wiper to the dropper 6k8 resistor for cleaner wiring, with another 6k8 resistor after the 50kl trim pot behind the diode. 12 k across both nets a nice healthy 25ma bias current and the amp sounds beautiful, all scope traces checked out wonderfully. i just wanted to make sure something else wasn’t drawing current but I can’t find anything else out of sorts. changing a resistor that drastically still confuses me.

[glass54]

Thanks for the photo Owen. Looks Neat.
Did someone say that "Electrons work in mysterious ways" ?? 
Glad you sorted it.
Regards
Mirek