Negative grid bias for 6550's

Chad Hauris

Chad Hauris

Super Member
We are working on some RCA amps using 2 6550 push-pull ouptut tubes.
Plate voltage is 530, screen is 240. These amps have adjustable bias controls for each 6550.
My problem is slightly distorted sound and kind of mediocre power output. I have it set for -21 volts at each grid and if the grid voltage is made more negative, it gets more distorted.
All capacitors in the amp have been replaced.

Does this seem like around the proper negative grid voltage? The GE tube book gives a value of -16 volts for a 450 volt plate voltage.
We don't have any brand new 6550's to try yet but subbing in other used 6550's just gives worse distortion. These amps were originally made for 6146s but someone field converted them to 6550.
 
Register to hide this ad
The bias voltage is not what you need to look at; it is the bias current. If this amp has a single bias pot per pair, then a 6550 should be biased to about 130 mA per push-pull pair. If the tubes are biased individually (and/or have separate cathode resistors on each tube), then they be biased for 65 mA per tube.

To determine the quiescent current, measure the DC voltage across the cathode resistor, and use ohms law to calculate the DC current through the cathode resistors.

You might have a very poorly matched set of tubes inside this amp. Also that screen voltage sounds a bit too low to me as well. :scratch2:
 
hey-Hey!!!,
I likely know the man who designed that amp. In his personal ones, he substitutes 0D3's to bring g2 to 300V with 6550's. Run it up to 50 mA per tube at least and see where your grid voltage is. My single still has 6146's( along with all its original caps ), so I am not using it.
cheers,
Douglas
 
And if everything was running correctly I'd suspect -25 to -30
would be closer to the correct -grid bias voltage on each tube.
 
Thanks a lot for your help...On this amp the cathode goes directly to ground but there is a cathode-break jack to hook up a milliammeter, there is no cathode resistor. On the amp panel it recommends 20 ma idle current per tube...I'll just try and take some measurements and see what the current is. It is a mono amp with an individual bias adjustment for each 6550.
We have some OD3's so I'll try those in the regulator section instead of the OC3's.

When I start to bring the bias up towards -25 to -30 volts the sound really starts getting fuzzy. I probably should test these 6550's and see what they read on a tube tester but none of the 6550's we have in the shop seem to give really great results in the amp.
Probably need to get some new 6550's but I have not seen weak tubes cause this much distortion in other amps, more like just weaker output.
 
Last edited:
hey-Hey!!!,
First things first...set idle current by measuring it directly. You'll need to make an adapter so you can plug your meter into the 1/4" phone plug jack. The actual voltage on the grid is pretty much irrelavent. It is the current you're setting through the tube that the amp cares about.
cheers,
Douglas
 
PakProtector said:
hey-Hey!!!,
First things first...set idle current by measuring it directly. You'll need to make an adapter so you can plug your meter into the 1/4" phone plug jack. The actual voltage on the grid is pretty much irrelavent. It is the current you're setting through the tube that the amp cares about.
cheers,
Douglas
Click to expand...

Completely agree, was just guessing at what the -grid voltage would be when
things are running correctly compaired to what I've measured in the past is all.
 
PakProtector said:
hey-Hey!!!,
First things first...set idle current by measuring it directly. You'll need to make an adapter so you can plug your meter into the 1/4" phone plug jack. The actual voltage on the grid is pretty much irrelavent. It is the current you're setting through the tube that the amp cares about.
cheers,
Douglas
Click to expand...
Exactly right. This is the same point that I was attempting to make.
 
Well...the grid voltage isn't *COMPLETELY* irrelavent, but nearly. If it is widely different between the two finals it could be an indication of poor matching...but with individaul adjusters, that is pretty trivial.

Measure current; ignore grid voltage; listen to the amp with proper current coursing through its veins...then report findings.
cheers,
Douglas
 
Chad, do you have a scope?

Have you looked at the signal waveform from the input all the way thru the amp? There could be many things causing this distortion.

If the plate voltage stays at 530 V with 65 Ma current draw, the plate power dissipation will be over 34 watts! I'm sure this is higher than what is recommended for long life.

I wire a 10 ohm resistor across those 1/4" jacks so that I can open them up and measure the voltage to determine the bias current (.65 V = 65 mA).

Does the 6146 use about the same output transformer? I have never heard of this type, so I looked it up. It is fairly close to a 6550, but the 6550 G2 Voltage could be raised about 50 V on your amp. Maybe 275 V to 300 V???

Richard C.
 
forget grid voltage and install some 10 ohm cathode resistors and measure the current of the tubes. The grid voltages your mentioning seem extremely low(not negative enough to me) unless the amp is partially cathode biased or something. My amps require -50 to -60 on the control grids with 500V on the plates.

I suspect you have some other issue causing your distortion though.

Craig
 
Hey Craig,
There is no cathode bias arrangement. Current gets directly measured. No Ohm's law involved.

Chad Hauris said:
On this amp the cathode goes directly to ground but there is a cathode-break jack to hook up a milliammeter, there is no cathode resistor. On the amp panel it recommends 20 ma idle current per tube....
Click to expand...

cheers,
Douglas
 
Kegger said:
But installing some 10ohm resistors and measuring that way wouldn't hurt.
Click to expand...

If I could get those measurement jacks as cheaply as I get precision 10R resistors, I'd never use the resistor again. Those jacks are a very nice feature, no sense removing them. A trip to RadioShack with $5 will solve this problem very simply( as in make the adapter tool).
cheers,
Douglas
 
PakProtector said:
Hey Craig,
There is no cathode bias arrangement. Current gets directly measured. No Ohm's law involved.



cheers,
Douglas
Click to expand...

Ah I missed that.

I was just looking at the 6550 spec sheet and operating in AB1 pentode with 600V plate and 300V screen they list -32.5 volts to the control grid so I guess were not too far off here. I'd put the control grid at -30 or so and start looking else where for a problem. If this amp is running only 20 mA current for each tube it must be heavily on the B side of the AB spec or possibly 100% class B.

One thing about those jacks is you have to confirm that they are indeed getting a good solid connection for accurate readings and stable cathode connection to ground they will get seriously oxidized over the years. I prefer a fixed resistor myself, a soldered connection is always better. The resistor will add a very small amount of feedback but very much negligible IMHO.

Craig
 
20 mA for 6146 is a bit different from 20 mA for a 6550. Again, in this situation, *IGNORE* the g1 voltage and instead measure current through the tube as the amp is designed to directly do.

And, if you're going to run 6550, you should put in two 0D3 to raise the g2 voltage to 300. The low voltage was used so that RCA's 6146 could be used in that RCA amp.
cheers,
Douglas
 
I haven't got a chance to get back to the amp yet but thanks again for your help. I'll try all your suggestions and report back.
 
You must log in or register to reply here.
Back
Top Bottom