strange biasing behaviour (beware, long post)

[jezzbo]

Hello everybody,

I'm rebuilding an amp for one of my friends.
It originally was a Fender Bassman 100.
but my friend wanted something else .
I kept the original eyeletboard etc and made a new turretboard.
The original PT was also blown. No idea what caused this!
The Pt was replaced with a 125P34AX (twin reverb).
(http://www.tubeampdoctor.com/en/shop_Ma ... an_100_111)

The OT is still the original.

So I made a new amp in the Bassman 100 chassis.
The output section of the amp and biasing is standard blackface (AB763 showman amp).
Preamp 1 is a kind of tweed preamp with a mastervolume control.
(Volume - Tone - Master)

Preamp 2 is standard blackface)

Poweramp = 4X 6L6GC.
Plate voltage = 478V
Screen grid voltage = 476V
NO "Vibrato" (tremolo)

Anyway, the amp is build, there are no short circuits (tested with and without a current limiter) No smoke etc

Now for my question:
When I try to biase the amp (using 1 Ohm resistors, and also tried with a bias probe) I get no or really low reading om my multimeter.
I should be able to get somewhere arround 37mA (actually mV) but I only get readings of 0 to 4,2mA.
Even when I turn the bias pot completely open.

So I checked some things.

1. Bias voltage from the transformer = 60V AC.
2. voltage at the 220K resistors goes from -77V DC to -57V DC. (bias potmeter fully clockwise and fully counter clockwise)
3. Biaspot is 10K-L like on the showman schematic)
4. the resistor at the bias pot = 27K
5. the resistor where the bias tap from the transfomer is connected to is 470R (like on the showman schematic)
6. The bias cap = 100uF / 100V.

Does anyone can tell me why I get such a low reading when meassuring the current to bias the amp?

Here are some pictures.

http://users.telenet.be/jezzbro/veine_amp/


Thank you for helping me.

Jesse

[Leo_Gnardo]

[quote="jezzbo
2. voltage at the 220K resistors goes from -77V DC to -57V DC. (bias potmeter fully clockwise and fully counter clockwise)
3. Biaspot is 10K-L like on the showman schematic)
4. the resistor at the bias pot = 27K
[/quote]

-57V isn't low enough to allow proper current flow in your output tubes. By reducing your 4. resistor at the bias pot 27K to a lower value I'm sure you'll be able to get the bias voltage you need. Quick & dirty you could parallel another resistor with the 27K, try 220K down to 100K. Dial the bias adust pot to the maximum voltage when starting each test session and adjust down until you get a sensible bias current. The "Twin-size" amps can run with as little as 25 milliamps per output tube, and you can increase that up to @ 40 mA if your ears detect an improvement. FWIW I usually set 'em at 30 mA.

[Structo]

Also make sure your power tubes are good.

[Phil_S]

1. Bias voltage from the transformer = 60V AC.
2. voltage at the 220K resistors goes from -77V DC to -57V DC. (bias potmeter fully clockwise and fully counter clockwise)

Jesse: I think you want to have Vg1 = ~ 37V. The range of 77 to 57 is good range but too much voltage (or too little if you are stickler for the negative aspect of it) and very cold bias.

I think you need to drop about 30V on the bias circuit. This is usually done with a large value dropping resistor between the 60V supply and the reverse diode. Look at how Marshall did it in many older models like the JTM45. The range of resistor is typically 150K to 220K. You will have to tweak yours based on grid current and it could be more or less.

[jezzbo]

@Leo_Gnardo: thanks for the info. will try this.
@Structo. Tubes are new Tung Sol (matched quartet.) I assume they are good, but how do I know for sure? thanks.

@Phil_S: thanks for the info, Maybe I'll try this first. (have to check what resistors I have at the moment)

[Leo_Gnardo]

@Leo_Gnardo: thanks for the info. will try this.
@Structo. Tubes are new Tung Sol (matched quartet.) I assume they are good, but how do I know for sure? thanks.

@Phil_S: thanks for the info, Maybe I'll try this first. (have to check what resistors I have at the moment)

I think Phil_S may have overlooked you have 6L6 in your outputs. -37V is often given as the bias voltage for EL34, -47V for 6L6. Of course you have your 1 ohm resistors to measure bias current so keep a voltmeter on one of these as you dial down the bias voltage. When you measure a comfortable current level, check the voltage at pin 5, and you'll probably find it within a couple volts of the expected value. Also remember the voltages given in old amp schematics are plus or minus 20% so don't panic if it's not exactly the voltage on the schematic. Current is what counts.

When I refer to higher or lower, it's "absolute value", that is disregarding the sign. Lower = closer to zero. It does cause some confusion so I hope that clears it up.

[Phil_S]

Here, I found a schematic of a bias circuit. Sometimes a picture is worth more than words. Values may be different from your circuit, but you should get the idea.

[Phil_S]

Leo: I am looking at RC-30 for the bias voltage. See the picture. I am not nearly as experienced as you. Perhaps I misread the manual? In any case, we agree, the 1 ohm resistors will tell the OP when he's hit the right voltage. I'm thinking he should center the 20V range with roughly -40V in the middle. Right now, the mid point is -67V, so a drop (really an increase) of 27V is going to get into the right range. If it turns out to be centered too high or too low, the component values of the bias circuit can be tweaked.

[Leo_Gnardo]

Leo: I am looking at RC-30 for the bias voltage. See the picture. I am not nearly as experienced as you. Perhaps I misread the manual? In any case, we agree, the 1 ohm resistors will tell the OP when he's hit the right voltage. I'm thinking he should center the 20V range with roughly -40V in the middle. Right now, the mid point is -67V, so a drop (really an increase) of 27V is going to get into the right range. If it turns out to be centered too high or too low, the component values of the bias circuit can be tweaked.

You read the manual right all right. And it has the SG at 400 with plate at 450. In the case we're looking at, Plate is 478 and SG 476, much more conducive to current flow. A condition not approached in the RCA manual. Lowering the G2 - screen grid - has the effect of reducing current flow, so in an amp with those values, -37 may be right on the money.

edit: Hmm or maybe not - by the book these conditions yield a no-signal plate current of 116 mA. That's a plate dissipation of 52W in a tube that's rated for what, 25? What color is that plate gonna glow? I wonder why RCA put these things in the manual... maybe RCA wanted people to melt their tubes so they can sell 'em more? end edit.

Just about all the old Fender schemos for large 6L6 amps mention a -47V bias value. Also, as the output tubes start to draw current, the B+ will reduce a bit, say to 450-460V, nothing to worry about. The SG voltage will track down similarly. In big Fender amps I've worked on recently, B+ @ 450V, bias voltage has ranged from -43 to -52V.

[Phil_S]

Aha! Thanks for the explanation. I hadn't thought through the consequences of 450/400 vs. a 2V difference.

[jezzbo]

Tung Sol 6L6GC-STR MAX Plate dissipation = 30W
(http://www.tungsol.com/tungsol/specs/6l ... ng-sol.pdf)

I was thinking to achieve a plate dissipation of 18W (60%)

[Leo_Gnardo]

Tung Sol 6L6GC-STR MAX Plate dissipation = 30W
(http://www.tungsol.com/tungsol/specs/6l ... ng-sol.pdf)

I was thinking to achieve a plate dissipation of 18W (60%)

That'll put you at an Ip of @ 40 mA, with 450 on the plates. We're right in the ball park. I'd start at 25 mA, increase a bit at a time, and listen at each adjustment. I usually wind up at 30 mA but, suit your ears, that's what counts.

[VacuumVoodoo]

edit: Hmm or maybe not - by the book these conditions yield a no-signal plate current of 116 mA. That's a plate dissipation of 52W in a tube that's rated for what, 25? What color is that plate gonna glow? I wonder why RCA put these things in the manual... maybe RCA wanted people to melt their tubes so they can sell 'em more? end edit.

These values are for two tubes in P-P. This should be printed in block bold letters.
RCA biased them hot because they never imagined anyone sending heavily distorted power cords into these tubes. GC was supposed to be 30W plate dissipation tube.
One has to read the small print even ion the tube data sheets, we don't like it. /uttered with a fake nasal stiff upper lip Queens English/

[Phil_S]

That'll put you at an Ip of @ 40 mA, with 450 on the plates.

Plate voltage changes as you adjust grid voltage. Remember to verify plate voltage as you tweak it.

[Phil_S]

One has to read the small print even ion the tube data sheets, we don't like it. /uttered with a fake nasal stiff upper lip Queens English/

Yes, I just got new glasses and the doc increased the reading add on the bottom to +2.75. All of a sudden, I can see the top half of the newspaper again without moving it around.