Problem: Weber 5F6A with very low output!

[Diablo]

So your bias voltage is now -40 volts? That sounds very low and will run the output tubes very hot and drag down the B+. Set your bias voltage closer to -48V, or if you have a bias pot installed, adjust your bias so each tube is drawing around 35 mA at idle. Then measure your B+ again and it should be in the range of 460V.

[tubeswell]

When you lower the bias voltage, plate and screen voltage will rise because there is more tube current. You really should be ascertaining the plate current in order to check plate dissipation. If you were thinking of adding a 1% 1R resistor in series with each cathode-to-ground (0.6W metal film resistors will work fine), now's a good time. Since they are 1R resistors, then measuring the mV across each of those resistors will tell you the tube current (in mA) (I = E/R, so 1 mA = 1 mV/1 Ohm).

The tube current is the total current flowing through the tube, and includes current drawn by the plate and current drawn by the screen. So you have to deduct the screen current from the tube current to get the plate current. (In that amp you can determine the screen current by measuring the voltage across each of the 470R screen grid resistors, and dividing that voltage by the resistance of each resistor to get the current flowing through each resistor. - same equation I = E/R)

The plate dissipation with be the plate voltage x the plate current. For Class AB1, the tubes should be biased around 70% of maximum dissipation at idle. So for a 6L6GC, which has a 30W max dissipation rating, this is 21W at idle (which is about 50mA plate current per tube if you have a plate voltage of 420V, or if the plate voltage is higher at say 470, then it will be 44mA). Of course you don't have to go to 70% dissipation, 17-18W will sound good as well.

[jibe]

I bought the 2w 5% 1 ohm at the same time I got the 220k bleeding restistors.
I guess I can go back and change to 1%. I need them to do my biasing.

The higher the value of bias resistor the more negative bias voltage?

I maxed out my pot, can't turn it anymore.
I guess I have to replace it, think I have a 50k pot now,
maybe I need get a 100k then?

I'd like to install a "half-power" switch as well, read som where that all you need in
a fixed bias amp is to lift the ground on one tube, correct?
In that case a simple on-off switch will do.


//Jibe

[Diablo]

I bought the 2w 5% 1 ohm at the same time I got the 220k bleeding restistors.
I guess I can go back and change to 1%. I need them to do my biasing.

The higher the value of bias resistor the more negative bias voltage?

I maxed out my pot, can't turn it anymore.
I guess I have to replace it, think I have a 50k pot now,
maybe I need get a 100k then?

I'd like to install a "half-power" switch as well, read som where that all you need in
a fixed bias amp is to lift the ground on one tube, correct?
In that case a simple on-off switch will do.


//Jibe

Yes, the higher the value of bias resistor the more negative bias voltage. Just get a 39K resistor and put it in series with your 50K bias pot. That will give you a bias voltage range of -46 to -54 volts...at least it does with my 5F6A clone. I don't think that half power idea works - I've never heard of grounding out one tube in a push-pull. If you had 4 power tubes, yes you could have lifted the grounds on two tubes on opposite side of the push-pull.

[tubeswell]

Hi Jibe

Maybe you are thinking about the triode/pentode switch (that takes out the screens)?

W.r.t. the bias range, diablo has already said what I was going to suggest for increasing the -ve bias voltage (which will make the bias 'colder', and allow you to run the tubes in it at a higher plate voltage.). I was just going to wait to see what your measurements were for the idle dissipation before suggesting whether you needed to go up or down.

On the other hand, if you just want the bias to be hotter, just turn the pot down (i.e.; towards 'zero'). But don't turn it all the way down - that will kill the bias (and the tubes)...

And yes, for the cathode resistor measurement method, the mV-to-mA interpretation works better if you have a 1% resistor at each cathode, otherwise there is too much uncertainty in the measurement. Those cheap little dime-a-dozen 1% metal film resistors are fine

[erndawg]

Installed two 220k B+ bleeders, worked fine, drained the system within 30 secs. Does this lower B+ somewhat?

Those 220k resistors will consume less than 2 watts total; that is an inconsequential amount of power and will only lower the B+ by a few volts.

B+ is low 420 and bias voltage -40, whats up with that? How come it changed
so much? Had 470 B+ and -45.7 bias the other day.

Something is not right here. Your B+ dropped because you are again drawing way too much current through each power tube. A bias of -40 VDC is way hot. You need to figure out why your bias voltage has dropped. Go back and revisit the bias circuit.

I maxed out my pot, can't turn it anymore. I guess I have to replace it, think I have a 50k pot now, maybe I need get a 100k then?

No, don't change to a 100k pot. Figure out why your bias went south. If you need more bias because Weber's design is flawed, then change the 15k R41 to 12k or 10k.

but I can't say it's about 40 watts. Maybe 15-20.

Again, something is wrong here. Sort out the bias first, then I'll tell you how to figure out the exact power output from your amp. It should be VERY loud.

[erndawg]

I need a selector switch and a sturdy robust RCA Jack if I want to be able to switch impedance.

Or just use 3 Switchcraft 1/4" jacks and connect each OT secondary tap to each jack.

[jibe]

What's the difference? Changing R41 or putting a new R in series with the bias pot?


I can't see anything wrong with the bias circuit. Fixing the short in the RCA connector and installing bleeders changed the bias. I can't figure out how. Maybe it was just the short, now I'm back to normal and have to bias the tubes.

I'll get new 1% Resistors and install them, then take the readings I need to do the math.
Calculate dissipation to see how much I'm getting.


As for the "half power switch", it was a four output tube schematic I saw. Fixed bias: Lift ground on 2 tubes, cathode biased was a little different.

//Jibe

[erndawg]

What's the difference? Changing R41 or putting a new R in series with the bias pot?

If you add a resistor in series with the bias pot, you increase the source impedance of the bias supply. Your bias supply should be as stiff as possible; don't compromise it.

Fixing the short in the RCA connector and installing bleeders changed the bias.

It should not make the bias go down. Something is not right. This is further evidenced by your subjective evaluation of 15 - 20 watts output. At -40 VDC bias with 420 VDC on the plates, you are running the tubes hot. They should not have low output power.

Compare your amp to an original or reissue, and note the differences. If you are happy with the way your amp currently is, great. If you want your amp to perform more like a Tweed Bassman, then I think you still have some troubleshooting to go. But you're probably close.

[Diablo]

Maybe I'm a little dense, but the bias resistor R19 is 56K, just like the original Fender 5F6A schematic. If you sub in a 50K pot instead of R19, then you're a little low on bias voltage. That's why I suggested using a 39K resistor in series with your 50K pot - now you have a pot with a range of 39K-89K, and a bias voltage in the range of -46 to-54 volts. The pot and resistor are in series to ground, and are part of the voltage divider with the 15K. I don't see any way this compromises the stiffness of the bias voltage. I also checked that the max bias voltage is -63 volts at C23, before the 15K voltage divider.

[erndawg]

You're right Diablo. I just revisited that Weber schematic and had forgotten that he has the bias not taken off the wiper of the bias pot.

[jibe]

Ok!

I'm going for the 39k in series to make up for the fact that the pot comes up short.
Will be installed together with my 1% 1ohms and then I'll do the math from there.



//Jibe

[jibe]

39k resistor installed, 1% 1ohm Cathode resistor installed.

New measurements:

35.1v over Cathode resistor nr 1.
36.6 over nr 2.

0.670v over 470 screen resistor nr 1,
0.689 over nr 2 gives 1.43 mA on nr1 screen, 1.47mA on nr 2.

35.1 - 1.43 = 33.67
36.6 - 1.47 = 35.13

Plate voltage today: 432v gives Plate dissipation on nr 1 14.55 and on nr 2 15.18

bias voltage = -39.
b+: 432v.
Turning my 50k bias pot to the max doesn't give me more than 442v on the plate.
I'm not sure why my B+ dropped after fixing my short and installing 220k bleeders.

still seems a bit low, still nice powerful sound but my 30w laney feels like more output.
The speaker cab I'm using is brand new, the speakers are not broken in. Maybe that could explain something.

Is something strange here or am I chasing ghosts?




//Jibe

[Diablo]

I think you must have a short somewhere that is pulling down the B+. Your B+ dropped down for no good reason from 470V to 430V. Whatever caused that needs to be found and fixed.

[jibe]

Could I have broken something when soldering 220k bleeders? How do I test if the power capacitors are working ok? Maybe I should lift the ground of the bleeders to see what happens.

I saw you had -63v before the 15k resistor in the bias circuit. I'll check that to see what I get. -39 bias voltage seems on the low side.




//Jibe