Tuesday, December 30, 2014

Biasing A Mesa Triple Rectifier.




I recently got a Mesa Triple Rectifier head in for a retube and general checkover. A little clarification is in order if the gibberish I have seen in researching this issue is any index.

Bias voltage is a negative potential applied to the grid of a tube, in this case power tubes,

A "fixed bias" amplifier only means that there is an external bias supply for the power tubes, and this can be obtained by means of a dedicated winding on the power transformer suitably rectified and adjusted, or by taking a little bit from one side of the power transformer high voltage and doing the same thing here-which means you have high tension in places where maybe it ought not to be. But nevermind. Fixed bias can be and often is adjusted with a potentiometer or, as in Mesa's case, with a fixed resistance value.

The other way is by cathode bias which means that a suitable resistance is placed in the cathode string of the power tube or tubes and this creates the desired potential. About this we need say no more.

One thing to do here is to measure the bias current and plate voltages in both tube rectified and silicon rectified modes before changing out the tubes unless there is other damage that prevents this. It gives you a basis for comparison, and there is a pretty wide spread between the two sets of values. We do not want to exceed the rated value in the silicon rectified mode, in case someone wants to use that and flips the switch. The tube rectified position will just have to be what it is going to be, and it will be significantly lower.

So....what to do?

You can install an adjustment pot in place of the 82k bias resistor which has the green test lead attached to it in the upper image, adjust and be happy, but you will have to remove that resistor and unless you remove the circuit board to get access to the underside you can damage the traces on removal and not know it.

You can also, as I did, take a different approach. I took some test leads and paralleled the bias resistor. The other end of  the resistor is the black wire on the 6L6/EL34 switch. Then, selecting a suitable resistance between 20k and 100k you can look for your sweet spot.

In my case a 39k resistor in parallel produced about 30 ma in silicon rectifier mode-which is good-and about 10 ma in tube rectified mode, which is not so great but will have to do.

I got out my magnifier, trimmed the 39k resistor and turned the ends into hooks and soldered it in.

Close enough but as I mused over the subject I got to thinking about the rather wide spread in bias values between tube rectified and solid state rectified it occurred to me that I did have three nice new Shuguang 5AR4 rectifier tubes that were worth trying in place of the 5U4GB Sovtek/Mesa tubes installed.

Well. Wonder of wonders. The 5AR4 rectified plate voltage was close to solid state, and the bias spread was only about 7 ma. It seems that the idea was good but the execution of the original was flawed. Now, I could get both plate voltage and bias schemes within spitting distance, stay out of crossover distortion country, and have the best of both worlds. I highly recommended it to the owner.


1 comment: