Bias cap value?

[Chuck H]

Yep, normal. In fact that's pretty tame (you must be playing softly ) The only advantage to the dual 10uf's instead of the single 22uf is slightly better filtering. Whether that's even necessary is usually debatable or a matter of individual cases. Your circuit, as it is, is pretty common and if you're not having trouble with it you don't need to change it.

Did the drift issue resolve with the pot change?

[daz]

Yep, normal. In fact that's pretty tame (you must be playing softly ) The only advantage to the dual 10uf's instead of the single 22uf is slightly better filtering. Whether that's even necessary is usually debatable or a matter of individual cases. Your circuit, as it is, is pretty common and if you're not having trouble with it you don't need to change it.

Did the drift issue resolve with the pot change?

Seems so. I changed the pot and the set resistor and biased it right at startup and it didn't seem to drift after a few minutes, at least not like it did. By the way, on a unrelated note (related to another post i made tho) I was at a electronics surplus place the other day and asked if they had any 9 pin sockets. So he says yeah, but i think they're cheap Chinese ones. He shows em to me and they're my fav...Beltons ! So yesterday i replaced all 3 preamp sockets. No more socket issues for a while.

[KevinOConnor]

Hi Guys

BIAS SUPPLY IS THE MOST IMPORTANT VOLTAGE IN THE AMP if it is a fixed-biased amp.

The actual value is unimportant PROVIDED there is sufficient bias voltage to control the tubes, i.e. to turn them off completely so you can control how 'on' they are.

When changing anything in the bias supply, you must TEST WITHOUT TUBES and verify that the bias voltage is present right at the tube socket. This checks that the entire grid control path is intact: bias supply working; bias-set network intact and working - test voltage sweep; grid-leaks and grid-stops intact; wiring generally good.

The preferred bias supply and bias-set network are separate circuits. The bias supply shown by daz combines them so that the "output of the raw bias supply" and the "applied grid control voltage" are one and the same. As Chuck stated, an extra RC of filtering is highly beneficial and is the usual way to do simple bias supplies if you want low hum. A further step towards low hum is to provide individual bias pots, which is where the bias-set network becomes more visible to most people. In this case, two pots have their X-ends wired together and their 0-ends wired together. The 0-ends go to the bias filter cap and the X-ends go to ground through a fixed resistor. The raw bias voltage is now what sits at the zero end of the pot, corresponding to zero tube current, zero tube heat and the universal zero-bias condition. The -end makes the tubes hot at possibly higher than desired idle current, but this depends on what the raw bias voltage is and what values the pots and range resistor to ground are. Each pot wiper feeds ONE grid-leak resistor for ONE power tube - or for the push and pull halves of the circuit.

To make the raw bias supply completely independent from the bias-set network, the simplest thing is to add a zener to clamp the cap voltages when the pots are not connected. The zener is only 'on' without the pots and it is not desirable to actually regulate this supply unless you use a tracking regulator.

With the individual bias pots, you set the first tube by meter to a known and safe idle current value for the specific amplifier. Set the second bias pot by ear for minimum hum, with any MV or volumes pots ste to zero.

In fixed-biased amps, the "target" control voltage for the tube grid is often around 10% of the absolute value of the screen voltage. To properly control the tube, the sweep of the bias pots should go to 15% (as the absolute value). As an example, say Vs=400V. 15% of this as a bias voltage would be 400 x 0.15 = -60Vb; must be negative to control the tube.This means that the parallel bias pots have -60V at the 0-end. There is no point allowing the grid voltage go too close to ground as that just makes the tubes very hot, so the range resistor is chosen to set the voltage at the X-end be no less than half what is at the 0-end.

The above example applies to any amp using large-bottle tubes and also for 6V6s. EL-84s require quite a bit less control voltage.

To monitor the tube current, install a 1R or 10R 1W resistor between the cathode and ground. Add a meter jack on the rear panel of the amp for each tube or for each half of the circuit, plus a third one for ground. The three meter jacks allow easy voltage measurement of these resistors, where for the 1R each 1mVdc represents 1mA through the tube. Similarly for the 10R, each 10mV represents 1mA. The bias pots should be mounted on the tube plane of the chassis for easy access.

In daz's amp, we are assuming that the plate winding of the PT has a CT that is grounded. he never stated this specifically, rather just that the bias supply is taken from one end of this winding. There are many 18W amps that use a full-bridge and "floating" plate winding rather than the half-bridge and CTed winding.

Have fun