Plate voltages in a AB763

[überfuzz]

A friend of mine asked me why there are tiny differences in the plate voltage of the different pre.amp tubes. Well, my take on this is, ask the pros of MEF.

Anyone care to shed some light?

[eschertron]

A friend of mine asked me why there are tiny differences in the plate voltage of the different pre.amp tubes. Well, my take on this is, ask the pros of MEF.

Anyone care to shed some light?

Well, the question is worded broadly enough that, while not a pro, I may answer something reasonable or at least not wrong.

Are these tiny differences 1/100V or 1/10V or what?

There are differences that can be attributed to cathode loading, for example. some valves are biased by a 1500R cathode resistor, while others are ganged together on 820R resistors. 820 is not exactly half of 1500, so the bias voltage created will not exactly match.
There are acceptable tolerances to be figured as well. not every resistor marked 100K will measure the same (Same goes for other values, too!). Tubes can have transconductances that differ from each other and from the data sheet. This will affect the operating point as well.
The plate voltage is a divider between B+ and the tube plus cathode component. The effective resistance of the tube depends (per above) on the cathode bias point and actual tube characteristics. Using a tube data chart, the load line and operating points for different cathode resistances can be figured. If the measured plate voltage is off from the data sheet, I'd blame the tube for not matching the idealized spec.

Does this help?

[JoeM]

Are you referring to differences among the AB763 amps? AB763 could refer to for example a BF Twin Reverb or a BF Deluxe Reverb.

[überfuzz]

I'm referring to the different plate 'supplies' on a signel AB763 board. Different supply in the sense that there are different voltage levels from the PT+rectifier.

[Enzo]

Someone else can handle the theory, but I'll comment on terminology.


When we say "plate voltage" we mean the voltage right on the plate of a tube, not the power supply voltage for that tube.

The high voltage is also commonly called the B+. It comes from the rectifier, where there is also the first filter cap. We call that cap the reservoir cap. As we flow through a resistor to the next filter cap, and then again to yet another, we call each cap junction a node. Often with a name, like the screen node and the trem/reverb node, or some other descriptive. That is usually based on the part of the circuit that node serves. An alternative to that would be many drawings that name the nodes, so we might see on one drawing: B+, B++, B+++. ON some drawings they might have letter names: A, B, C,...


Your typical preamp tube triode will have one of the lower B+ voltages, ie a later node in the string. If that node is say 300v, then the plate voltage on the tube is more likely to be in the 100-200v range.

[Tooboob]

If you're talking plate voltage, that 100K plate resistor changes the plate voltage 100V for every milliamp of current difference. Variation in bias voltage, which the tube specs only wave a hand at, can account for considerable differences at the plate. And having different values of cathode resistance or plate resistance makes these changes intentional. I sometimes think that changes in operating point due to variation in bias voltage from tube to tube is the biggest reason why people can change a 12AX7 for a 12AX7 and hear God or demons.

No insult intended to the tube sniffers. TungSol tubes really are redolent of cherries.

[überfuzz]

ON some drawings they might have letter names: A, B, C,...

Yes you're quite right I've seen that to. In the filter of many tube amps there are different levels of voltage supply going to the boards. Like Enzo wrote sometimes these nodes are called A B C etc. This is what I (tried to) asked about. Why are many circuits designed like this? Why no just go with the same supply for all. The differences in voltage are often so low that they are within 10 to 20% which in turn is the uncertainty of many circuits.

Edit, thanks for the lingua Enzo, English is not my native tongue.

[Chuck H]

Yes you're quite right I've seen that to. In the filter of many tube amps there are different levels of voltage supply going to the boards. Like Enzo wrote sometimes these nodes are called A B C etc. This is what I (tried to) asked about. Why are many circuits designed like this? Why no just go with the same supply for all. The differences in voltage are often so low that they are within 10 to 20% which in turn is the uncertainty of many circuits.

Edit, thanks for the lingua Enzo, English is not my native tongue.

The different nodes provide isolation for stability and noise rejection. The voltages drop proportional to current for the different circuits from each node of the power supply. Decreasing, of course as the nodes recede from the HV source, but the actual plate voltages vary more relative to individual circuit current.

[exclamationmark]

I don't think the voltage differences are intentional, but were born out of necessity. In order to prevent motorboating and other oscillations, you usually only have 2 triodes connected to the same filter nodes. Filters are usually implemented by RC filters, which will incur a voltage drop if any DC current is drawn from them (ie, triodes drawing quiescent current).To provide even more filtering to input stages, these filters are usually cascaded, which leads to a bigger voltage drop. Usually as you head towards the phase inverter, the voltage at these nodes will start to get slightly higher.

[eschertron]

I don't think the voltage differences are intentional, but were born out of necessity. In order to prevent motorboating and other oscillations, you usually only have 2 triodes connected to the same filter nodes. Filters are usually implemented by RC filters, which will incur a voltage drop if any DC current is drawn from them (ie, triodes drawing quiescent current).To provide even more filtering to input stages, these filters are usually cascaded, which leads to a bigger voltage drop. Usually as you head towards the phase inverter, the voltage at these nodes will start to get slightly higher.

What he said.

The voltage at the plate of a tube can vary from almost zero (well, a low number) to almost B+ for a highly-amplified signal. This would be a signal that's clipping at the limits of headroom. The voltage here is going to wiggle the voltage at the other end of the plate resistor around a little, despite what the filter capacitor wants to do, which is keep the voltage at that B+ node constant. That's why the filter caps other than the first (reservoir) cap are called de-coupling caps, they reduce the interaction between the voltages among different filter stages.

And while the B+ voltages at the beginning of the signal chain are generally lower, that's OK because the signals are smaller, and don't need as much headroom. In fact, if we WANT stage 2 to clip, then we WANT the B+ voltage at that node to be low enough at that stage to reach max headroom. I followed another thread that discussed taking the preamp voltages of a Marshall amp (2204 I think) down by 100V in order to emphasize the preamp clipping.

[Gingertube]

A picture is worth a 1,000 words - but te picture is'nt great without some words.
Words and Pictures by the Wiz here:
http://www.valvewizard.co.uk/smoothing.html
Cheers,
Ian