Voltage doubling rectifier

[dwinstonwood]

So, I was reading about voltage doublers in Blencowe's Power Supply book and started thinking about the Marshall 1959 MK II rectifier.

Has anyone used this rectifier configuration to make two separate B+ voltages for anodes and screens? I'm just wondering whether it makes sense for certain tube types that have lower screen voltage ratings.

Thanks!

[sluckey]

I've seen a few amps that use that circuit to run the screens at half the plate voltage. That 1959 MK II does not use that circuit though.

[dwinstonwood]

That 1959 MK II does not use that circuit though.

Thanks sluckey. Other than the two .22uF caps before the diodes, what is the difference in the rectifiers?
Yes, the Marshall does not use the 1/2 voltage tap between the two reservoir caps, but otherwise aren't they the same?

Thanks!

[WimWalther]

That's just a standard circuit using a bridge rectifier with a CT transformer to produce a pair of complementary supplies of opposite polarity. This is how we created the driver +325V / -325V supplies in all of the audio amps we built at my previous employer.

The design requires a bipolar supply as the output section bias is derived from the cathode circuit of the direct-coupled driver tube. It also provides plenty of voltage to run a differential cascode voltage amp with an active CCS.

[sluckey]

The circuits are the same. Just a FWB that uses the PT center tap to force the voltage to split evenly across the two series caps. (You've probably seen a lot of circuits using two 220K resistors to do the same.)

[dwinstonwood]

...to force the voltage to split evenly across the two series caps. (You've probably seen a lot of circuits using two 220K resistors to do the same.)

Yes, I guess I have. I just never stopped to think about it the way you worded it there. 

Another mental project I'm working on is the effect that screen voltage has on power tube performance (or, a preamp pentode for that matter). Chisel away, bit by bit...

Thanks

[WimWalther]

Another mental project I'm working on is the effect that screen voltage has on power tube performance (...)

If you have a HV power supply, it can be interesting & informative to plot DC curves for various pentodes at different screen / plate voltages. Try to find an older tube-era supply that furnishes a full set of A, B & C supplies - like those made by Heathkit, Lambda and others.

Also, check out a few power pentode / beam tube data sheets. Some of them have data & curves for triode mode, where G2 is connected to A.

[dwinstonwood]

Some time ago I had read this 12 1/2yr-old thread: https://el34world.com/Forum/index.php?topic=12281.0

Apparently, it sat somewhere in the back of my mind, still ticking away.

So, there's a condition on the 6550 datasheet where 600 anode volts and 300 screen volts produce 100 watts form a fixed-bias pair: https://frank.pocnet.net/sheets/049/6/6550.pdf

[pdf64]

Actual voltage doubling rectifier and split rail operation https://el34world.com/charts/Schematics/files/Musicman/Musicman_2100_75.pdf

[HotBluePlates]

Another mental project I'm working on is the effect that screen voltage has on power tube performance ...

Take a look at the 6L6GC data sheet, top of Page 6.

"E_c1" is "G1" or the "control grid."
"E_c2" is "G2" or the "screen grid."

Here, E_c1 is at 0v which we expect to coincide with the moment of peak plate current when we're driving this tube to maximum clean output power.

There are a series of curves for different values of "E_c2" answering the question, "What happens when we change the screen voltage of the 6L6?". We see that as screen volts increase, plate current increases.

Plate current also increases when we increase "E_c1" the control-grid voltage, as we see from the graphs on Page 7 of the same data sheet.
   - I notice the plate current on the graphs of Page 7 ranges from near-zero to about-300mA as "E_c1" varies from -60v to -4v: a 56v change.
   - I notice the plate current on the upper graph of Page 6 ranges from near-zero to about-300mA as "E_c2" varies from 0v (or 50v) to 350v: a 300v or 350v change.

   - I conclude changing the screen-grid voltage changes the plate current, but slower than changing control-grid voltage.
   - Over time I've come to realize the screen voltage of power tubes is mostly about the peak plate current we want the output stage to achieve.