300vdc plate, 19vdc heater psu

[b3groover]

I think the above schematic is that Navy book's penultimate tube based B+ power supply. It uses everything the book covers in Chapter 3. It is overkill.

I was looking at that Leslie 21H power supply again and that seems very simple. I like the idea of using a tube rectifier but I'm not stuck on it.

Here's something I worked up in PSU Designer:

Am I on the right track?

 

[adamasd]

you are getting closer, but with a full wave bridge rectifier on a 300-0-300 transformer you are going to get more then 300 volts. somewhere in power supply designer it should show you output voltages. With that setup you are going to over shoot your 300 volts by quite abit.

adam

 

[b3groover]

You're right. Here's a refined one:

According to the program, under constant current, at I1, the voltage readings are as follows:

max: 300.03 | diff: 300.06 | mean: 287.62 | RMS: 289.90

Is it the max that matters, or the mean?

 

[mcs]

[quote author="b3groover"]I like the idea of using a tube rectifier but I'm not stuck on it.[/quote]
So just use a tube rectifier. Try using a 600Vct transformer, a GZ34, a 47µF cap, a 5H choke and a 220µF cap. That should work fine. You can also replace the choke with a resistor. PSUD will tell you when/if you make the resistor too small. If you want a cheaper tube rectifier, you can also use a pair of 6AX4s or similar damper tubes.

Best regards,

Mikkel C. Simonsen

 

[PRR]

> something I worked up in PSU Designer:

While PSD will let you design a resistor-input rectifier and filter, it is usually not a good plan. Voltage drop is large, filtering not great. And this plan is probably a lot of ripple for a single-ended mike amp. (And that 300V-19V mike-amp does not eat anything like 100mA.... probably more like 3-4mA per channel.)

Split the 220u cap into two 100u caps. Put one RIGHT across the diode bridge. That maximizes your raw DC voltage. Then through about 200 ohms, and another 100u cap. The second R-C stage adds a lot of ripple filtering. Fudge the R to get the right voltage. Add another R-C stage to get lower ripple.

> Is it the max that matters, or the mean?

300.03V, 287.62V, all "the same". When a tube-amp plan calls for "300V", they usually mean 270V-350V, anywhere in there. And with real-life wall-outlets and power transformers, the actual value will be 10%-20% off from your precision calculation.

In this case, the ripple (not given explicitly) looks high. Guessing that Mean and RMS are in the center, and Min is as low as Max is high, that is 4V peak-peak ripple, a lot to put into a single-ended preamp stage.

 

[b3groover]

Thank you, PRR! Is this close to what you're talking about?

The numbers in PSU Designer look great. Right around 300v. Nice!

Is it a big deal to use AC for filaments? Or should that be converted as well?

 

[mcs]

Your R1 (200 Ohms) looks small to me. You will not get a lot of ripple reduction from such a small resistor. A value in the 1-10k range would be better I think. Or you could just get a small choke...

Also remember that a 250V transformer will not output exactly 250V. If the transformer is small the voltage will change a lot with load. So if you want exactly 300V, you'll have to buy a transformer and measure the voltage under load.

Best regards,

Mikkel C. Simonsen

 

[PRR]

> R1 (200 Ohms) looks small to me. You will not get a lot of ripple reduction from such a small resistor.

True. With those values, I eye-ball about 10mV ripple, which is a lot for some mike-amps.

And ripple is usually more important than exact voltage. Most tube amps will "work" over 2:1 or more range of voltage, but any PS buzz comes right through barely reduced. The B+ often needs to be "clean enough to listen to".

However PSD does not make it easy to determine ripple. Try reading the "Diff" column.

This plan would be typical for a tube preamp and seems to give 0.1mV RMS ripple.

 

[b3groover]

Thank you, PRR! That looks simple enough. I'll give it a shot!

 

[b3groover]

All right, so I finally built my tube PSU using a 5Y3 rectifier and it works well. I have not tried to power up my old tube preamps with it yet. I read that a good way to do this is to use a variac to slowly ramp up the voltage. This is kind of a stupid question, but I just plug the preamps into the tube power supply and that power supply into the variac and slowly ramp up the voltage, right?

 

[Sleeper]

Thats Right :grin:

sleeper

 

[b3groover]

Thank you! I will give it a try.