Hmm, sounds like I've got some research to do.
Going back to the original topic of the post, it's now looking like the big regulator will be 701a based, with a 6GW8 CCS loaded by a 6CL6. The smaller regulators will use a 4D32, driven by a 6BM8. Whether they will also be CCS loaded is in the air. I suppose the only thing to do now is build and test.
Going back to the original topic of the post, it's now looking like the big regulator will be 701a based, with a 6GW8 CCS loaded by a 6CL6. The smaller regulators will use a 4D32, driven by a 6BM8. Whether they will also be CCS loaded is in the air. I suppose the only thing to do now is build and test.
See if you can find the schematic for the Tek 504 (and possibly 503) scope from the late 60's, early 70's. It had a SMPS made with a 6DQ6 vacuum tube. I fixed these things when I worked in the cal lab at a Motorola plant from 1975 to 1984. The 6DQ6 was the primary failure item, usually low emission causing a dim display. The production line equipment ran 24/7. Yes, my "crash cart" to deal with factory shutdowns contained vacuum tubes, a 12AX7 for the VTVM's and a 6DQ6 for the scopes. We also used about 100 HP200AB audio oscillators, and we had to swap 5AR4's, but not often enough to keep them in the cart.
It would be possible to build a buck converter using vacuum tubes. It would be more efficient than a linear room heater, but not as efficient as a mosfet buck.
Yes, they would need to turn rather slow for 60 Hz. Not sure how much voltage it would produce at that speed. Early (60's 70's) GM alternators were Delta wired, while Ford and Chrysler were Wye. A GM alternator will make 120 VDC when the engine is at about 2500 RPM (alternator spins 3 to 5 times faster) and the field is wired directly to 12 volts. I used to run power tools this way from my 1949 Plymouth. It had a GM alternator and a 12 volt battery.
If you were building a purpose built power supply for an amplifier you would use a frequency well above the audio band to reduce the transformer size and eliminate that pesky 60Hz hum forever. You would also try to get as close to pure square waves as possible to improve efficiency and reduce the need for output filtering.
There have been several attempts to make heater supplies for DHT's that supply pure sine waves at a high frequency. There are two ways to do this. One involves a high frequency (50 to 200 KHz) "audio" amplifier feeding the filament directly or through a small ferrite transformer. The other involves a digital pulse train with strategically placed missing pulses to approximate a sine wave. It is then low pass filtered into a good sine wave.
Now, back to the regularly scheduled big a$$ linear regulator broadcast.
I wasn't really thinking of individual power supplies for amps, rather a tube based replacement for a rotary converter. Convert 240v single phase to 4kv single phase, rectify, feed big power triodes, 3 phase ac out to step down transformers for 480v or so. If doing so for an individual amp, I would indeed increase the frequency considerably.
My tektronix 543 and 545 both use a roughly 50khz oscillator to run a transformer, which has the hv winding for the crt, and a few single turn windings for the rectifier filaments. They are red directly with hf ac.
If I go high frequency, I'd stick with single phase, since filtering hf is easier than 60Hz, 3 phase would be unnecessary. I eventually plan to have 3 phase at home, so I don't plan on focusing too hard on switching power supplies for individual devices.
Also, it sounds like you've had a very interesting life!
My tektronix 543 and 545 both use a roughly 50khz oscillator to run a transformer, which has the hv winding for the crt, and a few single turn windings for the rectifier filaments. They are red directly with hf ac.
If I go high frequency, I'd stick with single phase, since filtering hf is easier than 60Hz, 3 phase would be unnecessary. I eventually plan to have 3 phase at home, so I don't plan on focusing too hard on switching power supplies for individual devices.
Also, it sounds like you've had a very interesting life!
Last edited:
It has occurred to me that since b+ was increased to 800v, the 350v regulator won't be able to keep up without exceeding plate dissipation ratings. Also the 300V regulator would only have a 20mA load max.
The 300V regulator supplies the preamp and phase inverter tubes. 3x 5751. Could these be safely run on 350v? Rated max is 330v.
If so, I need 200mA 350v for preamp, inverter, and driver (4x 6SN7GTB). Considering I need to drop 450v at 200mA, that's 90w plate dissipation. So I could eliminate the 300V regulator, then have 2 parallel 4D32s in the 350v regulator, correct?
The 300V regulator supplies the preamp and phase inverter tubes. 3x 5751. Could these be safely run on 350v? Rated max is 330v.
If so, I need 200mA 350v for preamp, inverter, and driver (4x 6SN7GTB). Considering I need to drop 450v at 200mA, that's 90w plate dissipation. So I could eliminate the 300V regulator, then have 2 parallel 4D32s in the 350v regulator, correct?
Last edited:
I think you're overestimating the current you need for the preamp stage, the 6SN7 has a rated MAXIMUM cathode current of 20mA per section, however 5 to 10mA is a more realistic figure.
The 12ax7 you mentioned take up about 2-3mA per tube.
This leaves you with a maximum of about 90mA, note that long tailed pair phase inverters will have a nearly constant current draw, so RC filtering is not out of the question and perhaps cheaper as well.
You could feed the 350V regulator with the regulated 600V from the main regulator. in that case you can manage with about 25W of plate dissipation, even lower if you soak up some of the power in resistors as well.
Note that 12AX7 can be run safely from a +350V supply, look at the Philips datasheet for the ECC83 and you will find a chart with cathode and anode resistors.
The 12ax7 you mentioned take up about 2-3mA per tube.
This leaves you with a maximum of about 90mA, note that long tailed pair phase inverters will have a nearly constant current draw, so RC filtering is not out of the question and perhaps cheaper as well.
You could feed the 350V regulator with the regulated 600V from the main regulator. in that case you can manage with about 25W of plate dissipation, even lower if you soak up some of the power in resistors as well.
Note that 12AX7 can be run safely from a +350V supply, look at the Philips datasheet for the ECC83 and you will find a chart with cathode and anode resistors.
Regulated b+ is 800v, not 600v. So at 100mA, 450v has to be dropped. That's 45W, we're back to one 4D32, but with no extra headroom. I can live with that.
At this point no filtering is needed since it's from a regulated supply, so I'd just be looking at dropping 450v. If the 350v preamp, inverter, and driver all had a constant current draw I could use a resistive divider, but since it isn't I'm using a regulator.
At this point no filtering is needed since it's from a regulated supply, so I'd just be looking at dropping 450v. If the 350v preamp, inverter, and driver all had a constant current draw I could use a resistive divider, but since it isn't I'm using a regulator.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.