Forced air on a 833A never hurts. There are heat radiating connectors (finned) for that tube on surplus sites or ebay... The plate seal usually is the hot one that destroys the tube...
Didn't look closely enough. Small screen here. I have worked on 1 kw AM broadcast transmitters (a few still running out there) that use the 833A...these transmitters used two 833A in push pull modulator and two more for the RF final...the final tubes ran at 2500 volts at 560 ma (for the pair) and the modulator current was variable depending on audio level...the HV supply used two 575A (or 872A) mercury rectifiers and was rated for one ampere continuous. A fan blew air into the transmitter from the top, over the plate and grid seals, and another small fan cooled the two filament connections. The seals were the weakest part of that tube. Loose connections at the filament caused trouble, and the plate dissipation was obvious.
I'll look again at the pics on another machine with a better screen. Glad you enjoy the 833A. It is a workhorse...
I'll look again at the pics on another machine with a better screen. Glad you enjoy the 833A. It is a workhorse...
Do you have experience with tube circuits? Voltages in this range are not for beginners. Mistakes could be fatal.
AStouffer's comment is correct. Mistakes can be fatal. Read the warnings at the top of the tubes/valves section carefully. Better find out the insulation voltage of that Lundahl output transformer and make certain it is at least double the plate voltage in use.. Some of those transformers are bifilar wound. You do NOT need a primary to secondary short and wind up with HT on the secondary winding, which can kill someone...
Does anyone know what the x-ray emissions are for these 833 amps? At 1kV and above, the soft x-rays are not so soft anymore (their wavelength goes below 1nm, so penetration increases readily and won't be stopped by glass alone)
Recheck this. It takes at least 20 kV for electrons to hit an anode hard enough to make x-rays. The old cathode ray color televisions were carefully voltage regulated to try to prevent this problem but one or two kV will not make an x-ray. If you want, you can calculate the kinetic energy of an electron accelerated through one kV or one thousand electron volts (electron charge times 1000 Volts) and compare it with the energy of the softest x-ray or even UV photon (multiply its frequency by Plank's Constant) and you will see the proof you have nothing to worry about.
A quick Google search for "lundahl 2765" brings up discussions of headphone amplifiers with current ranges from 30 to 60 mA. The 833A could set most headphones on fire without effort since I was getting 200 watts from mine. It takes 100 watts of power just to light the filament on an 833A, and another 200 to 500 watts at kilovolt levels for the plate supply. A headphone OPT is not suitable for the OPT or even a driver due to the grid currents involved with a tube designed to run in class B.
OPT's for this tube are difficult to wind due to it's high Rp. I paid a well known OPT winder some $$$ to make me one, and it was not quite "HiFi" quality due to parasitics, with a resonant notch about equal to a Hammond 1628SEA.
Clearly some lack of understanding is present here, but to answer your question, I ran mine at 270 mA on 1500 volts. Typical currents range from 200 to 400 mA depending on B+ voltage and expected power output. I would select a different tube for applications below about 800 volts.
This kind of project is not for an inexperienced builder. Even with careful attention to the voltages present and proper HV safety, a project like this has a low probability of success due to the magnitude of just about every variable involved.
Anyone contemplating such a build needs to follow the story here. It describes a long, expensive, big amp build that failed after initially springing to life, apparently because his custom OPT builder did not fully understand how an OPT of this magnitude must be internally insulated for 5000 to 10000 volts!
High voltage driver for AB2 operation GU81m tubes
Last edited:
I think mine was a case of beginner's luck. The Lundahl 2765 made for single ended use is good for a 45 triode which typically only draws 20 ma short of the 60 ma if you get the one made for SET operation. Take the 5k between the B+ for the 45 and the 45 plate and connect the 600 Ohm output between the 833A grid and ground. I only put about 1kV on the plate of the 833A and with zero grid bias it draws about 120 ma which is less of a challenge for the Hammond 1642SEA which is good for 400 ma. With very strict DC on the 833A heater I used a crude way to estimate whether it was reasonably linear. I used 9 Volt battery between grid and ground and reversed polarity and got the same increase (for + bias) as the decrease for - bias.) My instruments for measuring are limited and I went by how it sounds and it sounds more like a live orchestra that any other amplifiers I tried. Distortions, depending on what kind they are, might not be something most people can hear. At 1000 Volts and 120 ma this is less powerful than other less flamboyant choices of tubes but is is more than enough for Magnepan speakers.
If anyone is considering building an 833A SET it would be ideal for them to visit someone who has one working and give it a listen before deciding. In audio there is no such thing as a universal best amplifier design. No two people will have the same preferences and I try to respect that. Mine sounds ideal to me.
If anyone is considering building an 833A SET it would be ideal for them to visit someone who has one working and give it a listen before deciding. In audio there is no such thing as a universal best amplifier design. No two people will have the same preferences and I try to respect that. Mine sounds ideal to me.