I am contemplating using an RF beam power tube, the 5763, in a stereo SE amp, because I have them. In very rough terms, I would characterize the 5763 as a souped-up 6AQ5, just to give a frame of reference. Most data sheets only list RF specs for the tube, but the Brimar data sheet gives AF values as well (probably for modulator use). The problem is, I have a hard time believing the SE numbers. In particular, for the DC plate input the claimed output seems way too low. Example:
Plate V: 250 V
Plate I: 45 mA
Power out: 2.85 W
Efficiency: 25.3%
That's a plate input of 11.25W for 2.85W out, or an efficiency of only 25%. A 6AQ5 has these numbers:
Plate V: 250 V
Plate I: 45 mA
Power out: 4.5 W
Efficiency: 40%
6BQ5:
Plate V: 250 V
Plate I: 48 mA
Plate out: 5.7 W
Efficiency: 47.5%
Any thoughts on how the 5763 could have such low power output, with essentially the same plate voltage and current swings as the other listed tubes?
https://frank.pocnet.net/sheets/184/5/5763.pdf
Plate V: 250 V
Plate I: 45 mA
Power out: 2.85 W
Efficiency: 25.3%
That's a plate input of 11.25W for 2.85W out, or an efficiency of only 25%. A 6AQ5 has these numbers:
Plate V: 250 V
Plate I: 45 mA
Power out: 4.5 W
Efficiency: 40%
6BQ5:
Plate V: 250 V
Plate I: 48 mA
Plate out: 5.7 W
Efficiency: 47.5%
Any thoughts on how the 5763 could have such low power output, with essentially the same plate voltage and current swings as the other listed tubes?
https://frank.pocnet.net/sheets/184/5/5763.pdf
"The 5763 is a heater-cathode type transmitting beam power amplifier in the 9 pin miniature construction. It is intended for use in compact, low-power mobile transmitters and in the low-power stages of larger fixed station transmitters. The 5763 is particularly useful in the doubler and tripler stages of transmitters."
It's not an "audio" tube, and not used in audio.
Internal design is different, specifically for RF use.
It's not an "audio" tube, and not used in audio.
Internal design is different, specifically for RF use.
The other datasheets might show slightly higher power using different plate load.
and the trade off as usual would be more power, but higher distortion.
using lower values
assume plate loads around 4.7 K to 8 K
and distortion could be anywhere from 6% to 13%
I haven't looked at transformers for awhile
you could likely find something in the middle around 6K
or likewise with any basic ratio you can half or double the speaker load.
so technically you could have a 4k or 8k
depending on the transformer ratio and available output taps.
likely the tube can make more power, the data sheet was shooting for lower distortion
you also have plate voltage to play with, depending on regulation of the power transformer
you can raise or lower voltage or might have higer voltage around 270 volts.
Depends, if you use tube rectifier it can drop 20 to 30 volts
or if you use solid state rectifier it can raise 20 to 30 volts
and the trade off as usual would be more power, but higher distortion.
using lower values
assume plate loads around 4.7 K to 8 K
and distortion could be anywhere from 6% to 13%
I haven't looked at transformers for awhile
you could likely find something in the middle around 6K
or likewise with any basic ratio you can half or double the speaker load.
so technically you could have a 4k or 8k
depending on the transformer ratio and available output taps.
likely the tube can make more power, the data sheet was shooting for lower distortion
you also have plate voltage to play with, depending on regulation of the power transformer
you can raise or lower voltage or might have higer voltage around 270 volts.
Depends, if you use tube rectifier it can drop 20 to 30 volts
or if you use solid state rectifier it can raise 20 to 30 volts
Last edited:
Brimar gives 3 pages of data for AF use. People use sweep tubes and other RF tubes for audio all the time.
I think the difference is they expect the 5763 to have the control grid driven positive under normal use. This works for RF applications but not so much for SE.
You should be able to use them as long as you accept the slight difference in efficiency.
I have a pile of 5763s and have used them push-pull with good results. I think I used 12 tubes per channel in a car amp going for a specific look.
They may not blow up with 375 volts on plate and screen in UL use as long as you are not pushing dissipation limits.
You should be able to use them as long as you accept the slight difference in efficiency.
I have a pile of 5763s and have used them push-pull with good results. I think I used 12 tubes per channel in a car amp going for a specific look.
They may not blow up with 375 volts on plate and screen in UL use as long as you are not pushing dissipation limits.
I really don't care about Brimar or what people dream up on the internet.
The tube is, hold on.... INTENDED...... for Radio Frequency use.
It's certainly not a 6AQ5. Heater current is 65% greater; so is transconductance, and the plate curves are much less linear. Try drawing the load line on the plate curves, and you'll see why it won't make as much power - it can't deliver enough current without going into grid current.
Mike Got Lots. When NOS military stock hit the market some 20 years ago, I helped a guy advertise them on ham radio forums and he gave me a few cases of them in return.
Last edited:
I wouldn't put any stock in any of those efficiency numbers. The highest possible efficiency theoretically possible for a class A amplifier is 50%. That's for matched source and load resistances (not! impedances) and lossless coupling, etc. IOW, fictitious. Typical numbers for audio use, with matching optimized for linearity, might be half that, or even less.
All kinds of "transmitting" valves are used for audio, and have been since the 1920s. Given the current state of valve manufacturing, we're all going to have to accept that what we have is what we have, and leave the guitar amp valves for the guitar amps that need them. Lots of great alternatives for audio. No need to eat the seed corn.
All good fortune,
Chris
All kinds of "transmitting" valves are used for audio, and have been since the 1920s. Given the current state of valve manufacturing, we're all going to have to accept that what we have is what we have, and leave the guitar amp valves for the guitar amps that need them. Lots of great alternatives for audio. No need to eat the seed corn.
All good fortune,
Chris
Thanks, I'm definitely going to fiddle around with them. My scope has a waveform generator that I can use to drive one for testing purposes. Should be interesting. BTW, I remember seeing a 12AU7 used as a video amp in an old B&W TV. That 12AU7 didn't know or care what its "intended" use was.
Where do you get this idea?
The 12BY7A is an excellent pentode for audio (okay, it's a great pentode for a lot of things) and it was commonly used as a video amplifier and as the driver tube in a lot of smaller RF transceivers. That's what it was originally designed for. Most of the hybrid transceivers from the 70s used a 12BY7A to drive the 6146 output tubes.
The 6L6, incidentally, is more or less an RF tube as well. The 807 is a 6L6G with a plate cap and different base. I'm sure everyone knows that 6L6s and 807s are pretty good audio tubes. The plate cap lowers parasitic inductance, but more importantly reduces RF heating and arcs within the tube base, which proved to be a problem with 6L6s (and many other small tubes) when used in RF service.
Many of the bigger RF tubes work well at audio frequencies too, but obviously they're going to need some grid drive (good luck finding a big tube that doesn't).
I've done some preliminary calculations and modeling on a push-pull 829B amplifier with some crude bench testing to back it up, and the results look very promising. Originally the 829B was designed for use as a VHF power amplifier, but the main use of the 829B was as in hard-tube magnetron pulsers for radar applications. They proved to be quite suitable for this due to their generous cathode area and ability to hold off pretty high voltages (see Rad Lab volume 5).
LS50s (and therefore GU50s) are well-liked for audio use (though they have pretty anemic cathodes, so output transformers are a bit trickier), and they were originally an RF tube.
Oh, and one of the most beloved audio tube of all time, the 211? It's an early RF tube (1930s IIRC).
6AS7/6080? That was designed and intended for series pass elements in linear power supplies, but lots of folks use them for audio (though their $\mu$ is pathetically low) and get good results.
The point I'm trying to make is that there are a lot of tubes which were designed for one specific application (often RF) that perform very well in other applications. I've used EL34s in low frequency RF service (I never pushed them above about 500 kHz, but they very well may go higher than that). They definitely weren't intended for that use, but they worked decently well - their uncommitted suppressor grid is a nice touch.
When re-purposing hardware, it's always a case-by-case basis. There's s*** like the 845 that doesn't perform particularly well in any application, and there are beauties like the 715 that are spectacularly good for one thing and one thing only (pulse generators - see Rad Lab volume 5). There are also tubes like the 813 that work great for a lot of different applications.
A 'C-' engineer makes their decision based off the blurb on the first page of the datasheet. An 'A-' engineer looks at the plots before making that decision. An 'A' engineer looks at the plots, checks an app note and runs few back-of-the-envelope calculations before making a decision.
When the datasheet of a generic-looking tube says that it is intended for RF application, or even explicitily says that it is not intended for audio, I believe that they really meant that it is not screened nor optimized for low hum and microphonics, and they haven't spent any significant amount of time characterizing the audio amplifier application case, so it is up to the user to find the best working point for audio. I am currently using a trioded wired EF184 frame grid pentode on my headphone amplifier. It works wery well despite it was never meant or sold for audio use. But I had to test 20+ tubes to find two matched ones without any background noise or microphonics. I'm sure this screening would have been redundant if I had fitted a EL84.
I noticed that the curves for the four cases of tetrode connection of the 5763 in the Brimar Application Report (see post #1) show the screen grid currents at the stated power outputs to be 8.8 to 9 mA. With the screen grid voltages being (almost) 225 V, the screen grid dissipation is very close to the absolute maximum screen grid dissipation of 2 Watt.
So I would think that the screen grids are the limiting factor when it comes to maximum audio output power.
So I would think that the screen grids are the limiting factor when it comes to maximum audio output power.
My 833C transmitter tube monos are the best-sounding amps I've ever heard. My experience is that they love HIGH voltage (I run at 2300V B+) and can make up to 40W in pure class A1. I get up to 200W when A2 kicks in.
So...if you're not hesitant to crank up the voltage you can get stellar results from transmitters. It can be a bit unnerving at times, but 8 years after the build I'm extremely glad I took the plunge.
You can read more here in the build thread if you're curious:
https://www.diyaudio.com/community/threads/the-midlife-crisis-my-833c-amp-build.232484/
So...if you're not hesitant to crank up the voltage you can get stellar results from transmitters. It can be a bit unnerving at times, but 8 years after the build I'm extremely glad I took the plunge.
You can read more here in the build thread if you're curious:
https://www.diyaudio.com/community/threads/the-midlife-crisis-my-833c-amp-build.232484/
Last edited:
For a comparison of the 5763 to say a 6BQ5, a good starting point is to compare the plate curves
at the same screen voltage. What is the max current in the knee at zero bias?
The 6BQ5 is clearly better in that measure. So the 6BQ5 will will deliver more audio when properly
loaded.
at the same screen voltage. What is the max current in the knee at zero bias?
The 6BQ5 is clearly better in that measure. So the 6BQ5 will will deliver more audio when properly
loaded.
Attachments
Over 20 years ago someone gave me over 100,000 loose, unsorted, untested tubes. Over a 10 year period I went through all of them throwing out the trash (broken, corroded, or otherwise unusable tubes), sorting and testing a few of each of all the type numbers that I had never seen before. Ten years ago it became apparent that my 41 year engineering career was on life support and that we would eventually move out of Florida when it ended. That began a massive "stuff" reduction program. I sold, traded or gave away lots of tubes and only kept those that I thought I may use someday. I have been in West Virginia for 8 years now, and some of my guesses as what to keep were a bit off, but most were good. For some reason, I have about 1 gallon of 5763's. I probably lit some up and leaned on them pretty hard.
I believe the clue to sucess here is in the spec sheets. 25 watts from a pair in an AUDIO amp. OK, but it's in class AB2. Daqvin Carter picked up on this. You will need to drive G1 positive to get good efficiency from this tube. I use mosfets for this, but to each, their own.
I believe the clue to sucess here is in the spec sheets. 25 watts from a pair in an AUDIO amp. OK, but it's in class AB2. Daqvin Carter picked up on this. You will need to drive G1 positive to get good efficiency from this tube. I use mosfets for this, but to each, their own.
