Dome high efficiency audio amplifier

[John K5PRO]

Has anyone (besides R. Dome) tried this? In a 1959 IRE Transactions on Audio paper I found online, Mr. Dome of GE says he could take a pair of 810 triodes in push pull class B audio, and get 537 watts of audio. After adding his auxiliary square wave signal at high frequency to this, he raises the output to 750 watts, while not exceeding the rated plate dissipation of each tube (125W). He didn't discuss what frequency his aux signal was, but that it was super-audible and must be terminated in a high frequency load in the output to speaker.

Hope this link works for all:

http://ieeexplore.ieee.org/iel6/8335/26275/01166188.pdf?arnumber=1166188

[Tom WA3KLR]

I haven't read the link or any other stuff by Dome, but from what you stated John, it sounds like an early implemention of what we now call a Class D audio amplifier.

[Steve - WB3HUZ]

Not quite. It's more like Pulse Amplitude modulation than Pulse Width/Duration modulation.

I haven't read the link or any other stuff by Dome, but from what you stated John, it sounds like an early implemention of what we now call a Class D audio amplifier.

[WA1GFZ]

Huz,
Are they goosing the plate voltage during the peak of the waveform?

[Steve - WB3HUZ]

No.

Huz,
Are they goosing the plate voltage during the peak of the waveform?

[John K5PRO]

Well, not exactly. He put a higher frequency square wave ON TOP of the audio sinewave,so that the peak excursion (p-p) was larger, but so that there are also zero times where the amplitude is low. Ran this through a series network, the audio load, and a HF load consisting of an inductor in parallel with a resistor (like a parasitic choke). From this, claims of 10-20% efficiency improvement.

[WBear2GCR]

Can anyone email me a copy of the actual article?
I can get it by traveling to the engineering school's library, but I don't get up that way very often these days. Am curious.

Is he modulating the HF signal with input signal too??

                  _-_-bear

[Steve - WB3HUZ]

http://ieeexplore.ieee.org/iel6/8335/26275/01166188.pdf?arnumber=1166188

[WBear2GCR]

That's just the abstract.

If you want the full article you have to pay...


            _-_-bear

[Steve - WB3HUZ]

Hmmm. I was able to DL the full article. I'll send it to you.

[WA1GFZ]

sounds like too much work why not just do class D and get it over with.

[WBear2GCR]

Well, it is tubes and it ain't class D... (which still sounds like death in hi-fi apps...)

I read the article 1.5 times and still don't exactly see what he is doing, or more precisely what the mechanism is that he has working... I'll figure it after a few more passes. The figures he provides are suitably vague...

Thanks for passing it along...

Hmmm...   

              _-_-bear

[WBear2GCR]

Hey, I've read the thing three times now - so far I don't get it at all.
You run *more signal*, the audio signal plus a full amplitude ultrasonic square wave and you have less total dissipation? Huh?

I know I am missing something here.

In the class A illustration it looks like maybe he is negative amplitude modulating a sort of static PDM signal... but in the class B illustration it looks like modulating a carrier, but with peak clipping!!

The back to back diode "drive circuit" also leaves me scratching my head a bit...

Obviously, I'm not understanding something at all...

Signed,

           confused, not in peoria

[K1DEU]

Some people do not notice that tetrodes in common cathode configuration and crontrol grid driven (only) run a very cool, low distortion class B. It isn't new on the AM broadcast band where I saw a pair of 828's modulating 810's in a GE 500 watt output box. So basically use 20 Db. of feedback and increase your bias until you see/hear cross over distortion and back it down a bit. Your quiescent/idling plate current will be so low you are definitely Class B with very little drive!

   I do this in my 200 watt output DX-100 for the 36LW6 modulator tubes which have 1075 VDC on the plates and a poorly regulated 150 (on purpose) on the screen grids (tru 1,000 resistors). 73  John

http://hamelectronics.com/k1deu/pages/ham/transmitters/am/pages/modulation_power_mods.htm

[Steve - WB3HUZ]

Yup. The GE BT-20 ran I pair of 828s as modulators. The driver was RC coupled, but the resting currently was something like 25 or 50 mils per tube - very low for an AB1 operation. I'm thinking you could do something similar with 814s too.

Bear, the arrangement in the Dome amp is more like Pulse Amplitude Modulation (PAM) than PDM. In PDM, the peak level is constant and the pulse width changes. In PAM, the peak level changes (just like normal AM) and the pulse width is constant. And this is exactly what happenns in the Dome amp. I didn't see any clipping anywhere.

Some people do not notice that tetrodes in common cathode configuration and crontrol grid driven (only) run a very cool, low distortion class B. It isn't new on the AM broadcast band where I saw a pair of 828's modulating 810's in a GE 500 watt output box. So basically use 20 Db. of feedback and increase your bias until you see/hear cross over distortion and back it down a bit. Your quiescent/idling plate current will be so low you are definitely Class B with very little drive!

   I do this in my 200 watt output DX-100 for the 36LW6 modulator tubes which have 1075 VDC on the plates and a poorly regulated 150 (on purpose) on the screen grids (tru 1,000 resistors). 73  John

http://hamelectronics.com/k1deu/pages/ham/transmitters/am/pages/modulation_power_mods.htm

[K1DEU]

        The GE BT20A  500 watt output transmitter had 828 tetrodes. Control grid driven with a standard push pull  primary, split secondary  audio driver transformer for the cross connected negative feedback ladders. The Common B+ was 1500 VDC and the Class B idling current was 20 ma. per 828. Nearly biased to cutoff, definitely not AB1, manual also states Class B. The push pull drivers were 6J7/6SJ7's.

        Many grid driven tetrodes from 6L6's, 807's, 6146's, 813's on upward to handles (on the anodes) work very well Class B with a good push pull driver and 20 Db. of negative feedback. 73  John

[WBear2GCR]

<snip>

Bear, the arrangement in the Dome amp is more like Pulse Amplitude Modulation (PAM) than PDM. In PDM, the peak level is constant and the pulse width changes. In PAM, the peak level changes (just like normal AM) and the pulse width is constant. And this is exactly what happenns in the Dome amp. I didn't see any clipping anywhere.

Steve,

Yeah, I wanted to say PAM, but I didn't think it was a legit mode!

Still do not see how it makes the amp have any more capability exactly, but I'll think about that some...

                    _-_-bear

[Steve - WB3HUZ]

It may have been Class B but the drivers were RC coupled. So, the 828 couldn't have pulled much grid current. IIRC, older handbooks showed listed the 828 at about the same parameters and called it AB1. Dunno.

The drivers in the BT20A were 1620s which is a low noise version of the 6J7. The RF power output was 250 watts, although I could get 300 out of the one I ran on 160 meters.

I often wondered why similar set ups weren't more prevalent in amateur circles.

        The GE BT20A  500 watt output transmitter had 828 tetrodes. Control grid driven with a standard push pull  primary, split secondary  audio driver transformer for the cross connected negative feedback ladders. The Common B+ was 1500 VDC and the Class B idling current was 20 ma. per 828. Nearly biased to cutoff, definitely not AB1, manual also states Class B. The push pull drivers were 6J7/6SJ7's.

        Many grid driven tetrodes from 6L6's, 807's, 6146's, 813's on upward to handles (on the anodes) work very well Class B with a good push pull driver and 20 Db. of negative feedback. 73  John