Unsual 807 PP Amp.

[Diabolical Artificer]

Here is a schematic of a 75W amp that uses 807's as the OP valves, that I found the other day on the interweb. The circuit is unlike any other amplifier I've seen.

I find the phase splitter arrangement odd, as instead of using a double triode as the phase splitter, it uses two 6SN7's as drivers/phase splitters; that and the NFB going to the final drivers instead of the input stage.

Lastly, the OPT secondary shows an impedance of 600 ohms. Therefore is this a PA/ guitar amp or hifi amp or something more esoteric?

Any thoughts most welcome, Andy.

[GrimJosef]

It's not a guitar or hi-fi amp Andy since they'd be designed for driving conventional low-impedance speakers. It'll be for some sort of PA use I imagine. Its performance isn't really hi-fi either - the third and fifth harmonics appear at the level of several percent each. So even with 20-30dB of NFB you'd still be left with several tenths of a percent distortion.

The reason the NFB doesn't go all the way to the front of the amplifier, by the way, is that this is really a preamp followed by a power amp. NFB (at least in its simplest form) tries to correct for any 'distortions' in an amplifier. So if it went all the way to the front of the pre it would be trying to reverse the effects of things like the volume and tone controls. To avoid this its entry point has to come after the last of the controls and correct only the power amp stages, which it does.

The arrangement with the double triodes is needed because this is a Class AB2 amp. AB2 means that the 807s are driven so hard that for part of the audio cycle their grids go positive of their cathodes. When this happens quite a bit of current suddenly starts flowing into the grids and the driver stage has to be capable of supplying that current without its output signal voltage drooping under the sudden strain. So the first half of each 6SN7 is used as half of a double-triode phase-splitter and the second half is wired as what's called a 'cathode follower'. Here the signal for the next stage is taken from the cathode of the valve. The cathode behaves like a low-impedance source, so it can supply lots of current without its voltage being affected very much. Simplistically this is because if the cathode voltage did change then that would affect the grid-cathode voltage, and the valve's current flow is very sensitive to small changes in the g-k voltage, so the effective impedance (which is the small voltage change divided by the large current change - R=V/I) ends up being very small (low).

Executive summary - this would be a LOUD amplifier, but not a very lovely one.

Cheers,

GJ

[Radio Wrangler]

Output transformer data says 12 henries primary inductance Let's assume that is anode to anode, and 4500 Ohms load to the valves is also anode to anode, the the O/P stage is 3d`b down at 4500 = 2*PI*F*12 so F=60Hz

If 4500 Ohms is per side and 12H is per side than the result is the same

if 4500 Ohms is per side and 12H is anode to anode then the LF limit moves up by a factor of 4 to 240Hz.

Both will we shifted by feedback, but it still reduces the available power at LF.

Not a lovely amplifier as it stands.

David

[m0cemdave]

That circuit has appeared, with minor variations, in various textbooks and Brimar valve manuals. It probably came from a manufacturer's application note.

The zero bias 807 amplifier in your diagram 807-43 is also well-known, but whoever drew it has got the two HT supplies mixed up...

[Bazz4CQJ]

I don't have the expertise to comment, but, if all that's wrong with it is falling off below 240Hz, how good would it be as the modulator of an AM Tx?

If we said that what we want there is efficiency and a passband from ~250 - 3000Hz (?) is this a design to do that?

Barrie

[G3VKM_Roger]

I think the 1950s Tiger TR200 transmitter used the zero-bias design shown earlier in this thread and that worked very well.

73

Roger

[Radio Wrangler]

I did a quick back of envelope calculation because the documents gave some output transformer characteristics. The LF response could be tailored to whatever LF response is needed by changing the transformer design, and also checking the timeconstants of the coupling and decoupling capacitors in the rest of the circuit. Anyone wanting to build the circuit would need to source or have made a suitable transformer anyway. The secondary doesn't have to be 600 ohms either, though a recalculation of the feedback resistors would be needed.

The bias point is going to make it fairly high in open-loop distortion, and the output transformer is going to limit the maximum possible amount of feedback without risking stability. As G-J has pointed out, the design is workmanlike, with cathode-follower drive to the output grids so they don't non-linearly load the phase shifter stage.

The phase shifter is unusual in that it's a pair of cascaded (not cascoded) inverting amplifiers rather than the usual concertina or long-tailed pair type. This has generally been seen as less balanced than the usual circuits, that the phase shift in the second inverter will spoil the accuracy of the 180 degree split at higher frequencies. I suspect that this design is intended for speech service, covering 300-3000Hz.

For music, I think I'd stick to the Mullard/Leak/Radford designs despite lower power from class AB. The 807 is an excellent bottle and has the advantage of not being overpriced. Perhaps its bass isn't fast enough.

David

[Diabolical Artificer]

I understand the circuit a lot better now thanks to all your reading of the circuit. I must admit to having a bit of a bee in my bonnet regarding 807 valves. I've been looking for a decent 807 based hi-fi amplifier design for some time since I saw the Williamson, which is a great amplifier but a little under powered for my purpose and needs a expensive, complex OPT.

The 807 is a nice looking valve, power-full and relatively cheap, as David says- though they are increasing in price. And for those of us unable to afford KT66/88's I think they have potential. Of course it has drawbacks too.

Thanks for all your input,Andy.

[Bazz4CQJ]

Originally Posted by G3VKM_Roger

I think the 1950s Tiger TR200 transmitter used the zero-bias design shown earlier in this thread and that worked very well.

Thanks for that info; I have latent desires to get an AM rig running on 80m; not imminent, but down the "To Do" list

[Radio Wrangler]

As the 807 takes the 6L6 design and adds better insulation, and isolation between grid and anode wiring, then it's a 6L6 with its characteristics extended to higher voltage. You'll find an 807 is right at home in any 6L6 circuit.

David

[Malcolm G6ANZ]

Bazz, I have an AM modulator sitting not doing anything. I built it years ago to modulate a qqvo6/40 on 2m, when you could use AM on 2. Its a pair of KT88s driven by a 6SN7GT into a Woden modulation transformer. There is also a speech processor built into it.Design from the VHF/UHF manual.
If your interested PM me.

Malcolm

[TrevorG3VLF]

I do not think that this is a modulator since it does not have a floating output winding. Would this be 100V line?

Surely, you can still use AM on the 2m band.

[G6Tanuki]

It does look like a PA/audio-distribution amplifier [the 600-ohm OPT winding is a dead giveaway - that's not really a typical impedance for AM modulator duty].

I'd suggest it was intended for public-address/background music in factories/small public-venues such as railway-stations.

A pair of 807s with control- and screen-grids strapped and used in "zero bias" Class-B [750V on the anodes, expect them to pull 250mA on peaks] was a favourite 1950s-solution for UK amateurs who wanted to run the 'legal limit' [150W DC input to the PA] and then AM it on the amateur bands.

See here:

http://n4trb.com/AmateurRadio/RCA_Ha...amtips0702.pdf

for the RCA article on how to hook up a pair of 807s as zero-bias triodes. I've used the same circuit for a 150-watt [grin] guitar-amplifier!

[Malcolm G6ANZ]

Trevor, you probably can use AM on 2M. But the actual transmitter I built to go with the modulator has long since gone. Twas a simple thing solid state VFO followed by a series of frequency multiplier stages driving a QQVO3/10 then the modulated 6/40 PA.

Malcolm

[G6Tanuki]

Originally Posted by TrevorG3VLF

Surely, you can still use AM on the 2m band.

Most definitely - whether classic "full-fat" double-sideband-and-full-carrier, one-sideband-and-full-carrier, both-sidebands-and-no-carrier, single-sideband-and-no-carrier, one-or-both-sidebands-and-reduced-carrier or independent-sidebands-and-no-carrier - they're all still permitted.

[It's only a few days since I operated a 100-watt double-sideband-full-carrier AM transmitter on 2-Metres].

[Bazz4CQJ]

Browsing through an old Brimar valve databook the other day, I stumbled upon this design for a 75W "HiFi" amp (Brimar's view not mine) using a pair of 807's. There is some additional info which I'll scan if anyone wants it.

B

[Radio Wrangler]

Odd sort of hifi that, Bazz. Lots of power but at the cost of lots of distortion. It seems GEC and Mullard had a lot better idea how to do such things than Brimar did.

David

[GrimJosef]

Yes, it does seem odd to suggest that the Brimar 807 amp is hi-fi. But they could design nice kit when they wanted to. Their 25P1 monoblock power amp, using a 6BR7, a 12AX7 and a pair of 5B/255Ms, all powered by a 5R4GY, could deliver 25W with 0.1% THD at 1kHz, hum and noise 85dB below 20W and an output impedance of 0.4 ohms at both 50Hz and 10kHz. The details are in their Valve and Teletube Manual No 8, issued in 1960 I think.

Cheers,

GJ

[barretter]

The Brimar diagram doesn't say "hi-fi", it just says AB2.