UK Vintage Radio Repair and Restoration Discussion Forum

UK Vintage Radio Repair and Restoration Discussion Forum (https://www.vintage-radio.net/forum/index.php)
-   Homebrew Equipment (https://www.vintage-radio.net/forum/forumdisplay.php?f=99)
-   -   807 tube amp, any ideas? (https://www.vintage-radio.net/forum/showthread.php?t=178887)

nzoomed 13th Apr 2021 10:40 am

807 tube amp, any ideas?
 
Been looking at building a hi-fi amp with 807's mostly because i have stacks of the things and they look cool, especially with their blue glow.

Probably would set up in class AB/AB1 configuration at 400V, ultra linear and cathode bias.

I know there is some debate about running the screen voltages over 400V on the 807, but I know they run perfectly OK at 400V without any issues and I have built a 100W marshall clone amp using them at 400V and its poking out over 160W from a quad.

AWA Radiotron in Australia even released a technical bulletin regarding the screen voltages in this regard too supporting this.

Anyway, I was looking at using 6J7 / EF37A as the driver and a 6SL7 as the phase inverter.


Any other suggestions? Ive never built an amp from scratch before, but am looking at similar designs to the Mullard 5-10 amp.

I expect I should get a good 25-30W per channel from a push pull pair of 807's

Radio Wrangler 13th Apr 2021 12:14 pm

Re: 807 tube amp, any ideas?
 
If you use ultra-linear mode, with the screens on output transformer taps, the transformer action ensures the anode is always more positive when the screen voltage is high. The electrons find the anode more attractive and screen dissipation is kept down.

The normal screen voltage ratings on a valve are for the case of fixed voltage screens. They don't count for triode connection or ultra-linear operation.

David

Diabolical Artificer 13th Apr 2021 2:25 pm

Re: 807 tube amp, any ideas?
 
4 Attachment(s)
Personally I'd use two 6SN7's, one for the IP for both L&R channel, one for the PS. Phase splitter run from CCS and - 80v negative rail. An EF37A & 6SL7 have far too much gain unless your intended to amplify a TT cartridge.

An 6SL7 is a high gain low current valve, you want the opposite to drive an OP stage cleanly. You want lots of grunt, IE high current to drive the OP valves inter electrode capacitance. An excellent choice is an ECC82 or better a 12BH7 or a 6BL7, though a 6SN7 does the job.

Any valve amplifier depends on the OPT, get the best you can and the biggest, EG for a 20w amp get a 100w jobbie. I've attached a schematic for an 807 amp I built. Recently lent it to a mate who paired it with his sensitive speakers, he said it's as good if not better than any valve amp he's listened to, including Quad, McIntosh etc. I built it from scratch, pretty easy to do if you have time and patience.

There's a few mistakes on the schematics, they're for guidance only.

Andy.

Ed_Dinning 13th Apr 2021 7:42 pm

Re: 807 tube amp, any ideas?
 
Have a look at Geloso amps. They did a host of designs in the 50's using these valves for instrument, PA and Hi Fi use

Ed

nzoomed 13th Apr 2021 9:49 pm

Re: 807 tube amp, any ideas?
 
Quote:

Originally Posted by Radio Wrangler (Post 1363853)
If you use ultra-linear mode, with the screens on output transformer taps, the transformer action ensures the anode is always more positive when the screen voltage is high. The electrons find the anode more attractive and screen dissipation is kept down.

The normal screen voltage ratings on a valve are for the case of fixed voltage screens. They don't count for triode connection or ultra-linear operation.

David

OK, thats good to know, because I had been reading another thread where people were advising someone not to use ultra linear at a high voltage.

Actually makes total sense as you say.

In the case of my guitar amp, they are running with their own grid supply.
I will have to post some pics of that in another thread.

nzoomed 13th Apr 2021 9:58 pm

Re: 807 tube amp, any ideas?
 
Quote:

Originally Posted by Diabolical Artificer (Post 1363915)
Personally I'd use two 6SN7's, one for the IP for both L&R channel, one for the PS. Phase splitter run from CCS and - 80v negative rail. An EF37A & 6SL7 have far too much gain unless your intended to amplify a TT cartridge.

An 6SL7 is a high gain low current valve, you want the opposite to drive an OP stage cleanly. You want lots of grunt, IE high current to drive the OP valves inter electrode capacitance. An excellent choice is an ECC82 or better a 12BH7 or a 6BL7, though a 6SN7 does the job.

Any valve amplifier depends on the OPT, get the best you can and the biggest, EG for a 20w amp get a 100w jobbie. I've attached a schematic for an 807 amp I built. Recently lent it to a mate who paired it with his sensitive speakers, he said it's as good if not better than any valve amp he's listened to, including Quad, McIntosh etc. I built it from scratch, pretty easy to do if you have time and patience.

There's a few mistakes on the schematics, they're for guidance only.

Andy.

I guess I was wanting to go for these tubes since I had plenty sitting around and wanted to make use of them, plus they kinda look cool :D

Anyway, it was actually your thead i stumbled on earlier when I was looking at designs that use these types of tubes, I know you went away from this entirely, as it was a prototype, but your schematic gave me some ideas nonetheless.
https://www.vintage-radio.net/forum/...4&postcount=38

Diabolical Artificer 14th Apr 2021 6:18 am

Re: 807 tube amp, any ideas?
 
I refined the 120w amp over several years, put a hell of a lot of work into it and hours of research, I'd advise against using that schematic,it was about Mk5, the finished amps were about MK79 : ) 807's do have character, thing is if you want to run 6 you'll need to go through about 100 valves to get six matched.

I've bench tested numerous 807 valve amp circuits from extended class A, Class AB2, Class AB2 ultra linear which killed several 807's and several other mad ideas, in the end they're best run straight in PP with a regulated G2. The extended class A article is worth reading though, it worked quite well, but you need the right OPT. The problem running 807's in UL is that Vg2 max, it means you have to keep your HT within 400v, 400v HT is not quite enough to get the best out of a pair or more 807's. My thinking is your already chucking about 10w per valve away on the heaters,you've gone to a lot of trouble and time to build a chassis, prototype etc,why not run the valves at their optimum configuration?

I reluctantly ditched the idea of running 807's with Vg2 over 300v, it just doesn't make sense, for a start no new 807's are being made, they are a finite resource, running them at full whack is a bit cavalier.

I hope I don't sound too discouraging, and am glad some of my original work inspired you to build something. I'd advise you to download the STC 807 application report, find a set of parameters that's about right, then find a decent pair of OPT's, then build the OP stage and PSU and crack on from there. Designing an amp from scratch is good fun, educational and immensely frustrating at times,bit like bringing up kids or keeping chickens.

All the best, Andy.

Radio Wrangler 14th Apr 2021 9:05 am

Re: 807 tube amp, any ideas?
 
I once made an amp with quad 813s. One pair each side of push-pull. The difference between anode and screen voltage rating is even more severe. There was no thought at all of ultra linear. The screens got a supply regulated by some discharge tubes. and a fair bit of art went into adding stoppers to keep it stable at RF. Rather special transformer design was needed at that HT voltage - 1.8kV at nominal mains. The very high HT meant that the anode voltage mostly stayed well above the screen volts. 10v @ 5A heater per bottle and graphite-block anodes. The high HT helps with distortion. If low distortion had been needed, I'd have gone with a pair of 3-500Z triodes, or I'd have gone with grounded-grid driving directly heated filaments (and their individual heater windings) from a step down transformer from a smaller power amp.

The 807 ultra-linear versus rating issue is a point on a curve. It gets easier for smaller valves, it gets less important for bigger ones.The 807 may be the worst point on the curve.

David

nzoomed 15th Apr 2021 2:19 am

Re: 807 tube amp, any ideas?
 
Quote:

Originally Posted by Diabolical Artificer (Post 1364131)
I refined the 120w amp over several years, put a hell of a lot of work into it and hours of research, I'd advise against using that schematic,it was about Mk5, the finished amps were about MK79 : ) 807's do have character, thing is if you want to run 6 you'll need to go through about 100 valves to get six matched.

I've bench tested numerous 807 valve amp circuits from extended class A, Class AB2, Class AB2 ultra linear which killed several 807's and several other mad ideas, in the end they're best run straight in PP with a regulated G2. The extended class A article is worth reading though, it worked quite well, but you need the right OPT. The problem running 807's in UL is that Vg2 max, it means you have to keep your HT within 400v, 400v HT is not quite enough to get the best out of a pair or more 807's. My thinking is your already chucking about 10w per valve away on the heaters,you've gone to a lot of trouble and time to build a chassis, prototype etc,why not run the valves at their optimum configuration?

I reluctantly ditched the idea of running 807's with Vg2 over 300v, it just doesn't make sense, for a start no new 807's are being made, they are a finite resource, running them at full whack is a bit cavalier.

I hope I don't sound too discouraging, and am glad some of my original work inspired you to build something. I'd advise you to download the STC 807 application report, find a set of parameters that's about right, then find a decent pair of OPT's, then build the OP stage and PSU and crack on from there. Designing an amp from scratch is good fun, educational and immensely frustrating at times,bit like bringing up kids or keeping chickens.

All the best, Andy.

I only plan on running a pair on each channel, so matching wont be much of an issue. Believe it or not, you may be surprised to learn that 807s are still made in Russia and China, but dont know what their quality is like, but ive got a lifetime supply of the things anyway.

From what Radio Wrangler is saying, it sounds like running 807's in UL configuration is the same as a triode so the screen voltages dont matter., or am i missing something?

I guess I could try at a higher voltage if thats the case, nonetheless, my guitar amp has them running at 400V fixed bias and not UL configuration.

I have all the gear to wind transformers, and all the bobbins, laminations, etc, but dont know where to start on winding one yet.
I know some people who can wind a decent OT if needed.

Radio Wrangler 15th Apr 2021 4:43 am

Re: 807 tube amp, any ideas?
 
Ultra-Linear is sort of part-way to triode connection. The screen follows a percentage of the anode swing. For triode connection, that percentage is 100% for tetrode operation it's 0%

For ultra-linear operation, different people have pinned the tail on the donkey at various percentages. The tetrode end of the range (0%) tends towards more gain and more output power. The triode end of the range (100%) tends towards lower distortion at the expense of lower power and lower gain.

David

Diabolical Artificer 15th Apr 2021 5:46 am

Re: 807 tube amp, any ideas?
 
This explains UL - http://www.oestex.com/tubes/ul.html an article by Hafler & Keroes. Inna nutshell UL or distributed loading is supposed to be a way to get the most power out for the lowest THD, you loose a bit of power OP for a gain of better THD.

I've looked for 807 dynamic characteristics in UL @ HT 400v but can't find one, but am sure I've seen a graph. Maybe you could use UL curves for a KT66 or 6L6? Can't find a schematic for an 807 UL PP amp either which might be indicative. I did some experiments with this but have lost my notes, but did do a video on YT,not sure how much use it'd be. Essentially the problem you have to solve is to stop G2 getting too much current when the anode swings to 0v ish. In UL G2 is already at 100v over Vg2 max if HT is at 400v. HT has to be at 400v ish to get some decent power out, so you'll need a big grid stopper, which if memory serves caused THD to go up.

Andy.

joebog1 15th Apr 2021 7:33 am

Re: 807 tube amp, any ideas?
 
Radiotron Australia posted some pics ( which I have) of 100 or so 807,s connected as triodes after the Williamson circuit was reproduced here. They officially moved the screen dissipation UP to suit Williamson amps.
I will search for the pics, but they are in almost 70 years worth of Radiotron Valve Data books, from Franks vacuum pages, it will take some time to find.
Its here if you wish to lose yourself for 50 years or so.

https://frank.pocnet.net/other/AWV_R...ics/index.html

AWV, Amalgamated Wireless Valve Company

They are arguably one of the best multi purpose valves ever made. Audio, Radio transmitters, linear regulators, electronic loads, etc etc, and virtually indestructible ( within reason ) I have run a few radio stations using RCA transmitters where the 807 pre modulators would show dull red plates for MANY years, without failing
There are a few places in NZ that have a stash or two I know, but have never been able to find a site.

Joe

Ed_Dinning 15th Apr 2021 8:19 am

Re: 807 tube amp, any ideas?
 
Hi a good reference manual on transformer design is " Transformers for Tube Amps" by Popovitch tubeampdvds@yahoo.com.au

Pricey with postage to the UK but I believe cheaper in A, NZ

Ed

nzoomed 15th Apr 2021 11:28 am

Re: 807 tube amp, any ideas?
 
Quote:

Originally Posted by Radio Wrangler (Post 1364469)
Ultra-Linear is sort of part-way to triode connection. The screen follows a percentage of the anode swing. For triode connection, that percentage is 100% for tetrode operation it's 0%

For ultra-linear operation, different people have pinned the tail on the donkey at various percentages. The tetrode end of the range (0%) tends towards more gain and more output power. The triode end of the range (100%) tends towards lower distortion at the expense of lower power and lower gain.

David

Yes that makes sense, so basically the screen voltage is the voltage potential across the end of the winding on the transformer and the tap to the grid.

I dont know an awful amount about UL, other than its mostly seen in HiFi use and not so much in guitar amps.

I guess there is a bit of experimenting to get the sweet spot, unless the theory can calculate it accurate, quite often in this field, many break any of the "rules" to get things sounding better than how they theoretically should going by the book as it were.

Quote:

Originally Posted by Diabolical Artificer (Post 1364474)
This explains UL - http://www.oestex.com/tubes/ul.html an article by Hafler & Keroes. Inna nutshell UL or distributed loading is supposed to be a way to get the most power out for the lowest THD, you loose a bit of power OP for a gain of better THD.

I've looked for 807 dynamic characteristics in UL @ HT 400v but can't find one, but am sure I've seen a graph. Maybe you could use UL curves for a KT66 or 6L6? Can't find a schematic for an 807 UL PP amp either which might be indicative. I did some experiments with this but have lost my notes, but did do a video on YT,not sure how much use it'd be. Essentially the problem you have to solve is to stop G2 getting too much current when the anode swings to 0v ish. In UL G2 is already at 100v over Vg2 max if HT is at 400v. HT has to be at 400v ish to get some decent power out, so you'll need a big grid stopper, which if memory serves caused THD to go up.

Andy.

I would imagine 6L6 would be very similar. I know an 807 is not identical to a 6L6 and there are a few differences.

Typically UL does not use a grid stopper, so I expect it would affect the THD.

Quote:

Originally Posted by joebog1 (Post 1364486)
Radiotron Australia posted some pics ( which I have) of 100 or so 807,s connected as triodes after the Williamson circuit was reproduced here. They officially moved the screen dissipation UP to suit Williamson amps.
I will search for the pics, but they are in almost 70 years worth of Radiotron Valve Data books, from Franks vacuum pages, it will take some time to find.
Its here if you wish to lose yourself for 50 years or so.

https://frank.pocnet.net/other/AWV_R...ics/index.html

AWV, Amalgamated Wireless Valve Company

They are arguably one of the best multi purpose valves ever made. Audio, Radio transmitters, linear regulators, electronic loads, etc etc, and virtually indestructible ( within reason ) I have run a few radio stations using RCA transmitters where the 807 pre modulators would show dull red plates for MANY years, without failing
There are a few places in NZ that have a stash or two I know, but have never been able to find a site.

Joe

I know an old guy here in NZ who said the old 807 could take a beating and could glow red forever without failure, he told me they liked to be driven hard and not to worry about it when I showed him my guitar amp build.
This thing puts out close to 200W when overdriven from a quad of them and the tubes flash bright blue when there is high current flowing through them as you play heavy notes.

Yes I know a couple of people very well who have stashes of 807's, they are not going away anytime soon.

I know the one weakness with the 807 is it needs alot of gain (or current?) to drive them.
My amp uses a 12AT7 phase inverter to drive a quad of the tubes.

If i build this HiFi amp im going to be using cathode bias which will not require matched tubes, which will be at the expense of reduced wattage, but I can still get the best part of 30 watts out of a pair.

G6Tanuki 15th Apr 2021 6:48 pm

Re: 807 tube amp, any ideas?
 
In times-past I built a bunch of push-pull-807 amps: control- and screen-grids strapped together triode-style, zero-bias Class-B with 750V on the anodes.

120 Watts output!

It always seems deeply-sad when I see a couple of 807s used in a circuit that only delivers a few tens of Watts output.

nzoomed 16th Apr 2021 12:28 am

Re: 807 tube amp, any ideas?
 
Quote:

Originally Posted by G6Tanuki (Post 1364736)
In times-past I built a bunch of push-pull-807 amps: control- and screen-grids strapped together triode-style, zero-bias Class-B with 750V on the anodes.

120 Watts output!

It always seems deeply-sad when I see a couple of 807s used in a circuit that only delivers a few tens of Watts output.

WOW, Im assuming you must have had a fair few tubes in parallel to get that much power in triode mode?

Most audiophiles appear to go for lower wattage, more inefficient Class A amps that are supposed to be the best "sounding"

I think there is some truth to this, but IDK how much better a class A amp sounds over a class AB or B amp for example.

joebog1 16th Apr 2021 1:11 am

Re: 807 tube amp, any ideas?
 
Class B likes to run flat chat!! not much distortion ( relatively ) at low levels class B can sound pretty shocking. Your best bet is AB1 which will fetch about 30 or 35 watts with 400 - 450 HT volts. There are 100's of circuits on line. AB1 is also pretty easy on the valves and is pretty easy to drive.

Joe

nzoomed 16th Apr 2021 5:31 am

Re: 807 tube amp, any ideas?
 
Quote:

Originally Posted by joebog1 (Post 1364866)
Class B likes to run flat chat!! not much distortion ( relatively ) at low levels class B can sound pretty shocking. Your best bet is AB1 which will fetch about 30 or 35 watts with 400 - 450 HT volts. There are 100's of circuits on line. AB1 is also pretty easy on the valves and is pretty easy to drive.

Joe

AB1 is what im looking at using.

I was looking at modding this design to suit 807's.

Its based on the mullard 5-20 design, im assuming its AB1 looking at it, but im not 100% sure.
http://diyaudioprojects.com/Schemati...-Schematic.htm

joebog1 16th Apr 2021 5:51 am

Re: 807 tube amp, any ideas?
 
Have a look here
http://r-type.org/articles/art-101.htm

Its a very nice amp with all the details. 5B/255 is a "modern 807. No mods needed. The EF86/6BK8 is probably too much gain for todays signal sources. A single triode would probably be enough. A 6C4 or similar. A 6C4 is half a 12AU7 in a seven pin bottle. Depends on what preamp you will use, if any.
Anyway, its my way of helping you along.

Joe

Radio Wrangler 16th Apr 2021 6:12 am

Re: 807 tube amp, any ideas?
 
Joe mentioned the versatility of the 807.

The most surprising place I found them was one as each Y amplifier of the old Cossor 339 oscilloscope. Running very low power. There doesn't seem much the 807 cannot do though the heater power may make the efficiency daft in low power applications.

There's a lot of 807 still in the wild, but no-one's making any more. They are, however very long lived and very hard to kill. 5B/254 and 5B/255 are rarer, and also not being made.

David


All times are GMT. The time now is 4:20 am.

Powered by vBulletin®
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Copyright ©2002 - 2021, Paul Stenning.