Looking for a circuit how to drive a 845 in class-A PP, with B+ at 800V. My initial thought was to build a differential amplifier with a 6H30, gain stage with EL84 and drive the gate of the 845 with a source follower (yes, a MOSFET.... - similar to the Kronzilla amplifiers)......
any better ideas? thanks for your help
any better ideas? thanks for your help
The 845 is a nasty load: high input voltage at the grid due to the low voltage amplification, large internal capacitances, and it likes to pull grid current even before Vgk goes positive. About the only viable choices are a stiffly driven IST or a source follower.
Go with the source follower.
Go with the source follower.
I generally think that the requirements for driving the 845 is a little bit exaggerated.
According to the RCA data sheet: Cgp=13,5pf, Cgk=6pf, u=5,3
This makes for a input C of less than 80pf+strays. For the chinese tubes, his is probably on the low side of reality but it is not that far off.
I think the problem is to acheive the big voltage swing into a relatively low grid resistor with decent linearity.
And of course they DO draw some current before 0v at the grid so you have to live with that they clip at Vg 2-3V below zero. Or design a A2 driver.
I have driven 845pp with a 5687 diff pair with good reults. I later tried 12b4 and 6bx7, the latter being VERY good.
/Olof
According to the RCA data sheet: Cgp=13,5pf, Cgk=6pf, u=5,3
This makes for a input C of less than 80pf+strays. For the chinese tubes, his is probably on the low side of reality but it is not that far off.
I think the problem is to acheive the big voltage swing into a relatively low grid resistor with decent linearity.
And of course they DO draw some current before 0v at the grid so you have to live with that they clip at Vg 2-3V below zero. Or design a A2 driver.
I have driven 845pp with a 5687 diff pair with good reults. I later tried 12b4 and 6bx7, the latter being VERY good.
/Olof
zigzagflux said:
And the 845 would follow it quickly
You're right ... if it were a 'Lytic !
Bent side is supposed to denote the outside foil.
Yves.
Monster MUstage Driver!
Hi Yves,
On your two drawings you could put a choke coupled MuStage Driver like a EL86 or a 7591 and run the same voltage as the output stage !! You would have a monster voltage
swing, about 760v&960v respectively.
Hi Yves,
On your two drawings you could put a choke coupled MuStage Driver like a EL86 or a 7591 and run the same voltage as the output stage !! You would have a monster voltage
swing, about 760v&960v respectively.
Re: Monster MUstage Driver!
Hi Ben,
Not exactly what you suggest, but I've also tried that wich is also able to deliver huge drive:
It had my preference, but the builder was affraid by the choke load 🙁
Yves.
Ben Reginato said:
Hi Ben,
Not exactly what you suggest, but I've also tried that wich is also able to deliver huge drive:
It had my preference, but the builder was affraid by the choke load 🙁
Yves.
Don't get me wrong the 12HG7 has got some huge balls for such a small tube, but isn't the 813 an easier to drive than the 845. I could very well be wrong. I am sure something of the sort could drive a 845 though. If you want over kill you could drive it with a KT66 or KT88 triode strapped, or you could leave it a pentode, could choke-load or CCS-load for max effect!
Somewhat, the GM70 is a monster, it is on my list to build. The GM70 as far as I know, is easier to drive than the 845; it also has a higher max plate dissipation at 125watts vs. 100watts. Designed right and it can kick out close to 50watts of SET magic. One of the special things about this tube is that it can be run at very low voltages to start. You can put 400v-450v through the GM70 and you will basically have an over-grown 300B. To get the max power you need closer to 1kv, but unlike the 845 and 211 it doesn't really NEED 1kv to "come alive". They are also a lot cheaper than 845's, but the base is very unique.
Could you just use an interstage, or still not meaty enough?
Cheers
James
edit: ^^^ got to it before me with the 6V6, especially choke loaded it could handle it with ease.
Here's a preliminary 845 SET design I'm working on. 845 Prelim Design.
Here, I'm using source follower drive to take care of any grid current, input capacitance, and overdrive. This will be running at a rather stiff Vpp= 1.1KVdc since that's where 845s really like to run to be at their best.
To help take extreme voltages off the front end, this uses balanced +/- supplies. The second voltage amp is cathode coupled to avoid the largish capacitance that this stage would otherwise have, and active plate loading for the best linearity and to accommodate the big voltage swing needed at the 845 grid.
Here, I'm using source follower drive to take care of any grid current, input capacitance, and overdrive. This will be running at a rather stiff Vpp= 1.1KVdc since that's where 845s really like to run to be at their best.
To help take extreme voltages off the front end, this uses balanced +/- supplies. The second voltage amp is cathode coupled to avoid the largish capacitance that this stage would otherwise have, and active plate loading for the best linearity and to accommodate the big voltage swing needed at the 845 grid.
Hi Miles,
Obviusly, you're not a "minimal design" afficionado 🙂
All that stuff to drive a MOSFET follower !
Caricatural provocation, I admit !
Cheers, Yves.
Obviusly, you're not a "minimal design" afficionado 🙂
All that stuff to drive a MOSFET follower !
Caricatural provocation, I admit !
Cheers, Yves.
Yvesm said:
Too many compromises in those "minimal" schemes: either capacitor coupling to a fixed bias final (blocking problem on over drive) and/or magnetics in the signal path (precludes NFB).
Like Einstein said: "Keep it simple, but not too simple" 😉
Wow, this seemingly simple question has evolved quite a bit.... thanks for all the good ideas?
Miles, what MOSFET are you using in your schematic?
Miles, what MOSFET are you using in your schematic?
^^^^
I'm considering a Fairchild FQPF3N40 for now. It has a nice, low Crt that smooths out nicely above Vds= 30V. This is tenative, and might be changed when the design is finalized.
I'm considering a Fairchild FQPF3N40 for now. It has a nice, low Crt that smooths out nicely above Vds= 30V. This is tenative, and might be changed when the design is finalized.
Does the 845 require such high plate voltage
if one is willing to drive the grid into a2? That
grid can handle a few watts I am certain of it.
----------------------------------------------------------
I seen a NOS 845 at an electronics fleamarket
a few weeks ago, if I had realised then it was
a big audio triode, I mighta jumped on it...
Apparently needed special permission to buy
or sell one back whenever it was made...
Stamped with many warnings. Don't remeber
the specifics of the markings now. Think it
was an RCA.
I do remember the graphite plate, some springs
above the top spacer, and biggest grid structure
I've ever seen.... And furthest from the Cathode.
And dense too, no seeing through the mesh.
Didn't know at the time that it was the only grid.
I assumed that far out from the center, the "real"
grid musta been further inside. I was drawing the
wrong conclusion that I was looking at a Pentode...
if one is willing to drive the grid into a2? That
grid can handle a few watts I am certain of it.
----------------------------------------------------------
I seen a NOS 845 at an electronics fleamarket
a few weeks ago, if I had realised then it was
a big audio triode, I mighta jumped on it...
Apparently needed special permission to buy
or sell one back whenever it was made...
Stamped with many warnings. Don't remeber
the specifics of the markings now. Think it
was an RCA.
I do remember the graphite plate, some springs
above the top spacer, and biggest grid structure
I've ever seen.... And furthest from the Cathode.
And dense too, no seeing through the mesh.
Didn't know at the time that it was the only grid.
I assumed that far out from the center, the "real"
grid musta been further inside. I was drawing the
wrong conclusion that I was looking at a Pentode...
kenpeter said:
You can operate Class A2. However, getting a SE OPT that can handle some 40W and do it while sounding good is problematic. Ain't easy to do so for just 20W. Unlike PP xfmrs, the SE xfmr has to handle both DC and AC core magnetization, and maintain a decent Lp. 845s seem to sound best when run at high Vpp. That really seems to get them into the most linear part of the plate characteristic.
Sould have. Oh well, live and learn.
I used to have a PP813 triode that used the 12HN7 in pentode similar to the Millett design above. Best sounding amp I've ever heard, with a caveat; I didn't need a lot of power as I ran very high eff speakers so I never pushed them hard, or near A2.hesener said:
For your example, I my first thought are;
- run a higher B+ (if you can get the power without A2, why add complication?)
- use a beefy IT like a Lundahl 1660 or the one specially made for the 845 - forget model #
- first stage would be CCS loaded 6SN7 or 6H30
- driver would be EL38, trioded, perhaps EL35 or 6550 trioded.
Cheers. It was a great amp, but it was huge, heavy and generated a lot of heat. Once I began multi-amping, it became impractical.JoshK said:
Not mine, but lots of similarities.....
http://www.izzy-wizzy.com/audio/powamp813.html
- Status
- Not open for further replies.