Single Ended 6550 (and other!) amplifiers

[dividebytube]

just a little update on my current audio situation -

I recently made some single ended EL156 amplifiers that I loved - but with my current speakers I ended up preferring some Threshold gear that I got a great deal on. So I sold my DIY amplifiers to a friend - however I modified them to run 6550s instead. He has a big stash of NOS tubes that he wanted to check out - now he can 6550s/EL34s/6L6GCs/etc etc. I even made a pair of octal adapters so he can burn up 6AR6s.

This 'design' is a bit of borrow-and-blend - the Ultralinear output stage is something I always wanted to try on an amplifier and the pentode front end is remarkably quick and clean. Don't let the ol' triode only bias take effect - this amp sounds remarkably un-tubey with a sound that bridges between the best of solid-state and tube. The big Hammond output transformer has some surprisingly deep bass response and the feedback loop certainly helps controlling some trickier drivers.

Schematics:

the circled components are the changes I did from my original EL156 schematic.

Anyways - with a pair of original Tungsol 6550s they sound very very good - almost makes me wish I never sold 'em!

 

[ejfud]

I like!

 

[GordonW]

Well... the 6550/KT88 has some of the lowest inherent distortion of any "normal" power tube out there... so, if there's one that should work well as a single-ended tube for good linearity, that would be it.

Unlike some tubes (the 6L6 comes to mind) that are optimized to work in push-pull (the 6L6 was designed deliberately in a way that increased even-order distortion, because it allowed the designer to drastically decrease odd-order distortion... then, the even-order products pretty much canceled out in PP operation- a pretty brilliant use of a trade-off, IMHO), the 6550/KT88 is pretty much straight-ahead... no need to "compensate" for anything... in fact, here is a link to a ham-radio engineer who tested a whole lot of different tube types, for intermodulation distortion:

http://www.qsl.net/n6ev/SwpTube.txt

The interesting thing is, that the 6550 had lower distortion (signified by the distortion products being a greater number of dB down than others) than any other pentode or beam power tube... and ONLY the really tight-tolerance, huge, transmitting triodes (which HAD to be clean, otherwise they "walk on" adjacent channels!) significantly bettered it.

The more I research it, the better the 6550 looks, IMHO...

Regards,
Gordon.

 

[dividebytube]

The (original) Tungsol 6550 is certainly one of the better power tubes I've heard - sweet, refined but still powerful. Those engineers of yore certainly knew what they were doing back then!

Oddly enough my pal also has a pair of original GEC KT77s and preferred the Tungsols (at least in this circuit).

 

[Kegger]

Very nice!
Good pentodes running Single Ended UL can be remarkable amplifiers!

I've ran 6LU8, 6LR8, 6MF8, 6JZ8 and EL500/6GB5 all in SE UL, great results.
(along with kt88, el34, 6l6gc, kt66, kt77, 6550)

And Tungsol sure did make some good tubes, I just switched from a GE 6336A
in a single ended amp to a Tungsol with quite a full upgrade in sound for sure..

 

[vincei]

Nice! Makes me want to go and build one...thanks for sharing.

 

[fsjonsey]

How much power can a 6550/KT88 produce in SE UL?

 

[WopOnTour]

How much power can a 6550/KT88 produce in SE UL?

Running my Citation II (6550/KT88) in triode yields ~30-35wpc (down from 60wpc in pentode)
Can you expect something similar in SE UL, so ~15-17 wpc for a single tube?
It would be an interesting experiment, as I have a spare pair of Citation II OPTs I've been thinking of utilizing ...
WopOnTour

 

[Kegger]

In pentode a KT88 running single ended can get yu roughly say 16watts,
but in UL I'd say more like 12watts or so running wide open, still not bad.

 

[soundmotor]

I see no schematics?

 

[bharper]

would bypassing the zener in the ul tap path not help some on reducing the zener switching noise from modulating the screen grid? maybe something like a .47 600v solen, perhaps even paralled with a .01 600v solen.

just a thought, that zener I would expect would be a noisy bugger.

but then again, I guess being single ended that the zener may be staying reverse biased, and not switching at all.

thoughts?

 

[AmadeusMozart]

Apologies to resurrect an old thread but....

Apologies to resurrect an old thread but I am just wondering if a CCS in the output tube would be a good solution rather than using a variable wirewound 5 watt resistor. Has anyone an opinion on this or even better has anyone ever tried using a CCS in the cathode of an output tube?

Many thanks in advance
AM

 

[Les Lammers]

I have had many SE triode amps...and they can be wonderful...but a SE Pentode has more 'grunt' and the tubes are a lot less $$.

 

[Jim McShane]

would bypassing the zener in the ul tap path not help some on reducing the zener switching noise from modulating the screen grid? maybe something like a .47 600v solen, perhaps even paralled with a .01 600v solen.

just a thought, that zener I would expect would be a noisy bugger.

but then again, I guess being single ended that the zener may be staying reverse biased, and not switching at all.

thoughts?

Correct, the zener is always on (assuming the amp is operating normally!). The zener assures a 50 volt drop - the plate will always be at 50 volts (less the small drop in the OPT winding between the UL tap and the plate) DC less than the plate. It really has nothing to do with it being single ended, the same technique can be applied in P-P amps. Even in class AB if the signal is cut off to one tube the DC voltage on the OPT primary is well above the zener voltage.

I don't know that a cap would be of any value at all. You could use a string of (8) 6.2 volt zeners - they are reputed to be quietest out there. I can't vouch for that, but I've seen the claim made.

In an old article in Sound Practices written by Bill Kleronomos (who's a pretty smart guy and radio station engineer) he found a significant reduction in distortion using the zener instead of a screen resistor.

 

[jaymanaa]

Nice simple design. I especially like the use of the pentode up front. I have a EL34 SET (strapped) amp that I like very much. I thinks it has as much detail as anything I've heard.:yes:

 

[gearhound]

An EF86 is one of my favorite tubes.
If you've ever heard one, in a guitar application, you'd know why?


Steve

 

[AmadeusMozart]

Correct, the zener is always on (assuming the amp is operating normally!). The zener assures a 50 volt drop - the plate will always be at 50 volts (less the small drop in the OPT winding between the UL tap and the plate) DC less than the plate. It really has nothing to do with it being single ended, the same technique can be applied in P-P amps. Even in class AB if the signal is cut off to one tube the DC voltage on the OPT primary is well above the zener voltage.

I don't know that a cap would be of any value at all. You could use a string of (8) 6.2 volt zeners - they are reputed to be quietest out there. I can't vouch for that, but I've seen the claim made.

In an old article in Sound Practices written by Bill Kleronomos (who's a pretty smart guy and radio station engineer) he found a significant reduction in distortion using the zener instead of a screen resistor.

Hi Jim, first thanks for the wonderful Tungsol 6L6GC-STR, 12AX7 and 12AU7 I got from you.

Instead of using a zener how about using one of those voltage regulator tubes like the OA2 or OB3, what you think might happen there? Sticking with semiconductors I wonder how a mosfet with zeners at the gate and a cap would work out....

Have yet to see an answer to my original question that caused the resurrection of this thread...

Marinus

 

[Jim McShane]

Hi Jim, first thanks for the wonderful Tungsol 6L6GC-STR, 12AX7 and 12AU7 I got from you.

Instead of using a zener how about using one of those voltage regulator tubes like the OA2 or OB3, what you think might happen there? Sticking with semiconductors I wonder how a mosfet with zeners at the gate and a cap would work out....

Have yet to see an answer to my original question that caused the resurrection of this thread...

Marinus

Thanks for the kind words, I'm really glad you are enjoying them!

On the VR tubes - I've never really thought it through. It might be possible, but what tube would you use? I'm not sure you want the screen voltage to drop too low. The 0G3 is the lowest I know of and we're still talking 85 volts drop or so. Plus the 0G3 has very low current ratings. Maybe an 0A3 would work - 75 volts nominal and higher current ratings.

I have no personal experience with a CCS in the output stage - sorry!

 

[AmadeusMozart]

Thanks for the kind words, I'm really glad you are enjoying them!

On the VR tubes - I've never really thought it through. It might be possible, but what tube would you use? I'm not sure you want the screen voltage to drop too low. The 0G3 is the lowest I know of and we're still talking 85 volts drop or so. Plus the 0G3 has very low current ratings. Maybe an 0A3 would work - 75 volts nominal and higher current ratings.

I have no personal experience with a CCS in the output stage - sorry!

I've seen (but do not have the time right now to chase it up, have to go out) some set up with 6L6GC that had 375V on the anode and 200V on the screens and the result was pretty low distortion.

Later I'll get some more info.

 

[AmadeusMozart]

Maybe an 0A3 would work - 75 volts nominal and higher current ratings.

I now remember where I aw this - it was in one of the amateur radio books that I have but since we are in the middle of moving (we've just sold and are building a new place) I cannot get to them. From memory it was in a transmitter where the screen voltage needed dropping a lot so it was not from the UL tap. I suppose one has to be very carefull about the max and minimum current and the start voltage required since the voltage regulator tubes need a higher than the regulating voltage to get started.

Then there is the other aspect that a zener in reverse direction works like a diode and some people are putting diode in that screen - UL tap.

AM