![]() ![]() |
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.206.64.37
Anyone tried to use a 6SN7 in place of the 6922 in an Audio Research SP10 or Audible Illusions M3 ?
Tube Depot sells 6sn7 to 6922 converters for $12 ea:
https://www.tubedepot.com/products/6sn7-to-6922-adapter
Follow Ups:
nt
If you're trying to scratch a tube-rolling itch, may I suggest the JJ E88CC/6922?
They're a great sounding 6922. I've mothballed all my Telefunken 6922's in favor of the JJ's.
IMO, a much better sounding tube than any Russian I ever heard.
The SP10 (and AI M3 series) drive tubes very hard. Most NOS and new tubes may sound wonderful, but do not last more than 100-200hrs in use. It becomes an expensive exercise trying to balance longevity vs ultimate sound quality.
Search goes on.
"Most NOS and new tubes may sound wonderful, but do not last more than 100-200hrs in use."
This unit could undoubtedly be easily modified to work properly with those tubes (or a different type). No way would I put up with that if it were mine.
I would sell to a collector that will never use it and get something better sounding.Looking at the schematic, it appears that the only tubes that are run hard are the four 6DJ8 paralleled cathode followers. The other 6DJ8s are run conservatively. As these are cathode followers without any gain, you can run super cheap 6DJ8s in positions V7, V8 V11 & V12 in the short lives sockets. Also the power supply regulator tubes 6L6s and the on 12AT7 can be cheap.
Good 6DJ8 tubes in V1, V2, V3, V4, V5, V6, V9, & V10.
(Also, is this using Q14 as a voltage clamp, like a diode, in that position? It would seem as though a diode would be simpler (and superior))
And running the signal through a photo cell at the end is fun, maybe it wants you to listen with the lights off.
![]()
Edits: 03/05/25 03/05/25 03/05/25
"Looking at the schematic, it appears that the only tubes that are run hard are the four 6DJ8 paralleled cathode followers."
The voltages and current shown on the schematic don't indicate unusually harsh conditions for those. The follower is dissipating slightly less than 1.8W. The data sheet shows a rating of 3.6W total.
Audio Research gear is always over-complicated with multiple areas of potential failure. Probably worth measuring voltages around the tubes that are failing before doing anything else. If everything seems OK, I'd be tempted to use a different tube type in the follower positions. The 6GU7 has similar curves, although filament current is higher. It's rated for 6W total.
The data sheet I was looking at said "both plates (both units operating) max 2 watts". So while not exceeding max, ts is the only place where you could say relatively "running hard".
"The data sheet I was looking at said "both plates (both units operating) max 2 watts"In that case, you're absolutely correct, it's operating on the edge. The Philips data sheet doesn't show that limitation, and maybe the AR designers didn't see it either. In any event, the intention was obviously to squeeze as much drive capability out of the 6DJ8 as possible. At this level though, they should have used a different tube.
Here's the simplified schematic with all the extraneous circuitry removed:
![]()
I modeled this in SPICE and made a few changes to the cathode followers. Simply increase the value of R13 to 180K, and increase the value of R14 to 39K. Both new parts should have a 1W rating. This will reduce the dissipation in the followers to about 70% of the original value and extend their lifespan considerably. The reduction in current of course means less drive capability, but it will still drive 18V RMS cleanly into a 10K load.This change will also reduce standing DCV on the cathodes to about +100V. The heaters are elevated to about +122V, so this is well within the H-K ratings for the tubes.
As a final note, this has virtually no impact on line stage gain. The difference is about 0.1dB.
Edits: 03/06/25 03/06/25
That makes sense.
Isn't a cathode follower the only "acceptable" use of a 12AU7 (5.5 watts both plates combined max). Would seem to be a better choice.
Yes, a 12AU7 would probably work well. I would need to toss it into the model to determine appropriate values. Other tubes might work as well, such as the 6BQ7A. The idea was to do the minimum necessary to make the original tube type work. The high Gm of the 6DJ8 is beneficial in this circuit.
One the more rugged 6922 variants is "supposedly" the Russian Reflector 6H23P-EB's. $50-$60 a pop though last I looked.( Viva and The Tube Store)
That is beyond brutal.
KP
Use 6CG7 instead. Already a nine-pin and electricity equivalent to the 6SN7.
But as others have said, just because you can doesn't mean you should. I HIGHLY suspect you will not like the results.
Couldn't agree more. An adapter doesn't magically make a 6SN7 100% compatible in a circuit designed for a 6922. Just buy the best 6922 or equivalent you can afford.
Those tubes have wildly different heater ratings.
The 6922 draws 0.3A per tube, the 6SN7 draws 0.8A per tube. You need to start thinking "What if I burn out the heater winding on this power transformer" before you go all "lets hear the sonic differences and see if we want to keep 6SN7s in there".
Really, the power transformer can probably take it, but any temperature rise is just eating away at its life. I would not recommend it.
Those two tubes are very different. If the circuit values are optimized for the 6922 I would doubt they would be for the 6sn7.
What are you looking for with this change?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I agree. I prefer the 6SN7 to the 6922 but only if the circuit is designed for it and the tubes are way too different for that
Any adapter like that should come with a disclaimer about performance. (Maybe it does, I haven't checked the site.)
![]()
pop it in there and off you go.
I call BS
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
KP
.
Have Fun and Enjoy the Music
"Still Working the Problem"
nada aqui
.
Have Fun and Enjoy the Music
"Still Working the Problem"
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: