• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

6U8 and 6GH8 as 7199 sub data

gadget73

gadget73

junk junkie
Subscriber
So I know this has come up a bunch, but for unrealated reasons I had knocked together a test fixture using Dynaco Mark IV schematics and figured this might be a useful place to drop some data about how these actually perform. Something I maybe should have done but did not was see what the output was without changing the screen resistor. My values were determined by adjusting the screen resistor to get max symmetrical output at the cathode of the inverter section.

so for starters, this is the Mark IV schematic.

1757728512647.png

Changes to this just for completeness
The stuff in the red box is not there since this isn't a complete amplifier.
The 270K resistors go to ground, not to a negative bias supply.
The 620 + 47 resistors in the cathode chain have been replaced with a single 680 ohm resistor.
The highlighted 1.5M screen resistor has been replaced with a 470K + 1M pot.
The 82pf + 18K shelf network is not installed.

None of those changes should affect the results in a meaningful way but I'm trying to be transparent here.

Here are the results. Not sure it means a ton, but the screen resistor value to get max output is quite different, ranging from 780K to 1.3M. Values either side of optimal produce clipping or visible distortion on one side of the waveform and overall less output.

Voltage output is peak to peak measured to ground. Grid to grid would be double this. An ST-70 or Mark IV needs about 70 volts grid to grid, the worst tube here is capable of 270 volts so there is a huge amount of headroom which makes this not completely critical. Fair guess if you stick a 1M in there, it'll work well enough with most tubes.

Just as a mention though, GE#2, the one that wants the smallest screen resistor, when set to 1.5M the best I can get out of it with no visible waveform distortion is about 10v p-p, which is well below the point that is needed for max clean output in this circuit. Beyond that it has an overly fat shape to the negative peak, so some tubes are absolutely not going to play nice with stock values here. At 1M this tube will do 100v p-p which is plenty


If you want to get it closer to optimized without actually bothering to measure per-tube, subbing the 1.5M for a resistor + pot arrangement and adjusting it to get 332 volts at pin 1 would likely get you there. 332v is just the average of all voltages in my little selection here, but its also the value with the least variation % wise.

Voltages at lugs 6 and 8 were determined by working backwards from the voltage chart in the Mark IV manual. Power supplied from a Heathkit PS-2 to lug 8, and a 47K dropping resistor to a 22uF cap for pin 6, also duplicating the Mark IV circuit.

1757729780673.png

highlighted values are the averages. Not sure they are all that significant but here you go.

WB just means White Box. Sorting is more about the box than whats in the box. Last column is any other details.

Unrelated to anything, I did find that the Japan tubes are both prone to some 60hz hum which is extremely bad if the heater supply is left completely floating. Not sure how these would behave in the amp.
 
Register to hide this ad
Very useful info! But what a shame we don't have distortion numbers for the pentodes. I say that because the 7199 was designed for audio, whereas these are TV tubes. And while the latter types are often useful in our work, my general experience in comparisons has been disappointing. I highly suspect that audio pentodes like the 7199 and particularly the EF86 are simply better in this regard.

Jack
 
A few years ago i made another comparation between 7199 replacements. The circuit was a sca35 and mesured was
dist at 1 5 and 15watts, also Va Vg2 on the pentod plate and Va on the triode.
All taken with NFB still connected ( it would have been more instructive with NFB disconnected)
 

Attachments

TriodeLuvr said:
Very useful info! But what a shame we don't have distortion numbers for the pentodes. I say that because the 7199 was designed for audio, whereas these are TV tubes. And while the latter types are often useful in our work, my general experience in comparisons has been disappointing. I highly suspect that audio pentodes like the 7199 and particularly the EF86 are simply better in this regard.

Jack
Click to expand...
I suppose I could collect that but I'd have to decide on some signal level as a standard. Also I don't have any 7199 as a reference so it wouldn't be a complete comparison. Likely it would have to be measured at the triode cathode for meter loading reasons.

Might be interesting to see if the point of peak swing is also the point of best distortion. Just visually i can say it has an effect on the output waveform, though how much of that is because of bias on the triode and how much is the pentode I don't know. I assume its more to do with the triode though
 
gadget73 said:
I suppose I could collect that but I'd have to decide on some signal level as a standard. Also I don't have any 7199 as a reference so it wouldn't be a complete comparison. Likely it would have to be measured at the triode cathode for meter loading reasons.
Click to expand...
I made a related measurement recently after breadboarding a triode-mode EF86 (actually the Russian 6J32P) driving a 6BQ7A cathode follower. Substituted a 6EA8 and 6U8A resulted in about four times as much THD at 1V RMS output. I was hoping the EF86-6BQ7 combination could be simplified to a single pentode-triode tube, but no go. I do have some good used 7199s somewhere. I'll try to revisit this (and I'll do it in pentode mode) when my current project is done.

Jack
 
We're the 7199's used originally because they were inexpensive at the time?
They weren't even in production anymore when they started using them in the ST-70 and mono blocks were they?
 
Single tube is still likely less expensive than a configuration that uses multiple even if that specific tube isn't the absolute cheapest in the family.

The 7199 was specifically designed for audio use though, likely as a refined version of the TV types.
 
Anecdotal evidence, but I can't tell a difference between an RCA 6GH8A with a socket adapter and an RCA 7199 on the SCA-35 where I did a good amount of listening tests. Obviously some differences electrically, even distortion and such that can be measured, but the end result in my experience was they sounded exactly the same. I think anyone would be hard pressed to tell them apart in a blind test.
 
albusark said:
Anecdotal evidence, but I can't tell a difference between an RCA 6GH8A with a socket adapter and an RCA 7199 on the SCA-35 where I did a good amount of listening tests. Obviously some differences electrically, even distortion and such that can be measured, but the end result in my experience was they sounded exactly the same. I think anyone would be hard pressed to tell them apart in a blind test.
Click to expand...
Which is confirmed in my previous measurments. The SCA35 has stabilized pentod idle point by feedback thus makes the circuit insensitive to tube parameters. The concertina part is by design also insensitive to tube properties as long as it's grid is
at a suitable voltage.
 
It fixes the DC part of the equation but it adds an AC complexity. The SCA35 has some positive feedback through that connection which can cause problems with some tubes. @dcgillespie did a deep dive on it in his sca35 thread and i don't recall the fine points but something about oscillation problems under certain conditions.

Could possibly partially implement it but with a bypass cap to kill the AC portion of the feedback. Maybe some evening I'll play with that. Currently digesting too much breakfast then I need to go wash and wax the car. Too nice to be sitting inside at the bench fooling with phase splitters for an amp I don't own.
 
It would be great to see the same with an actual 7199 for comparison. Sadly I don't have an extra to contribute.
 
TriodeLuvr said:
Pretty sure the 7199 was still in production into the '70s. That was well after the ST-70 was designed.

Jack
Click to expand...
There must be a huge amount of Dynaco amps out there that have used them up. Or they just didn't make near the amount of 7199's compared to other driver tube types. You would think if they were made into the 70's there would still be a healthy amount of good 7199's out there.
Other than Dynaco amps, what other brands used them, do you know?
I don't thi know I've ever run across anything else that used them.
 
Lavane said:
There must be a huge amount of Dynaco amps out there that have used them up. Or they just didn't make near the amount of 7199's compared to other driver tube types. You would think if they were made into the 70's there would still be a healthy amount of good 7199's out there.
Other than Dynaco amps, what other brands used them, do you know?
I don't thi know I've ever run across anything else that used them.
Click to expand...
Acrosound S1001 Preamplifier used the 7199. Gibson and Ampeg guitar amplifiers used them.
 
Part of it may have been that audio gear was the only use, so less were made.

Didn't some Scott gear use them? I know they ran triode pentodes in some models, but thinking some versions ran a 7199 specifically.

Fisher used them on the early TA600 and I think the TA800 as well. My 600 is a late version with the 7247.
 
Sunn guitar and bass amps were based on Dynaco designs and usually used Dynaco transformers. The lower powered models (EL34) used 7199's, the higher powered models (6550) used 6AN8's. I don't recall the driver circuit values changing between the types, just the B+ being slightly higher for the 6AN8 models.

All of my Ampeg fliptops used 7199's. Most have been converted to the 6AN8 or 6BH8/6AU8 types.
 
Dynaco used basically that driver circuit on most of their amps. the ST-35 I think is the odd one out with the 7247 twin triode. Never really studied the values on the Mark III circuit but they may have a few differences. Same circuit though.
 
You must log in or register to reply here.
Back
Top Bottom