6AS7GA as push pull output?

[artosalo]

Produced a high pitched wine which would come and go at random. Others have reported similar experiences in headphone amps using these valves.

I have not heard about such problem from the guys who built those PP-amps.
But we used Soviet made 6N13S tubes. Also no problem with 6080.

[Shoog]

Originally Posted by artosalo

I have not heard about such problem from the guys who built those PP-amps.
But we used Soviet made 6N13S tubes. Also no problem with 6080.

Has any of them run them on very sensitive speakers or as headphone amps ?
Unless they have their experience isn't comparable.
I have used just about every type of 6080 variant and they all show similar behaviour.

If you do extensive searching on 6AS7 based headphone amps you will come across people reporting the same thing.

Even though I loved my 6080 PP amps, I still think that there are better candidates for the job. I personally think that the PL84/EL86 is a better bet and not excessively expensive. Dealing with the massive drive requirements of a 6080 amp makes for a more complex build in the end. The MU of 2 will really equate to a gain of 1 or less in a real world circuit.

Shoog

[artosalo]

Has any of them run them on very sensitive speakers or as headphone amps ?

I don't know about their speakers. Mine are 91 dB Radiotehnika Giant FS-100.
Giant FS-100N | VEF Radiotehnika RRR

But I have measured the amplifier's S/N, which is > 80 dB (unweigted).

Dealing with the massive drive requirements of a 6080 amp makes for a more complex build in the end. The MU of 2 will really equate to a gain of 1 or less in a real world circuit.

True. The driver stage needs to be carefully designed. The MU in actual circuit is (in my case) 1.44 as can be seen from the schematic.

EL86 as a PP/UL is very good choice, better that typical EL84. With 300 V suppy voltage, more than 20 W is available.
With EL84 and +Ub = 300 V, some 12 W is max.

[soulmerchant]

Hi mr2racer. you probably know this OTL circuit:

6AS7GA OTL

Did you ever think of trying 6AS7 as a driver?

[rsumperl]

Good morning everyone,
Here's a really old data sheet on the 6AS7 that has audio info on it. Hope this helps.

Ray

[Bigun]

You can also use this tube in a SE configuration: my CELLINI triode amp

I still use it occasionally.

[mr2racer]

Thanks everyone for the help. Especially the data sheet with audio information. I was confusing internal plate resistance and plate to plate load resistance. That data sheet was a bit of a disappointment though as it warns to never use the tubes in parallel. Another disappointment was the expected life of the tube of only 400 hours.

As for the noise/whine Shoog mentioned I've found posts dealing with it. Most are for OTL headphone amps. They seem to find it is a product of individual tubes rather than the design of the tube itself. Also, rolling 6AS7, 6080 and 5998 tubes to find examples that are quiet. So it doesn't seem to be limited to 6AS7. Winged C built 6AS7G which would be relatively late model tubes. And yes they are pretty!

On the up side the tubes are inexpensive when compared with other power triodes like 2A3 and 300B. They are NOT DHT tubes. Built a couple of those amps and don't want to again. Also, both power transformers and output transformers would be fairly straightforward. As for the drive requirements from the interelectrode capacitance's, (I'm on thin ice here.) they don't seem more difficult to drive than 2A3?

The audio datasheet mentions the low mu/voltage swing problem and recommends using an interstage transformer to swing the grids. So I would probably start with a known quantity differential amp like Pete Millet's uniamp. Or possibly design from scratch using EDCOR line matchers as interstages and 6N6P as drivers, I have a few.

I've wanted to build a ten watt pure triode amp for a while. Which would be relatively high power I might add. One without feedback. And the differential push pull layout seems the most attractive option. Plus the amp could be done with only four tubes. I'm going to get a quote on the transformers and see where it takes me?

I really didn't need another project but this one feels good!

[audiowize]

Originally Posted by mr2racer

That data sheet was a bit of a disappointment though as it warns to never use the tubes in parallel.

No big deal to use four tube per channel for parallel push-pull, just use individual cathode resistors and bypass caps.

Originally Posted by mr2racer

Another disappointment was the expected life of the tube of only 400 hours.

I would take that with a grain of salt. Expect 5,000+ hours.

Originally Posted by mr2racer

On the up side the tubes are inexpensive when compared with other power triodes like 2A3 and 300B. They are NOT DHT tubes. Built a couple of those amps and don't want to again. Also, both power transformers and output transformers would be fairly straightforward.

I'm not sure how much better/worse this would be than using something like a pair of EL34's. Consider that your push-pull output transformer needs to handle a lot of current but only be sized for modest output.

Originally Posted by mr2racer

As for the drive requirements from the interelectrode capacitance's, (I'm on thin ice here.) they don't seem more difficult to drive than 2A3?

A 2A3 sits at 50-60V of bias usually depending on the design, while you may have 100V of bias on the 6AS7. The 6AS7 sure will need more drive voltage, and if you use some global feedback around the circuit, then you'll need even more!

Originally Posted by mr2racer

The audio datasheet mentions the low mu/voltage swing problem and recommends using an interstage transformer to swing the grids.

I think there may be an assumption in that datasheet that one would want the simplest possible amp, and they may be thinking that you could run something like a 6J5 off the same B+ supply as the output stage, then just step that first stage up a ton to get the required drive voltage for the output stage.

Originally Posted by mr2racer

I've wanted to build a ten watt pure triode amp for a while. Which would be relatively high power I might add. One without feedback. And the differential push pull layout seems the most attractive option. Plus the amp could be done with only four tubes. I'm going to get a quote on the transformers and see where it takes me?

You may be a little disappointed when you put your efforts up against a half way decent 300B amp (which can also easily be made to make 10W).

[artosalo]

The audio datasheet mentions the low mu/voltage swing problem and recommends using an interstage transformer to swing the grids.

The datasheet and recommendation is from 1947 when it was common to use interstage transformers.
As you can see from my schematic, there is 370 V supply voltage available for the driver and required drive signal for the output tubes is 245 Vpp, so no problem at all.
The driver stage has also a huge effect to the overall THD. Now it is biased to perform most of the distortion cancelling. The output stage alone has some 3.5...4 % THD at 10 W. With properly biased 6SN7 driver the overall THD is below 1 %.

With an interstage transformer as a driver you won't get such improvement to THD.

Attached the test results of my amplifier.

[Kay Pirinha]

What is even more dissapointing in the datasheet is the large power loss in the cathode resistors. But this is common to all self-biased power triode applications. Thus, I'd prefer smaller cathode resistors, primarily for balancing purposes, and fixed bias, assisted by a servo circuitry.


Best regards!