srpp & distortion

[vitalstates]

I'm not sure whether this should be posted under srpp, distortion, or oscilloscopes.

The only thing really linking them together here is that I know squat about any of them....

The prologue:

I've built a few srpp amps in the recent past and two have been crackers and the last one not so very good.I've recently read that srpp is only really good into its own specific load and that variance from this load will produce mediocre results. As the documentation on the interweb is scarce I thought I'd do some experimenting and search for some truth..

Act 1:

build the amp(6sl7 srpp, 2k2, 470k load, bias = 1.6v) - established/published srpp circuit.
measuring just the srpp, monitor o/p at load point.
srpp clipping - arrived at 0.535vrms i/p when bottom of wave at o/p started to go flat
this showed 32.4vrms(91vpp)at output. Seemed ideal as it would clip at same time as 2a3 biased at 45v. This point showed 5.27vrms into 8.25R(3.36watts)

so far so good...1st question, why does srpp clip at 1.5vpp input when the input valve is biased at 1.6v, would have thought there would be some headroom??

Havn't managed to find a definitive(easy to read) description for finding a reasonable op point in srpp. The only load line info I can find seems to imply a vertical load line????

moving on, undeterred

Act 2

measuring output vs input - ch1 connected to i/p ch2 connected to output.
adjusted channel 2 to get similar trace to channel 1.
inverted channel 2
added channel 1 to channel 2
fine tuned v/div on channel 2 to get approx straight line
increased i/p voltage until leaving straight line - this voltage = clip voltage from act1.

2nd question, this measurement method is pure guesswork, is it a good way of detecting onset of distortion??

0.53vrms is too sensitive for i/p in my set up, so I'm going to go back to 6sn7 for this amp. but will measure 6sl7 at different settings and load to see if there is a pattern.

3rd question, Is it possible to do more scientific distortion measurement with just a scope and generator

[Merlinb]

Originally Posted by vitalstates

so far so good...1st question, why does srpp clip at 1.5vpp input when the input valve is biased at 1.6v, would have thought there would be some headroom??

It may be because it is not optimised, so with the given load you are not getting sufficient drive to the upper tube, so it runs out of headroom too soon.

Is it possible to do more scientific distortion measurement with just a scope and generator

I don't think so, at least not at the sub-1% levels you should be getting. You can use a computer though, and some free software such as RightMark Audio. I measured 0.01% THD for an optimised ECC88 SRPP driving a 15ohm load.

I'll have a think about the other questions.

[Miles Prower]

Originally Posted by vitalstates

I've built a few srpp amps in the recent past and two have been crackers and the last one not so very good.I've recently read that srpp is only really good into its own specific load and that variance from this load will produce mediocre results. As the documentation on the interweb is scarce I thought I'd do some experimenting and search for some truth..

This is one of those weird, trendy circuits. The best way to look at the SRPP is that it's an active pull-up/active pull-down circuit that isn't really all that different from the back end of a TTL gate. When used as a linear amp, the A Number One deficiency is that there is no such thing as a "P-Channel" VT.

That means that the SRPP is balanced for just three load conditions: dead short (not very useful) an open line (forces balance since the only path for the current is through the two tubes) and the one impedance for which it was designed and optimized. This is not a problem given the original intent to the design: a line driver that could easily be matched to the characteristic impedance of the driven line.

You can frequently get away with using an SRPP in those cases where the load approaches open line conditions. If you depart from that, expect increasing distortion with rising imbalance.

It's one of those trendy designs I would avoid. If you need voltage gain with a Lo-Z output, just use a conventional voltage amp, and connect it to a cathode follower.

As for documentation, Tube Cad Journal, Brosky discusses the SRPP at some considerable length, and there are more than one article over there about it.

[nigelwright7557]

I have had good results with a hybrid tube/MOSFET amp I designed that used an SRPP tube front end.
It gives a very good clear sound.

I am not loading it heavily so perhaps that is why it performs so well.

[vitalstates]

thanks chaps

Hi Merlin, any more regarding the questions would be appreciated, in the meantime the rightmark audio looks interesting, I'll have a play, thanks

Ed

[Merlinb]

Originally Posted by vitalstates

2nd question, this measurement method is pure guesswork, is it a good way of detecting onset of distortion??

Sounds good to me. I am more familiar with the X-Y method, but this sounds like basically the same thing.

Havn't managed to find a definitive(easy to read) description for finding a reasonable op point in srpp. The only load line info I can find seems to imply a vertical load line????

Well you know the quiescent anode-cathode voltage is half the supply voltage. The load seen by each tube is twice the load impedance, plus one Rk. If Rk is very small with respect to Rk then you could just draw a load line equal to twice Rload, starting at the supply voltage. Then read off the anode current when the anode voltage is half the supply voltage. Then you can find Rk.

However, if you're designing an optimised SRPP then you don't need load lines since the formulae do it for you, without any iterative annoyances.