drip fourseven HF Roll Off

[gg85]

If your scope is showing the frequency response is dropping off and you are measuring right on the transformer with no load on the transformer, your problem is the transformer.
Change it for something better.

My issue is at the input, but does this necessarily mean it's the transformer itself? I can't imagine Sowter making a transformer for the redd that wasn't actually suitable. I'll swap it out for the a-10 that I have and see if there's any improvement, although it seems more likely to be impedance rather than the transformer itself right?

I do have a question regarding the testing method I'm using. My LA-2A has utc a-10/a-24 (500 ohm in and 500 ohm out), same as the redd. When using the same method, I get a flat response (slight rise in the HF). Is this a characteristic of using a 12ax7 vs the ef86, or am I misunderstanding something here? I would've expected to see a similar response if impedances are the same between both units, and if it was an impedance issue with the redd.

Thanks again everyone

 

[gg85]

I've just noticed something I missed. On Sowters site under the 9970 specs, it states "Very high primary inductance for minimal loading on a source up to 600 Ohms."

and

FREQUENCY RESPONSE (Typical) 600 Ohm source 50 K Ohm load - 3 dB 5 Hz and 50 kHz.

But under the list of transformers they offer, it says Z Ratio - 200/10k.

Drip also has a new design, called the Ultra British, which is a pre/compressor combo, with the redd 47 as the pre, and he states the Sowter 9970 is 200 Ohms. So my input is actually 200 ohms and now this is making a lot more sense to me.

Going to do some testing now and see how I go!

 

[gg85]

Can I ask one more inexperienced question...in the redd, what is setting the input impedance at 200, and maybe more importantly, the secondaries at 10k? And how can I confirm that my circuit is giving me this?

 

[Winston OBoogie]

Can I ask one more inexperienced question...in the redd, what is setting the input impedance at 200, and maybe more importantly, the secondaries at 10k? And how can I confirm that my circuit is giving me this?

The input impedance of a REDD.47 isn't 200 ohms, that's the nominal source impedance that the input transformer and amplifier is expecting to work from.
The actual input impedance on an original REDD.47 amplifier was somewhere in the 4K5 region, with the exception of the few desk modules where the bass-cut filter was engaged (Echo send amps). In those cases the input impedance came down to around 3K5.


The input transformer is a 1:7 so a primary source impedance of 200 ohms will be reflected to the secondary as approx. 10K.

The turns and voltage ratio is 1:7, the impedance ratio is turns ratio squared so, 7 X 7 = 49.

 

[gg85]

OK, so the redd will give a "flat" response when it's seeing 200 ohms or less? Sorry, impedance is something I'm still getting my head around!

I've just tried an apogee symphony mk2 instead of the apollo. Apogee state it's 50 ohms line out, which should definitely be low enough for the redd to give a flat response, but I'm still seeing roll off. Not as steep, but down about 1dB at 20k. Is there something obvious I'm missing here and does this point back to the amp rather than the input transformer?

 

[radardoug]

I'm sorry to labor the point but..If you measure after the input transformer and you have a rolloff, then either your test hookup is rolling off, or the transformer is rolling off. Its that simple.

 

[gg85]

I apologise if I seem thick and you have to keep repeating yourself, but thanks for being patient!

Are you suggesting taking the secondaries out of circuit and seeing if I still get roll off? When I hook the input transformer straight to the output transformer I don't see any roll off.

 

[radardoug]

The best way to do this would be to work through systematically. Set up your input oscillator. Measure straight across that oscillator and make sure you have flat response. Repeat after every significant stage. Note, having an input signal that is flat is very important, and you need to verify that. Be aware of loadings, by your test gear and by the circuit itself.