input transformer straight to passive attenuator, manley?

as in: A log pot on the secondary of a transformer. the pot is chosen to present the propper termination load on the TF.

Ive done this in line level applications. It works especially well with a suitable 1:1 TF into 10k log or a 4:1 TF into a 1k log pot. My reamp box is nothing more than this.

but in a mic amp with a high ratio transformer there are some issues. the noise figure is worsened and the impedance goes way up as you turn the pot. I am trying to load 1:10 iron with 250k log pot in parallel with 1meg for 200k total. full up, the amp sees 20k source Z (200 mic reflected thru the TF). at 6 dB of attenuation it is roughly 145k. assuming the amplifier doesn't have distortion issues due to this impedance it still gets noisier and starts to pick up all sorts of RF trash.

I remember reading (in the manual) that the manley DVC works this way, but I asume it is a low ratio TF.

Does anyone know how the manley mono 60dB tube pre input attenuator is set up? Ive recorded alot with one but never opened it up. it seems like the input aten is pre-gainstage, as you can feed it super hot levels and never clip as long as the input knob is down.

mike p

Tons of vintage line amp units use this method also, but not common in vintage preamps meant for mics. I'd look at your shielding, grounding, and wire lead lengths for possible problems first. Should work fine.

What else is going on in your preamp circuit? Are you working with a new breadboarded design, or modifying an existing circuit? I'm not sure why you need the additional paralleled 1M resistor; is a 200K secondary termination crucial? This might be a potential problem spot.

[quote author="emrr"]
What else is going on in your preamp circuit? Are you working with a new breadboarded design, or modifying an existing circuit?[/quote]

Thanks for the reply. I am working on a new design. I have a "better" way of doing it already but I am drawn to the simplicity of a fully passive approach like this. assuming a high-headroom transformer, it is positively unclipable. I'd be a little surprised if nobody has EVER used this technique in a mic input. But the only pre I've ever used that behaves like it *might* be set up this way is the manley. Im going to be at a studio today that has some manley tube preamps, I'll take a look at how they do the input attenuation after the session.

mike p

Mike,

I have a Manley Dual Mono that I can pull from the rack tonight if you don't get to look at the others,

cheers,
Ruairi

I believe you are right about the Manley. In the old days, no one ever dreamed you could need variable attenuation at this point, so it only happens in line amps that would normally receive a fairly stout signal.

my radio designers handbook from 1954 shows designs that do this but says it is not ideal because frequency response and noise is not constant...the examples have the wiper feeding a triode grid...but "rock on" it also says in the glossary.

This idea?

http://www.jensentransformers.com/as/as054.pdf

I wondered about that but I guess conductive plastic pots are much quieter than pots circa 1954....

[quote author="ruairioflaherty"]Mike,

I have a Manley Dual Mono that I can pull from the rack tonight if you don't get to look at the others,
[/quote]

oooh, could you please please? I didn't have time after all.

I guess that jensen cascode circuit verifies the concept. but I can tell you from my experiments, conductive plastic or otherwise, the pot and traces/wires to the amplifier need to be very well shielded. and if you are using multiple JFETs in parallel for low noise (like I was trying) the capacitance of the gates could become significant relative to the additional source Z at mid position. you loose some high end.

mike p

Shortest possible wire runs and lowest possible capacitance in said wire; very important. And shielding. This is why I refuse to modify lots of vintage gear in this way; it goes to crap very easily, and you're better off starting with a new layout.