Mic Input transformer plate loads.

[Quayhog]

What is the affect of different plate loads on a mic input transformer?  I plan to set these up for 250 Ω inputs.

I have some Triad HS-1 with 86.5k secondary, an HS-4  with an 117k secondary, a pair of octal Stancor with an 80k, Berlant's with 100k, Altec 4665 with an 84k, the Green 4722 with 65k.

These will be feeding a12AX7 grid.  What do I gain or lose using too large a load.

I also have some 600 Ω  line outputs with similar load  issues.

I'm trying to optimize a new build mixer circuit for some newly acquired nice mics.  The mixer is based on the Altec 1567a, and I plan on using these for recording acoustic string instruments, primarily fiddle, guitar and mandolin.

Jonathan

 

[abbey road d enfer]

What is the affect of different plate loads on a mic input transformer?

Almost none practically. The only interaction is via Miller effect. Miller effect attenuates high frequencies; it depends on the actual gain of the stage and the anode-to-grid capacitance.  Reducing the plate resistor decreases gain and increases HF response.

These will be feeding a12AX7 grid.  What do I gain or lose using too large a load.

Excessive gain reduces HF response, but in addition, increasing the plate resistor actually may decrease headroom. Your vague question needs to be refined with more info, such as the topology of the mic pre. Use of NFB may considerably alter the operation.

I also have some 600 Ω  line outputs with similar load  issues.

What is it exactly you're asking?

 

[Quayhog]

What am I asking? good question.  I'm not quite sure.    I have collected over the years through modding old PA equipment a short tonne worth of transformers.    I'd like to find a use for them.  Some of these are very nice and highly sought after.

I have two  unmatched Berlant Concertone chassis that have open bias transformer coils.  These are useless for recorder restorations.  I'd like to turn those into mic preamps.  My choices for 250/600 ohm inputs have secondaries that are all over the planet.  From 15k to 118k.  Altecs, Peerless, Triads, UTC etc.  All would require a socket change or elimination.

  It was a general question about transformer to grid loading and its effect on frequency response,  and how to compensate for losses, or chose the best for the task at hand.

 

[emrr]

Hook them up with unloaded secondary to a tube grid, and measure with correct source Z.  If it's great, you're done.  Keep in mind the era of transformer you are using, and don't be nitpicky about any modern ideas of flat response.  The ancients didn't use it, and that gear sells for a lot as-is without flat response.  Easily 100% of the transformers you mention would have been expected to be used in an unloaded circuit.  Altec sometimes loaded them, sometimes didn't. 

If you don't like what you see, keeping reflected ratio in mind, insert initially a variable resistive secondary load of suitable bridging scale, and see if you like the look best at any point along that variable load.  If you do like a particular point, measure that resistance by itself, and calculate reflected Z to the primary windings. 

If there's almost a sweet spot, but you still object to a peak, you can add a series cap to that variable load and vary the C value to see if you find a best response balance.  This is a Zobel network. 

If the load value is delivering some undesirable value, now get more complicated and start scaling R and C against each other to see if the same response can be achieved at a better point.  When I have chased this, I have never had to move it much for fine tuning. 

I'm sure there's math beyond ratio conversion, but you need to know some unknowns about the transformers, and I can do all of what I described much faster than I can even start to find the equations which might theoretically deliver an ideal answer.  As I've said before, you'll learn a lot more by hooking some sh*t up and seeing if it catches on fire.  Within reason.