Royer Mod Transformers 12:1 vs lower ratios

Is there any particular reason for needing a 12:1 transformer in the royer circuit? If it's based off a U47(I say based off, coz I know it's a different circuit, different tube etc) wouldn't a lower ratio be better suited?
Most of the tube mic circuits I've read about round here use a lower ratio.

I've put a lundahl 1538(5:1) in mine and it sounds great, which at the end of the day is the most important thing.

I guess I'm trying to understand mics a bit better, and I know I've got a lot of reading and experimenting to do, but if there's a specific reason for the 12:1, that'd be good to know (or be pointed in the right direction)

I think Dave Royer chose that transformer because it readily available, does the job well, is relatively low cost and high quality.

I've built Royer-ised mics with both the Lundahl and Sowter 7:1 trannies, and both work very well. But it depends what you're trying to drive.

Obviously with a lower ratio you'll get more level but won't be able to drive very low impedence inputs successfully.

If you don't need to run long cables you can even skip the transformer and run straight to a high-Z tube input!

Regarding the turns ratio, apart from getting a lower output impedance, do you think he might have chosen a higher ratio so as to get more turns on the reversed secondary and thus hopefully more inductance?

I've been looking into a little about what models might be most suitable.

Roddy

work out the ratio for the 50 ohm output not the 200 ohm

How do we figure out the plate resistance for triode connection? All I can see on the data sheet is a quote of typically between 175K and 260K, which gives you something like ~72K out when using a 100K plate resistor. That sounds too high to me.

[quote author="rodabod"]How do we figure out the plate resistance for triode connection? All I can see on the data sheet is a quote of typically between 175K and 260K, which gives you something like ~72K out when using a 100K plate resistor. That sounds too high to me.[/quote]

I'm a newbie, so take this with a grain of salt, but:

I believe you can approximate it by determining the slope of the grid bias curve if the tube is cathode biased. You should be able to do this by measuring your B+ and your plate and cathode voltages - then draw your load line, find your grid bias curve where it intersects the load line, then approximate the slope of that curve to say about +/- 20V of the plate voltage. I think :?:

To plot out your load line you'd have to use Ohm's law to determine the current flowing through the cathode resistor. This will give you one point of your load line - the plate voltage (X axis) and the current (Y axis). Your other point is the B+ voltage at 0 mA current.

Of course, the grid bias curve is not completely linear, so this is really an approximation, but I THINK it should put you in the ballpark. This also only works if you assume that the quiescent grid voltage is 0V. If there is any kind of grid leak bias then I don't think this would work.

Mason