Let's try this again - Al's Mic Pre...

OK, here's a preliminary schematic. A few things:

1. The gain triodes are a 6N1P tube. The WCF will most likely be a 12BH7A.
2. I want to bias the WCF with the 185V coming from the previous stage and save us a cap in the audio path. Tube CAD gives me some not-so-good results, but I haven't built anything yet. I might end up just modifying the second gain stage to run the tube at a lower plate voltage......
3. .....or I may shove a pot before the WCF for level control and will be stuck with a cap. Two of them, actually.
4. I calculated the feedback resistors and approximated them to Mouser's stock Xicon 1/4W 1% resistors. They're sized for 3dB steps, but tube circuits being what they are, they may need some readjusting before the final version.
5. COMMENTS PLEASE!!! Last time PRR and CJ saved me hours of pointless prototyping!

Peace,
Al.

Why do you want a WCF. Maybe a TIP50 with a constant current emitter leg.

[quote author="Gus"]Why do you want a WCF.[/quote]

So the feedback network doesn't load down the gain stage, to get low output impedance, linearity, etc...

[quote author="Gus"]Maybe a TIP50 with a constant current emitter leg.[/quote]

I don't like transistors... I don't understand them!

Peace,
Al.

Here's a circuit of mine with a WCF output stage that has proven to be very capable.
http://electronicdave.myhosting.net/miscimages/lineampE.gif

Just arrange your second stage so that the plate sits at about half of the WCF's upper triode plate voltage, and you can direct-couple to the grid of the WCF.

But how to predict the WCF plate voltage? Easy. This WCF is meant to be run at 15mA of plate current. If your B+ is 300V,

V = I * R = .015 * 680 = 10V

therefore 300V -10V = 290V

and 290V/2 = 145V.

You could increase the size of the plate resistor on your second stage, or bias it richer so that more voltage is dropped across the plate resistor.

Thanks Dave! i got a couple of questions for you:

1. I've been reading that Broskie article on WCF's and the plate resistor on your WCF seems a bit high. I remember you talking about why you chose that value instead of 1/gm, though... Could you please refresh my memory?

2. Why a 2:1 transformer at the output? (as opposed to, say, a 1:1?)

Thanks!

Peace,
Al.

1. I found experimentally that a value somewhat higher than 1/gm gave higher symmetrical output swing before clipping, at least in my application.

2. A 2400 ohm load is easier to drive than 600. You can drive 600 directly with that WCF circuit, but maximum before clipping is only +24 to +25dBM into 600 ohms. The stepdown xfmr buys you a couple more dB of headroom. If you're using the preamp primarily with modern gear with a high-Z "bridging" input, by all means use a 1:1 transformer with a fairly high primary inductance, or omit the output transformer altogether. Into a high-Z load, the WCF can easily manage +30dBU or better. Its internal source impedance is about 75 ohms.

ha ha, you don't understand transistors, and I don't understand tubes! :green:

[quote author="NewYorkDave"]1. I found experimentally that a value somewhat higher than 1/gm gave higher symmetrical output swing before clipping, at least in my application.[/quote]

Yes, it's all coming back now. I'll definitely try different values out when I get a chance.

[quote author="NewYorkDave"]The stepdown xfmr buys you a couple more dB of headroom. If you're using the preamp primarily with modern gear with a high-Z "bridging" input, by all means use a 1:1 transformer[/quote]

I understand that... I meant to ask something more like: Did you actually try a 1:1 there?

[quote author="Svart"]I don't understand tubes![/quote]

These books are getting me started: Bruce Rozenblit's "Beginner's Guide to Tube Audio Design" and Morgan Jones' "Valve Amplifiers".

Peace,
Al.

Yes, I did try a 1:1, and it will only do +24 or +25dBM into 600 ohms. With a 2:1 transformer, it can go up to +27 or +28dBM before clipping. If you're never going to have to drive 600 ohms, it's not an issue for you.

Duly noted, Dave. Thanks.

Are you using the Edcor transformer on that circuit? How's it working for you?

Peace,
Al.

I used a 10K:600 with only half the primary connected, but that's not optimal. (Unused windings increase losses). I'm awaiting my custom 2400:600 from Edcor. In the meantime, I might give a 600:600 a try using half of the secondary... I also have a 600:150 I can try (most likely the best option). Basically, what's needed is a 2:1 ratio with a high enough primary inductance.