WE 407A Push-Pull Mic Preamp

[CJ]

Great! An amp a day....

Too much current thru that 2k pot, or is it a wirewound variac?

.05 with No feedback?
What if....

I can wind you a custom output for this project if you can give me a figure for
plate to plate resistance in this circuit?

You could get feedback ideas from the 5116, looks like you swiped the balnace pot already, so what the heck?

 

[Larrchild]

Actuall, the 2k pot should be a 1k with 500 ohms built-out on each leg. It's simplified there. All i had at the time.

 

[CJ]

Any of that WeCo "ring" when you tap on those tubes?

 

[Larrchild]

Lemme terminate the input iron and crank my newly acquired Tek DA 4084 Programmable Distortion Analyzer up until I can hear my voice in the room, and I'll tell ya=)

 

[Larrchild]

Yeah, CJ..the plastic alignment tool to the side makes a nice "ting" against the shield at 50db+analyzer gain, lol. But thats how we get our sustain!

Hey for an output transformer, I'm thinkin M6 10-15k pri (10k mite have more hf) Center tapped pri, and the tertiary winding need'nt be center tapped. (I need differential-feedback).. I can use 2 resistors to ground ala Fairchild sidechain amp.

 

[CJ]

I used vinyl grommets to make a tube damper for the socket and tube.
So you have to give the chassis a pretty good jolt to get it to ring.

I was thing same thing, steel core, but maybe an M4 C core from mag met.
One size bigger than Langevin so I don't go blind winding the coils.

Maybe an 0.001 gap just in case.
I need max voltage swing so I can get gauss and tube current so I can get wire size.
I guess 10 ma from both without looking.

 

[Larrchild]

10-20ma +20 dbm Tertiary can just be some outside wraps. 10%-20%.

If you build this circuit, use that general layout and grounding scheme and short leads.
When you use tv tuner tubes with gain above 100 mhz, you have to sort of build it like a tv tuner. :roll:

I may just shockmount that whole L bracket!

 

[CJ]

Yes, V76 trick. Good idea.
Whats your maximum plate to plate voltage swing?

 

[Larrchild]

Edited for updated schematic.
Here's the way I want to try feedback. Tertiary widing into 2 resistors, then to the cathodes of the 1st stages.

The P-P voltage will be higher on this output xfmr than the old one with 27k's in the plates.

 

[CJ]

I was thinkin, 0.5 percent THD, why mess with the feedback can O worms.
Plus, you want a little something in there, right?

Do you need pwr supply isolation resistor and cap for first stage with this transformer version?

I can do a ct feedback winding, since this is a dual coil design, just put 20 turns on each and splice them.


Maybe lower noise if you drop V1 B+.

 

[Larrchild]

I was thinkin, 0.5 percent THD, why mess with the feedback can O worms.
Plus, you want a little something in there, right?

yeah, I just wanna dial enuff in to take the bite off. That THD goes up fast with level.

Plus putting the output iron in the loop is cool.

Yeah, I'll build out the v1 B+ with an R and a cap and see what voltage its happiest at..

If the output xfmr is split-bobbin then yeah, rock with a center tap on the fb winding! excellent.

 

[CJ]

You will have the reduction from the step down and reduction from the feedback.

Trying to back calculate max voltage swing on the plates.

 

[Larrchild]

Right now, at full-steam, I have 20V rms @ 600 ohms on the output. 4:1 xfmr

So 80v x 1.414 = 113.2 P-P ac volts. But thats with the 27k's in.

 

[CJ]

OK, 8000 turn pri 2000 turn sec 200 turn nfb
4:1
80 henries to 5 henries

pri-sec-NFB--sec-pri

so each coil

2000-500-100NFB--500-2000

lets try 10 layers so 200 turns layer so 0.0030 wire pri- 40 ga

and 8 layers @ 125 turns per layer so 0.0050 wire sec- 36 ga

so single layer of 100 turns 0..0050 tertiary

well below 18 kilogauss, so should sound cool.

Lets try this 5116 C core...

 

[Marik]

Hey Chris,

Will it hold some residual DC?

Since we are on custom trafo, how about getting crazy and making the thing both anode and cathode coupled, a la Quad II? :shock:

 

[PRR]

> THD goes up fast with level.

Try a higher impedance. 10K is low for a pair of small triodes. WE's sheet suggests 40K. CJ will curse the mighty-fine wire, but the tubes will be happier. Gain won't go down as fast as ratio goes up: you are loading the tubes pretty hard, they will make more gain with a lighter load.

If you change the gain-control voltage quickly, the outputs will cut-off until the coupling caps charge to the new DC levels. You might want an interstage transformer if you are going where you seem to be going. (But if you do that, the NFB won't work.)

 

[Larrchild]

If you change the gain-control voltage quickly, the outputs will cut-off until the coupling caps charge to the new DC levels.

My first post here was a quote from you, PRR from an old forum describing how the long-tailed pair in the Altec 436 has that coupling cap cutoff malady! I dig what you are saying, man. The 10uf on the control line makes a smooth, if somewhat delayed mic gain response.

I updated that 2nd schematic on page 2.

 

[Larrchild]

Since we are on custom trafo, how about getting crazy and making the thing both anode and cathode coupled, a la Quad II?

I Do want to build a linestage with a Unity-Coupled Line output xfmr soon.=) (Driver is non-simple! or iron)

Good one!

 

[CJ]

Does that negative voltage thast controls the gain give you distortion at low settings?

What if I tap the NFB winding every 10 turns, you could then hook those wires to a rotery switch and have the worlds first transformer feedback controlled atten control?

40 K , your killin me! Than God for the turns squared thing.

 

[Larrchild]

Does that negative voltage thast controls the gain give you distortion at low settings?

Naw, it's not as bad as I would have thought if the tube sides are matched.
I'll send ya some curve tracer matched 407a's or 6BQ7's once I make a jig.

So I put a UTC ER-841 15k ct 600/150/50 in the output side so I could use one set of output windings as feedback. With just 6db of feedback, It's much lower THD.
I will post. Looks like the feedback topology works perfectly.=)