Western Electric 417A Tube Preamp

This is an attempt to un-confuse the One Bottle thread.
Please post any 417A project info here so Dave's amp can have it's own space.

Here is a scan that got me interested:

http://vacuumbrain.com/The_Lab/TA/Western_Electric/we417_1.bmp

and the circuit ....



Should have done this a while back, so Dave has a free hijack to any of my threads.


Data sheets lifted from the other thread:


http://www.pmillett.com/tubedata/HB-3/Miscellaneous%20Types/407-A.PDF
http://www.pmillett.com/tubedata/HB-3/Receiving-Type%20Industrial%20Tubes/5842_417A.PDF
http://www.pmillett.com/tubedata/HB-3/Miscellaneous%20Types/408-A.PDF

http://www.westernelectric.com/products/we417a/417A.pdf


And a pic of a proto 417 pre:

that's funny
:green:

cos recently I 've been talking to someone else about a one-bottle with 40 or so gain and true 600 drive

I think you will see it ... soon'ish

Yes, and here is a commercial product that uses the 417A:


http://www.enjoythemusic.com/magazine/equipment/0902/lammlp2.htm

And this Dodd headphone amp:

http://www.doddaudio.biz/products/headphone.asp

CJ, I asked this in the Dave´s thread, but here it goes again...

If loaded with a 10k:600 output transformer, how much gain will the active part of your circuit offer? Don´t count the input and output step-up and down. I would like to know what tube only gain figures you are getting.

Thanks for one more nice circuit!

Damn, CJ. That almost looks too neat to be one of yours. :razz: :thumb:

I know. It's tough to make 3 or 4 components look busy.

Raf, I will measure the gain tonight with scope and generator and subtract the transformer to see what we gots.

Did you ever get the C cores I sent?

Yes, I got the C cores. But this is out of topic... Guilherme is doing the first langevin transformer for me, but he is getting very high resistance with the wire info you informed. Maybe you got the wire size wrong, is it possible?

[quote author="Kev"]that's funny
:green:

cos recently I 've been talking to someone else about a one-bottle with 40 or so gain and true 600 drive

I think you will see it ... soon'ish[/quote]

Wow. Looking forward to that... the ultimate minimalist pre. :razz: Could you drive 600 ohms if you added the appropriate transformer to CJ's design?

Kiira

Raf, that old wire was probably a single build.
Today, most winders use a triple build.
So for the same diamater wire, you will have three times as much insulation, and 1/3 the copper. This is why his resistance measuements are so high.

This is why when I do a take apart, I try to not break a 100 ft section of wire during unwinding.
I measure the resistance and multiply by 10 to get the ohms per 1000 ft spec.
I can dig up the ohms per 1000 ft on the Langevins if you want.

I measured the gain of the 417 circuit last night.
Without the input trans, I get a voltage gain of about 24.
So with the 1:14 DuKane input, i get 14 times 24 = 336 overall voltage gain without an output trans.

If you look at the schematic, you can see that driving 600 ohms straight from the plate is probably not impossible, as the output impedance of the tube is fairly low. Look at the plate resistor.

CJ, I got some pics of the internals of my single bottle prototype, I'll get them up soon. still a little messy but very little noise. I used a toroid and a small PSU i built with plenty of capacitance and DC heaters.

I can dig up the ohms per 1000 ft on the Langevins if you want.

Click to expand...

Yes, that would be nice.

About the one buttle, it seems that the maximum gain possible with a 1:15 input and a 4:1 (10k:600) output would be 40dB. Not bad, but also not great for some apps. Maybe I could include a transformerless output for unbalanced 10k or greater loads, like direct to ADC converter connection, and that would give me 12dB more, for a total of 52dB. Not bad for one triode preamp, he?

ok nothing special:

www.theopiumdenproductions.com/SUC50490.JPG
www.theopiumdenproductions.com/SUC50491.JPG

hey CJ, what about paralleling these up for more drive and/or gain?

what's the sweetest passive EQ you can think of? thinking what i'm thinking?

> Without the input trans, I get a voltage gain of about 24.

My abacus says, Rk fully bypassed, un-loaded, 250V supply, Gv=30. Which agrees with your 24 if you had 10K load and lower B+.

It also means that a 10K:600 transformer will give gain in tube and transformer of about 6, or 15dB. Add that very high-ratio 1:14 transformer, Gv=23dB, total gain 38dB.

Also, with this resistor-loaded deal, maximum output is 7.5Vrms, about the least we would want. If B+ is 125V, max ouput after the transformer is only 3Vrms.

Also, with R-C coupled 10K primary we are likely to have real problems with bass resonance. ASS-uming the "10K" winding is 10K reactance at 20Hz, and the "obvious" 1uFd cap, we get a resonance at 17Hz with 10K reactance and ~2K resistance, Q=5, +14dB hump at 17Hz which will probably be +6dB at 34Hz, a substantial boom. The "neutral" tuning would use a 20uFd cap for Q=1 at 3Hz; that's a huge cap.

The "right" answer is to drop the plate resistor and coupling cap. Now we need a plate transformer that will eat 20mA-30mA.

150V 2.4K -1.4V 28mA
40V 16mA peak, 28Vrms
Rp = 2,000 ohms
0.32W +22dBm
in = 1.0Vrms
out= 14Vrms
gain= 14/1.0= 14
unloaded: Gv=21

2K4 28mA to 600Ω is not an easily found part. There are suitable loudspeaker transformers: 10+Watts SE at 2K5 or 3K, but 16Ω secondaries which is just too much loss (1:12). Wonder if any of those winders would consider using their stock primary, then wind two windings of the same gauge and half the turns, for a 2:1 2400:600 transformer.

Note also, that with a 1:14 input transformer, the maximum input is 70mV. Ample for dynamics on classical, but isn't going to eat LD condensers around a rock band. That will be true for any 46dB amp with 14Vrms max output.

[quote author="Kev"]recently I 've been talking to someone else about a one-bottle with 40 or so gain and true 600 drive[/quote]

My 12AV7 pre, with a 4:1 output transformer, does exactly that.

+20dB (or so) from input transformer, +32dB max electronic gain, -12dB insertion loss from output xfmr = +40dB. Max clean output is +18dBM (into 600 ohms), which is on par with other classic preamp/booster amp designs of the tube era.

Mine is one-bottle but two triodes within that bottle. Doing all of the above with a single triode is a bit of a challenge if you care about low distortion.

What's in your preamp, Kev? One triode or two? Maybe it's time to start yet another thread to talk about that one :wink:.

Yes was running about 140 volts B+ for low noise.
Cathode was not bypassed, so gain could go up.

A 2400:600 would mean 2:1 ratio, not hard to do.
How much pri inductance?

[quote author="CJ"]Please post any 417A project info here so Dave's amp can have it's own space.[/quote]
OK. I made this one a couple of years ago: http://stiftsbogtrykkeriet.dk/~mcs/MicPre/micpre_v2.gif

I never got the hum down far enough though...

This one works (but not for mics): http://stiftsbogtrykkeriet.dk/~mcs/Tandberg/Cookie2.jpg :green:

Best regards,

Mikkel C. Simonsen

[quote author="PRR"]...drop the plate resistor and coupling cap. Now we need a plate transformer that will eat 20mA-30mA.

...

2K4 28mA to 600Ω is not an easily found part. There are suitable loudspeaker transformers: 10+Watts SE at 2K5 or 3K, but 16Ω secondaries which is just too much loss (1:12). Wonder if any of those winders would consider using their stock primary, then wind two windings of the same gauge and half the turns, for a 2:1 2400:600 transformer.
[/quote]

I bet Edcor would do it. It might work on their low cost gapped series.
http://www.edcorusa.com/classx/tube_output/xse/xse.htm

PS.: If I weren't elbow deep in the AM864 limiter construction I would order a couple.

Well, I have a winding machine or two laying around.
And some 75EI M6 (Neve type output core).
It would take me about an hour to wind.
Pie-Sec-Pri-Sec ought to be enough sections for leakage.

Looks like 3000 turns on a square stack and a 3 mil gap would give 40 henries primary > 10 henries sec.

Cool, I will take one!