One Tube Pre

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
[quote author="cuelist"][quote author="toby"]just out of perverse curiosity would there be any major drawbacks to not using a transformer on the input? you could feed two differentially for a balanced input?[/quote]

One of the key functions of an mic input transformer is to step up the source resistance (low Z) to the optimum source resistance of the first stage of the amplifier thus achieving lowest possible noise voltage.

Transformerless transistor based pre amps uses specially selected transistors or multiple paralelled transistors to achieve decent noise results without a transformer.

With a differential input w/o transformer using tubes, there is no practical way of bringing the optimum source resistance of the tube to match the low Z of the mic.[/quote]


i understood this preamp had a low input impedance?
 
Read the paper I posted earlier in the thread. Pay particular attention to page 2.

For mic use, having low input impedance is not desirable; what is desirable is to have low noise when operated from a low source impedance. That ain't the case with tubes, so we use a stepup transformer. The added voltage gain is just a fringe benefit.
 
[quote author="toby"] i understood this preamp had a low input impedance?[/quote]

Load impedance, what the microphone "see's" and source impedance, what the first stage of the amplfier see's are two different things.

In the case of a mic pre with a transformer on the input side, the turns ratio squared is the impedance ratio. So, for 1:10 transfomer, the impedance ratio is 1:100.

So, let's say the source impedance (the mic) is nominally 200 ohms, this gets transfomed to 200 x 100 = 20k. This is the source impedance that the first stage of the amplifier see's. It is the first stage of the amplifier that determines the noise.

From the mic's perspective, the load impedance that it see's is transformed at an equal ratio. If the amplifier has an input impedance of, say 100k, this get transformed down to 100k / 100 = 1k. So the mic is loaded by 1k.

And like NewYorkDave said, the "free" voltage gain is a bonus (or curse in case of hot signals) rather than a reason for the transformer to be used.
 
i was talking specifically about analag's pre amp.

the reason i mentioned it is i could do with a booster amp for a specific mic, and i know the output impedance of this mic is 600 ohm (which i am aware is higher than usual).

in this situation i'm not sure i understand why there should be any more noise when the 150:600 ohm transformer is omitted, and the source is the 600 ohm secondary of this specific microphone, rather than the 600 ohm secondary of the origional input transformer?

also the acticle NYdave posted briefly mentions using parrallel input valves to reduce noise. does anybody have any experience of this specific to grounded grid micamps?
 
[quote author="toby"] ...in this situation i'm not sure i understand why there should be any more noise when the 150:600 ohm transformer is omitted, and the source is the 600 ohm secondary of this specific microphone, rather than the 600 ohm secondary of the origional input transformer? [/quote]

The transformer itself has no set impedance.

150:600 simply means an impedance ratio of 1:4. The source impedance of 600 ohm you mention is thus presented as 600 x 4 = 2400 ohms to the amplifier. Without the transformer it would of course be the raw mic source impedance of 600 ohms.
 
[quote author="cuelist"]Without the transformer it would of course be the raw mic source impedance of 600 ohms.[/quote]

Which seems about right for a no transformer ground grid input.
 
Again, read the paper I posted, especially page 2.

Good transformers are expensive, always have been. But there's a reason they were used throughout the tube era. As PRR says, the Dead Men knew things that we sometimes have to rediscover for ourselves. The grounded-grid amplifier has been around for a long time; the fact that we don't see it in old audio circuits has nothing to do with a lack of imagination on the part of the designers. Some study of the old literature (which is NOT outdated as the laws of physics haven't changed) shows why.

Otto Von Bismarck said, "Fools you are... to say you learn from your mistakes... I prefer to profit by other's mistakes and avoid the price of my own."

But hey, we're still gonna try things whether the conventional wisdom says they're good ideas or not. It's just human nature.

Even when I posted that first mic amp idea recently, I knew right off the bat that it wasn't optimal for noise because of the fact that the second triode was being driven from a low source impedance. But it made for an interesting mental exercise, especially when PRR came along later and explained exactly why it was a lousy idea on my part :wink:
 
[quote author="cuelist"][quote author="toby"] ...in this situation i'm not sure i understand why there should be any more noise when the 150:600 ohm transformer is omitted, and the source is the 600 ohm secondary of this specific microphone, rather than the 600 ohm secondary of the origional input transformer? [/quote]

The transformer itself has no set impedance.

150:600 simply means an impedance ratio of 1:4. The source impedance of 600 ohm you mention is thus presented as 600 x 4 = 2400 ohms to the amplifier. Without the transformer it would of course be the raw mic source impedance of 600 ohms.[/quote]

would the source impedance not be 1/49 (1:7 transformer in Mic) of the output impedance of the valve in the microphone rather than 600 ohm then?
 
> the valve in the microphone

Hold it.

If the "mike" is really a mike plus an amplifer, noise is different.

First: passive mikes. Dynamics, ribbons. When transformed to 150-200 ohms, they have a thermal self-noise of about 0.2uV. We can't do anything about that. The noise of a high gain low current triode is about 2uV. Therefore we use a 1:10 transformer to get the mike self-noise up near or above the tube self-noise, so the tube adds very little noise.

Now "active" mikes. As long as you have an amplifier in the mike, you may as well bring up the output level. At 74dB SPL, a dynamic makes around 0.2mV signal and 0.2uV noise, a big condenser makes 1.2mV signal and 1uV noise. Both give the same signal/noise ratio, the same 14dB SPL equivalent acoustic noise. But the high output of the amplified mike means our console input does not have to be so heriocally low-noise. We could use a 1:2 input transformer and get acceptable noise (as in analag's recent plan), or a 1:3 or 1:4 ratio to completely overcome console input tube noise (but not with the low input impedance of analag's common-grid plan).
 
NYD sez:
PRR came along later and explained exactly why it was a lousy idea on my part
And I followed you right into the gaping maw of wrongness, Dave. Posting low Miller circuits for ya and everything!

We should have remembered Dale's sage Miller in Mic's commentary. But alas, at the time, he was like Lassie, saying "Timmies' trapped in the mine!"..yet all we heard was "Woof!"

A few of those will really get you using search and meta's I'm finding.. :thumb:
 
[quote author="skipwave"]but I can't remember if I ever listened to the 5965s. [/quote]

I have a stereo optical comp with 5965's for gain, attenuator amp and 6N1P's for CF. The gain stage is hybrid with depletion mode MOSFETs in place of plate resistors. Tubezistor, warm like tube but fast like transistor, truely better than both.

analag
 
[quote author="analag"]I have a stereo optical comp with 5965's for gain, attenuator amp and 6N1P's for CF. [/quote]

That sounds interesting. If I didn't have a pile of projects I would whip up your 5965 pre.

I put the 5965s I have into the One-Bottle circuit as a test. They work fine, although with the gain cranked up they were noticeably noisier than the AV7s. That probably won't be an issue at the gain in your design. Either way, I'm shipping them off to Svart so he can tinker.
 
I noticed the same thing about the 5965(A), the were a bit noisier than the av7, at least on the NYD two-bottle.

I've got a few of these left so I wouldn't mind building your tubesistor comp one day.

Analag, what's the opto part of your compressor based on? something you whipped up or borrowed? How's the sound?
 
:thumb:

I can't wait for those firebottles either!

I'm really catching the valve bug.. :shock:

And I want to try the hybrid designs.
 
The opto part is driven also by 5965s through 0.1uF caps going straight into the vactrols, it was stated an LED could not be driven this way, but it was then discovered it is indeed a superb level detector.
The hybrid comp is clean, fast, warm, with nice low frequency definition.
http://home.pacifier.com/~gpimm/Active_loads_and_signal_current_control.html

analag
 
That sounds very interesting. Were there any threads about this LED system here or prototype schematics even? Sounds like a bit too big a design challenge for a beginner like me.
 
He's back!
Good to see ya!
Thank you Lord, for sparing me the guilt of driving off yet another member of the Lab.

This from "Tube Lore":

"5965- computer rated 12AV7 (GE,RCA) "A" has controlled cutoff balance (never heard that one, what is it?) and 10,000 hour rated life; is derated to 2.2 W per section (G.E, RCA, Sy, Tf) 6-2-1952 Also see 7062.

GE manual says frequency halfer. Now why would one tube be a better divider than the next?

Also says Ic=140 uA on the 5965 and Ic = 200 uA on the 5965-A?
What the heck? Since when do tubes have collector current? :oops:

Darn for a minute there, I thought it said "see 6072" dang it!

THC Dyslexia, teen fallot.
My math teacher used to be like that.
Hmmmm...
 
[quote author="CJ"]He's back!
Good to see ya!
Thank you Lord, for sparing me the guilt of driving off yet another member of the Lab.[/quote]

So it's been done before....can't give you the pleasure of drivin me off. I realize I love the "Lab" a lot more than I thought. Besides I represent the tube hybrid technology.
Hybrid%201.PNG
Here is a simplified version of my hybrid 5965 line stage. It's been running for a little over a year and a half without problems.

analag
 
Back
Top