Mic levels, Tx ratios and tube (triode) grid bias...

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

alk509

Well-known member
Joined
Sep 3, 2004
Messages
1,207
Location
MA, US
Hey boys and girls,

I was playing around with some ideas for a mic pre-amp and now I have some questions:

I've seen mic pre's that put the microphone signal through a high ratio (1:5 or even 1:10) input transformer, bias the grid of the input tube to about -2 or -3V and then run with the output to a second stage, maybe add feedback, etc., etc... Now, with modern condenser mics you can get sensitivities of >10mV/Pa with maximum SPL's of around 130dB... That's 640mV (RMS) at 130dB, which translates into 3.2V RMS - ~9V peak-to-peak - after a 1:5 transformer!

So, without further ado, here's my questions:

1. Is my math correct?
2. If my math is right, -2V of grid bias isn't gonna cut it, now is it? What's a good grid bias voltage (Vgk) range for a tube mic pre with an input tranny?
3. What's the Vgk on the first stage of, say, Jakob's G9? Or Vgk on the first tube of the "insert the name of your favorite tube mic pre here"?
4. Is all this reasoning just fundamentally flawed? How so? Am I just stupid?

Thanks for your patience!

Peace,
Al.
 
Just to clarify where I got those numbers from...

1Pa = 94dB SPL

So, if a 1Pa sound at the mic's diaphragm makes a 10mV signal, a 2Pa (100db SPL) sound will output a 20mV signal and so on

94dB SPL => 10mV
100dB SPL => 20mV
106dB SPL => 40mV
112dB SPL => 80mV
118dB SPL => 160mV
124dB SPL => 320mV
130dB SPL => 640mV

Then the 1:5 transformer raises that level to 5 * .64V = 3.2V (RMS)
And 3.2V RMS * 2.828 = ~9Vp-p
 
> Is my math correct?

Yes, though your assumptions are debatable.

Outputs at 94dB SPL can run up to 12 or 25 milliVolts, more than doubling your levels.

And you assume that these transformer/tube preamps were made to work with these mikes. They were NOT.

Lookit. There are two general classes of "Pro" mikes. Passive mikes (dynamic including ribbon) tuned to work over most of the audio band will give output at 94dB SPL of about 1 to 2 mV. On the other hand, there are mikes with active electronics inside, mostly condenser, which can be designed for any output level but in German tradition will output 10 to 20 mV at 94dB SPL.

In days when tubes were mainstream, in US and UK broadcast practice and the recording industry built on that custom, all mikes are passive. It is the preamp's job to make the most of a weak signal, which forces certain design decisions. And most joints would not allow a 130dB SPL source in the building. The US and UK tube preamps never expected to see a level over 50 or 100mV. Putting hot condensers through these inputs is clearly "abuse".

Some German studios were built to use only high-output condensers, easing line-noise troubles and board input design, at the cost of needing electronics and power in every mike. In effect they moved the first stage out of the console and into the mike. "Mike inputs" for these systems are more like high-gain line inputs, because they do not have to scratch for every last microvolt of noise.

Note another fact. The self-noise output of a passive mike will be 0.1 to 0.2 microvolts. The self-noise of the high-output active mikes is more like 1 to 2 microvolts.

The quick and easy fix is to use the -10dB or -20dB pad on the high-output mike. That's why it is there! These mikes will not interface to passive-mike inputs over the full range of musical levels.

A more elegant and marginally less noisy fix is to use a low-ratio transformer. Because the self-noise voltage of these mikes is as high as many amplifiers, you need little or no transformer step-up. 1:1 may be suitable. Low-impedance transformers tend to be flatter and less lossy. Because the self-noise of many solid-state amps is much lower than the self-noise of the high-output active mikes, a transformerless input can be built without the extreme design needed to do good transformerless from passive mikes.
 
Well the math looks OK. But we would normally use a 20dB pad on a signal thats 120dB SPL+. There's another end of this spectrum. Sometimes
we like to record quite sounds too. So the comprise is us a medium ratio transformer 1:5 or so, feeding the first stage. as far as grid bias on the typical first stage, 2 to 3 volts is about normal. It should be designed where normal operating conditions will not draw grid current.

hope this helps

RonL

http://nashaudio.com
 
Thanks PRR and Ron!

I can definitely see your point(s), I already gave in to the idea of having to shove a pad in there. And I realize that microphone output levels vary GREATLY between microphones...

Where I'm having trouble is deciding on a compromise that will best let me plug in a low output ribbon and super high output condenser in the same mic pre.

So it makes sense to use a 1:5 input Tx instead of a 1:10 (thanks, Ron), and put a -20dB pad at the input (thanks, PRR). But what about Vgk? What's a good range of grid bias voltage for an "all-purpose" mic pre? Does it make sense to bias the grid to -6V for a mic pre? How come nobody does it? Oh, somebody does do it, you say? Who?

Thanks again,

Al.
 
Hi Al,
But what about Vgk? What's a good range of grid bias voltage for an "all-purpose" mic pre? Does it make sense to bias the grid to -6V for a mic pre?

It all depends on the circuit and the tube used, most is already in the datasheets. With the usual preamp tubes, biasing to -6v does not make sense because you get into the cutoff area instead of gaining headroom. To get started: http://ax84.com/media/ax84_m226.pdf

HTH
;Matthias
Just a interested lurker having delurked...
 
Hey Matthias,

[quote author="EZ81"]It all depends on the circuit and the tube used[/quote]

Naturally. The question, however, is not how to decide on the operating point of a specific tube, but rather what grid voltage is a good voltage to use, assuming a 1:5 input transformer, so that the tube - any tube - will not have problems handling signals from a hot mic, but will still be able to take a low output ribbon mic and amplify its signal enough to drive other equipment without too much noise.

I guess the question answers itself if you assume you're always gonna use a hot mic. In my 10mV/Pa mic example, the output from the input transformer was 9Vp-p, so you need at least a -4.5V bias at the grid Of course, whatever tube and operating point you use has to be able to take the full negative signal swing without going into cutoff.

That same circuit will probably not have enough gain, however, for a 1mV/Pa mic. So you switch a pad in when you're using a hot mic and take it out when you're using a ribbon.

The question still is: What's a good compromise for Vgk, assuming you'll use a 1:5 input Tx and a switchable -20dB pad, so that your mic pre will be able to handle as many different signal levels as possible???

[quote author="EZ81"]With the usual preamp tubes, biasing to -6v does not make sense because you get into the cutoff area instead of gaining headroom.[/quote]

Actually, I didn't just pull those -6V out of my ass... I had the datasheet for JJ's 12AU7 in front of me and -6V looks like a good and definitely possible grid voltage for that tube...

...I think?

Peace,
Al.
 
Sorry, i got you completely wrong :( and came across like a idiot :oops: : Of course you're right.

The G9 seems to use a clever solution for this problem with the whole fancy gain switch/NFB network...


Peace :guinness:
;Matthias
 
[quote author="EZ81"] The G9 seems to use a clever solution for this problem with the whole fancy gain switch/NFB network...[/quote]

Yes! I was actually thinking along those lines too for my design (variable gain through negative feedback). Maybe with some kind of mu-follower or White CF at the output instead of an SRPP stage. Not because i don't like SRPP, but because I like to be "different" :wink:

Peace,
Al.
 
The V76 has a 1:30 input transformer and is made to handle the hottest mic.
The main point is you are missing is that you will never set your preamp on 10, put a snare drum on your hottest condenser with the pad off and put the output into the board.
The excessive gain possible is for listening to crickets in the next county.
If you are jacking up your level to 6 volts at the input of stage one, there is something wrong.
Also, remember that your transformer has an inserftion loss, so you will never see 6 volts at the other end.

Jensen vs Ollie Iron input:
jensen_tab2.jpg
 
Back
Top