Proper Way to Measure Preamp Gain

Hi Everyone,

I'm currently trying to set the (internal) gain of a mic preamplifier for exactly 65db of total gain. I have a Fluke 187 DMM which has the db metering function. What I've been doing is feeding the input a 1kHz tone from ProTools at -60db (trying to approximate a mic signal), measuring the db with the Fluke attached as follows:

RED: On pin 2 of input to preamp
BLACK: Attached to chassis ground of preamp

After I read about -55db on the Fluke, I attached the RED probe to pin 2 of the output XLR (with black probe still attached to the chassis ground of pre) and read about -7db with the volume of the preamp full up.

This seems weird because when I plug a mic into the preamp and mic something up, the pre seems to have way too much gain, even at a very low setting on the input gain knob. When compared to my Sytek pre (which is also supposed to have 65db of total gain), the Sytek is much less sensitive using the same mic, source sound, and gain setting. What might be going wrong here?

According to the Fluke, I should have 48db gain on my pre. Yet, it's overloading on most sources. Does it have something to do with the way ProTools is sending the signal to the pre? Or am I reading it improperly? I would do a simple voltage test on the input and output, then convert to db but I feel like the same problem will be present in my readings...

Thanks for any help.

Not sure of your preamp but, -60 from ProTools re-amped up to
-7 is going to sound pretty ratty. The best was is to use a pad of
say -30, -40, between Protools and the pre. This lets Protools
use more bit for your test signal. You can still measure you signal
at the pre input to set you input.

If you have a scope:
Bring the input up untill -1 before clipping (as viewed on the scope)
Measure the output. Then measure the input. Use the Fluke if it does
dBv or Use the formula:

dB=20log*v1/v2 to calculate gain in dBv.

Good Luck

RonL

It's possible the Fluke is reading noise at -55dB (I assume this is dBu, but your manual will tell you) rather than the actual signal. What does it read when ProTools is not generating signal?

I think it also might be worthwhile to hang the Fluke across pins 2 & 3 of input and output rather than from pin 2 to chassis ground. Is this a transformer-coupled preamp?

Finally, another good way to inject signal is to use a DI box, driven by ProTools.

Peace,
Paul

Amen to Ron's comments...take a STRONG signal (0 dBu is convenient), then PROPERLY pad it. PROPER means understanding the source and load impedances.

In my test bench, my function generator has a 50 Ohm output impedance, yet I need to drop the level as WELL as providing a nominal 150/250 Ohm output impedance from the pad into the mic preamp to "simulate" a typical mic.

Hence, I need a 50 Ohm-200 Ohm (I decided upon the middle range of typical mics) pad in order to properly drive/terminate the input of an xfmr. coupled mic preamp.

A Fluke, or whatever, multimeter isn't the best measuring tool. For one thing, the better ones are "broadband" and will allow the intrusion of spurious noise, well out of band from audio.

I am NOT saying that out of band signals should be ignored..just that Flukes, etc. are not intended for this application.

Bri

http://groupdiy.twin-x.com/albums/userpics/10031/preampgaintest.png

A proper AC millivoltmeter is recommended for measuring V1 and V2. The signal source should be a sine wave, at a level 40dB (100X) higher than the desired input level to the preamp.

Thank you all. Very helpful!

[quote author="NewYorkDave"]http://groupdiy.twin-x.com/albums/userpics/10031/preampgaintest.png

A proper AC millivoltmeter is recommended for measuring V1 and V2. The signal source should be a sine wave, at a level 40dB (100X) higher than the desired input level to the preamp.[/quote]

I know this is an awfully simple question but would this setup suit my 600 Ohm sig. gen. impedance-wise, or would it only be suitable for low-impedance voltage sources?

Thanks,

Roddy

As long as you can make accurate measurements of the 600-ohm generator's output voltage, who cares what its output impedance is? The exception would be if the generator were going into current limiting or something, or if you had no voltmeter and were relying on the calibration of a dial based on precise output loading. But then you wouldn't have a meter to measure your preamp's output anyway.

In any case, by the time you devise your 40 dB attenuator you can have reasonably low generator loading and what you want for output Z.

If you want an attenutator with a 150 ohm output Z and 40 dB of attenuation, that could be realized with a U atttenuator with two 3.01k's and one 154 ohm (picking the nearest standard 1% R values). Adjust the generator using your voltmeter with everything hooked up and away you go. You are loading your generator with no less than 6k.

EDIT: the R values will depend a bit on the actual generator output impedance. If it is really 600 ohms then the 3k R's should be adjusted downwards to 2.7k (2.74k nearest 1% standard value).

[quote author="bcarso"]If you want an attenutator with a 150 ohm output Z and 40 dB of attenuation, that could be realized with a U atttenuator with two 3.01k's and one 154 ohm (picking the nearest standard 1% R values). Adjust the generator using your voltmeter with everything hooked up and away you go. You are loading your generator with no less than 6k.

EDIT: the R values will depend a bit on the actual generator output impedance. If it is really 600 ohms then the 3k R's should be adjusted downwards to 2.7k (2.74k nearest 1% standard value).[/quote]

FWIW...
Needing a simple 40dB pad myself as well and being lazy I went to
http://www.uneeda-audio.com/pads/ for ready to go values
(40 dB att, 7500 - 150 - 7500, U-pad), which make of course sense - - had I been interested in quick i.s.o. lazy I had of course multiplied half of 150 by 100 and had directly arrived at this 7k5-value for the incoming legs.

The values you gave here above are different though, can't seem to make 40dB att from these, more like 32 dB att.
What am I missing ? Are you perhaps already directly compensating for an eventual unbalanced to balanced 6dB difference ?

For my 'situation' (single ended generator source impedance 600 Ohms, desired source-impedance as presented to the mic-pre 150 Ohms) I could compensate for the 600 by using 7k125 i.s.o. 7k5.

Regards,

Peter

Looks like I calculated for a factor of 40 (32dB) and not 40dB (factor of 100). Musta been loaded....

[quote author="bcarso"]Musta been loaded....[/quote]

Properly loaded, I should hope.

I come for the knowledge, I stay for the puns.

[quote author="skipwave"]I come for the knowledge, I stay for the puns.[/quote]
...and the lunacy comes free!

Using a FLuke to measure super-low levels is a bad idea... the leads are unshielded, and they can usually be expected to pick up mondo noise.

Keef

Hi,

And let's add the obvious: better loaded than terminated :razz:

Not meant as nitpicking, you know that... was wondering if I had missed something, hence my previous one.

But ahhhh, life is good - one less calculation-oops in this world & the weekend has started....


Enjoy,

Peter

[quote author="clintrubber"]Not meant as nitpicking, you know that...
Peter[/quote]

Hell no. And that's a big nit. It amuses me though that it took a year and a half for someone to notice it.

EDIT: If I had gotten lucky, someone would have asked for the values for an attenuation of 29.35307086 dB. Then if I had again mistakenly calculated the values for a factor of that same number, it would be the same answer.

It's an amusing exercise to calculate that value BTW.

So i need to do some 'proper' gain tests on a pre amp we built a few years ago.

I've a NYD Re-amp box at home which has a 10K:150 transformer inside. That would that give the correct load to the mic input without needing to make up a pad right?

(Test signal coming from Pro Tools)

mrclunk said:

So i need to do some 'proper' gain tests on a pre amp we built a few years ago.

I've a NYD Re-amp box at home which has a 10K:150 transformer inside. That would that give the correct load to the mic input without needing to make up a pad right?

(Test signal coming from Pro Tools)

Click to expand...

The turns ratio of a transformer is the square root of the impedance ratio so that transformer gives you about a 66:1 drop.

I prefer resistive pads as being easier to quantify termination impedance and ratios.

JR

It's very odd only one member noticed the fact that you ran UNBALANCED (pin 2 to GND) measurements on BALANCED connections (pin2 to pin 3). In the best case, you have 6dB error on the input and 6dB error on the output, so your pre has probably ca. 60dB gain, in the worst case, who knows?...

i got myself one of these for stuff like this:

http://www.alva-audio.de/en_products.php?page=content/products/en_cableware_tools_product&subpage=content/products/en_tools_miccal

greetings,

Thomas

why not just build a 65db pad, place it before the preamp, send a 0db signal in and adjust untill you get 0db out?...

...hmm...if that was a bad idea..technically, I just asked "why not"... ;D

j

thanks guys, ended up using the sig gen from the desk.
Turning it up until the mic pre output started clipping, then backed it off a touch.
Measured the voltage in and out. Calculated gain.

That Signal generator key fobs real cute though....