Constructing A Klein + Hummel Mic Preamp - Can Anyone Help?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

rodabod

Well-known member
Joined
May 12, 2005
Messages
2,896
Location
London
Hi guys,

I recently purchased this Klein and Hummel B74 microphone preamp module because it was cheap and had a Beyer transformer. It is designed to be plugged into a larger unit which would probbaly be a sound system or tape recorder, etc.

Anyway, I have the schematic for it which I will try and post here, and was wondering if anyone could give me advice on how to wire it as a standard mic preamp. I can deal with constructing a suitable power supply and making an enclosure for it, but I am unsure which connection on the board is which and also how to adjust the gain.

I have studied transistor theory (and have a degree in electronics) but I seem to have forgotten a lot of it and don't understand the circuit 100%.

Any help would be greatly appreciated.

Roddy

b746yp.jpg
 
Yes, it is standard preamp.
But C4 is important.
Gain is set by R1,R2 divider,
R6 is synthetic transformer load (have
lower noise figure than simple resistor)
C5 is for stability (reduce gain for high freq.)

Only one mystery, why R7. For stability pay by worse noise
figure? why. If we design transformer, we want to have low
winding resistance. And then decrease it by one resistor....
Get it fucked. May be for mass-production reasons. Bridge it by small ferite choke, pearl on the wire or etc...
If not oscillate, all is O.K.

xvlk
 
I looks to me as if it were not that easy to modify this thing for variable gain. Do you have any info on the transformer, i.e. turns ratio? If the overall gain is not too high, a good solution might be to add a second stage with fixed gain, say +20 dB, and wire a pot in between. This would have the advantage of isolating this circuit from heavy loads - it does not look like it were designed to interface to the outside world (as you already said in your post). For lower gains, throw a 20 dB PAD in front of the input transformer.

Samuel
 
from Samuel:
I looks to me as if it were not that easy to modify this thing for variable gain.
A bit of the same story for the circuit of my Beyer-preampcard.
So if your Beyer-TX has a low ratio (say 1:1) you could just use the TX and use this circuit:
p13_fig1.gif

http://sound.westhost.com/project13.htm
As PRR has pointed out before, you can't use higher TX-ratios because input-Z would become too low - or redesign & optimalization of the ESP-circuit would be in order.

If you have a 1:~3.16-TX (10dB step-up) a circuit with say a 5534 could be nice. There are various circuits out there, and just also saw a Studer-schematic. Or have a look at the Jensen-app-notes for the active part.

BTW, it looks like this Beyer doesn't need a Zobel either (the TXs on the card I have don't have anything Zobel either).

Let us know what you end up with.

from xvlk:
R6 is synthetic transformer load (have
lower noise figure than simple resistor)
Right, please refresh me on this technique. How did it work again ?
Am I confusing it with - or - is it 'electronic cooling' ?

Regards,

Peter
 
Thanks very much for your replies, guys. I really appreciate it.

The Beyer transformer is model: TR/BV35806

I don't know if this means anything.

I don't know how to calculate "turns ratio" - i'm assuming it is the same ratio as the resistance of each coil? No? I will try and find out how to calculate it.

The input resistance on the transformer (measured between e and f on the schematic) reads 60 Ohms.

The output resistance on the transformer (between a/b and c on the schematic) reads 3.5 KOhms.

According to the specification sheet, the preamp has 60dB gain.

What is my best option for making an ok preamp?

Roddy
 
The Beyer transformer is model: TR/BV35806

I don't know if this means anything.

It's literally not in 'my' book I think (see the META on TXs for the Beyer-TX-pdf-download)

I don't know how to calculate "turns ratio" - i'm assuming it is the same ratio as the resistance of each coil? No? I will try and find out how to calculate it.
See the META's



According to the specification sheet, the preamp has 60dB gain.

That's a lot, it would mean that the TX has (60 - 28 (from R1,R2)) = 32 dB of gain. You sure it's 60 dB ? Or could it be 60 times ? Or are there other gainstages following ?

The input resistance on the transformer (measured between e and f on the schematic) reads 60 Ohms.

The output resistance on the transformer (between a/b and c on the schematic) reads 3.5 KOhms.

As discussed here before, it might be better not to tax TXs too much by raw DC-currents. It's not clear where it gets not-OK, but better be safe than sorry.
From your measurments it looks like it's defin. a step-up (unsurpr.), so the ESP-circuit is not suited.

What is my best option for making an ok preamp?

You want a gain-control, right ? Simplest would be a piece of perfboard, add the TX and a 5534(low-noise version).

Regards,

Peter
 
I've had a look a the transformer meta. As I thought, the turns ratio is equal to the square root of the impedance ratio. With my resistance (not impedance) measurements, this gives a ratio of 1:7.6

I had a look at the Beyer transformer manual and I am now guessing that mine is a 1:10 or 1:6 ratio model and has a 200Ohm impedance.

I missed out extra model numbers on the lower part of the transformer. The full number is TR/BV35806 10 72

So, is there a suitable circuit I should use the transformer with, or should I use the original circuit?

I would like to use the preamp with (don't laugh please!) some of my ribbon mics and dynamic mics only, so I would like at least 60dB gain.

I could use the Jensen NE5534 circuit:

http://www.jensen-transformers.com/as/as017.pdf

Would that be good?

Thanks,

Roddy
 
Yeah, I know you need to use an actual audio frequency signal since it is impedance and not resistance. I was just hoping for a (very) rough value which would not take the imaginary part (inductance and capacitance) into account.

Sadly, I don't have a signal generator at home.
 
After reading through the Beyer transformer guide again and again, I now believe that I have found the correct type.

I think the transformer ratio is 1:7 and the transformer is supposed to have a "transformed impedance" of 10 KOhms.

So, should I use the Jensen 5534 circuit? Or can I use the original one?

Thanks,

Roddy
 
I would go with the Jensen circuit. The original one is not of very high performace.

Another suggestion is a schematic I posted some time ago: Shared_Gain_PreAmp_r0.gif

This one is even higher performace and you can use an easy-to-source linear pot instead of the revers log. However, this schemo needs some refinements for your app. As I'm on holidays tomorrow, someone else would need to do that.

Samuel
 
Another suggestion is a schematic I posted some time ago: Shared_Gain_PreAmp_r0.gif
That looks like a neat one.

The '5534-Studer-circuit I saw with a Beyer 1:3.16-TX used an opamp with fixed gain followed by switchable gain in the second opamp. And the recommendation to lower gain of the first one (wiring a jumper I thought) with hot mics...

So that linked lin-pot circuit is more elegant.

The original one is not of very high performace.
Woulnd't know that right now, but yes, possibly.
With a few discretes very nice circuits could be made of course.
Maybe that other posted circuit (2*NPN, 1*PNP) is better, I should still fire it up.

Regards,

Peter
 
from Roddy:
I think the transformer ratio is 1:7 and the transformer is supposed to have a "transformed impedance" of 10 KOhms.
That makes sense: 200 Ohms to 10k (= about 7*7*200).

Hmm, 1:7, wasn't that also the ratio of one of the A*P*I-input-TXs ?
You could add a 2*5*2*0 discrete opamp.
 
I am not sure about three things with the Jensen 5534 circuit.

Will I need to change any of the values of resistors / capacitors at the input because I am using a different transformer?

Will I get higher gain because I am using a 1:7 transformer? The Jensen transformer has 20*log 4.8 = 13.6dB gain. My Beyer has 20*log 7 = 16.9 dB gain. Will I get 3.3dB more gain?

I don't know if the Jensen circuit will have enough overall gain for my mics. Should I use a different circuit with more gain?

Thanks again,

Roddy
 
Will I need to change any of the values of resistors / capacitors at the input because I am using a different transformer?
Yes. The choice is not trivial, you need either manufacturers data or a lot of testing. As a first guess, I would use a single 100k resistor. For more info, try a search here for "zobel network" and "ringing".

Will I get 3.3dB more gain?
Basically yes. For a more detailed analysis you'd need to include resistive losses.

Should I use a different circuit with more gain?
Well, the one I posted would for sure be better suited to your needs. I had a few free seconds left, check this: Shared_Gain_PreAmp_r1.gif

I changed the opamps to OPA604, as the 5534 would make DC issues. These amps are a bit noisier, but it's less than 2 dB with a 150 ohm mic. As dynamic ones have higher source impedances, the difference is even smaller for them.

I added some caps as well. I suggest bipolar types, if you cannot source them, use polar ones instead. In this case, make sure that C5/C6 have the + polarity looking to the output, this will make them survive a phantom power attack.

Note that this design inverts polarity - make sure you wire the primaries of the input transformer "the wrong way around".

As shown, this design should be quite a good performer and easy to build. Nonetheless, I drew this a bit in a hurry - can someone double-check my schemo?

Samuel
 
Hi Samuel, I really appreciate you making that schematic up for me.


Like Samuel said, can anyone double-check this schematic for me?

I thought it looked fine, although I am not an expert.

If someone can confirm it is ok, then I will get on with making my preamp!

Thanks guys,

Roddy
 
Roddy, not sure you're still interested in this but looking at it two weeks later I do not see any errors in my schemo.

I stated that a 5534 will produce DC issues; this was a quick guess and unfortunately a more-or-less wrong one. Both 5534 and OPA604 should work well here. If you use a 5532, you even save one chip.

If you go with two OPA627s, you could skip C1 (and even C5/C6, if you don't need to be phantom power proof). I think that's the amp I would choose if I were to build this.

Samuel
 

Latest posts

Back
Top