OA10 Mic Preamp

[isophase]

Hi all,
I'm thinking of building the OA10 mic preamp and I have a few questions. Does this amp really have 60db of gain with a single opamp and 1:2 ratio input transformer? where would be the best place to put an attenuator, right after the input xfmr at the output or in the loop? if i add an output transformer 1:1 ratio (600:600)  to balance the output, i assume i can get rid of the output cap, anything wrong with that?
thankyou!

 

[gyraf]

You'll want to keep the output cap, as the circuit operates from single-supply +24V (and OA10 output will be somewhere around ½ that).

Jakob E.

 

[isophase]

thanks Jacob, ok so assuming the load impedance is 10Kohm (input impedance of the line amp following the micpre) 10uF cap gives -3dB @ 2hz, i will try out different values and see what it sounds like.
Attached is the OA10 micpre schematics with added attenuator pot at the input and output transformer.
i have no idea how to calculate the correct value for the attenuator, i'm sure a 100K log will work but i think this would mess with the secondary widing impedance reflected to the mic output, mabe a T-pad attenuator will work better with constant impedance?


 

 

[abbey road d enfer]

thanks Jacob, ok so assuming the load impedance is 10Kohm (input impedance of the line amp following the micpre) 10uF cap gives -3dB @ 2hz, i will try out different values and see what it sounds like.
Attached is the OA10 micpre schematics with added attenuator pot at the input and output transformer.
i have no idea how to calculate the correct value for the attenuator, i'm sure a 100K log will work but i think this would mess with the secondary widing impedance reflected to the mic output, mabe a T-pad attenuator will work better with constant impedance?

The pot is wired wrong. As it is, it loads the secondary for attenuation, as you feared, creates havoc on the input impedance.
You may wire the pot correctly (secondary to top of pot, wiper to opamp input. This is not the best scheme in terms of noise, but it works. Better is making NFB variable. You can replace the 22k NFB resistor with a 22k Log pot wired as a rheostat (variable res). You may want to have  a padding resistor in series of about 600r, that would give variable gain of 23dB to 53dB (excluding xfmr gain). There should be no stability problem.
Regarding the value of the output cap, the inductance of the xfmr's primary will determine the LF frequency response. You will certainly need a much higher value. I can't say more, not knowing what xfmr you intend to use. A standard 600 ohms output xfmr, combined with a 47-470uF needed for proper LF extension will have a very pronounced VLF resonance that needs to be tamed by loading the secondary with a resistor..

 

[isophase]

Regarding the value of the output cap, the inductance of the xfmr's primary will determine the LF frequency response. You will certainly need a much higher value. I can't say more, not knowing what xfmr you intend to use. A standard 600 ohms output xfmr, combined with a 47-470uF needed for proper LF extension will have a very pronounced VLF resonance that needs to be tamed by loading the secondary with a resistor..

the output transformer i would like to use is a toroid VT24499 that came from neve module. i just know it has a 600:600 1 to 1 ratio and windings have a measured DC resistance of 10,7ohms and 6,2 ohms respectively. Don't know about the inductance, when you say 47-470uF is it a range? or you mean a 47ohm resistor and a 470uF cap? sorry for my total ignorance.
if i choose 470uf coupling cap so that the resonance is way below the bandpass (if i understand correctly) then wire a variable resistor on the secondary winding of the output transformer (say 500K lin?) set my multimeter to read RMS voltage on the secondary winding and sweep low frequencies with oscillator to the primary in series with the cap to read maximum voltage on the multimeter, and then use the pot to determine the resistance value that gives minimum reading? will that work?
thanks a lot for your help guys! learning learning ...
its great i have almost all the parts to build four preamps all from spares and surplus from other projects. the 4 DOAs came from 169eq project and the transformers i had for a long time waiting for the right project for them, i think thats it i realise i even have a spare mnats PSU board with 0-24V and phantom power PERFECT!

 

[isophase]

will redraw the shematic with corrections and post it soon.

 

[rafafredd]

Hey, forget this original single supply neumann schematics. Go for an API like, standard opamp mic pre with dual supply, get rid of the output cap, as you have beefy output transformers, or make it at least 100uF. You will also want to make sure you use a proper C-type anti-log gain control in the feedback loop. The 1x12 lorlins are specially suitable for this. No for the in/out attenuators in this circuit. Use good quality and big cap for the feedback loop to ground, after the gain control, I mean, and it will sound wonderfull. I like Nichicon, Panasonic, Elna and Rubycon.

You could look at the Jensen/JH 990 PDFs. It was at the John Hardy web site last time I checked years ago. 990 application data or something like this. You could try and look for the original Jensen papers also. I have it somewhere in my multi-gigabyte-20-or-so-HDDs DIY archive. It is somewhere in the past. Maybe someone could like to it online? There is a nice circuit there that you can use. I like the preamp with the input "bias" control to get rid of the DC on the output or make it to a minimum mV value. Then no cap or servo is needed on the outputs. Pretty simple circuit, I use it all the time. Sounds wonderful.

Nice transformers. What are those? I see later neve/amek pumpkin toroids at the output. Am I right? Should be awesome. And thse inputs are???

 

[0dbfs]

These things can have a lot of gain. My API's often need an inline-mic-pad at the input. I think mine are -20db but have doubled them up too on occasion.
Ends up being an extra switch but a defeatable/variable pad before the ip-xformer can be quite useful.
Some people may say that an output trim or pad is helpful so you can run the pre hot and trim back the output... IMO that may be fine on occasion but generally don't really want that on 99% of the signals. Output trim pot +/- 5 or 10dB would be great though to dial in your levels exact if you have a switched pad on the input and switched sensitivity (nfb) control...

Won't mention phantom... oops!

Cheers,
jb

 

[rafafredd]

Yes, single 20dB pas on mic inputs is a must. I was talking about having no gain control and using a multi-turn switch for input attenuator as the only way to control "level" in that circuit. But if you already have a gain control, an input "PAD" is always welcome.

 

[Harpo]

..You will also want to make sure you use a proper C-type anti-log gain control in the feedback loop...

..neg.log taper in the shunt arm, or a log taper for feedback. I'd prefer the 1st so gain drops to maybe unity instead of full boost if the pot wiper should fail. YMMV.

 

[rafafredd]

Yes. Neg log it is. Not anti-log.

I always use a 1x12 non-shorting switch in there, for proper matched gain stereo recording. Or very good low tolerance plastic ones, like OMEG offers.

 

[isophase]

thanks for your comments guys, understanding lots of things.
i will post an update soon so you can make more critics
tempted to try out with single 24V rail for two reasons 1st: i already have otherwise useless and already stuffed PSU board putting out correct voltage and with phantom power. 2nd: this is a neumann circuit so it must be right ;D
input transformers are belclere TF10015. apparently the OA10 circuit has lots of gain i will consider the 20db input pad i like switches as a gain control, will find out what to do after i do some more searching reading learning...
i suppose bipolar +/-24V will have more headroom, any other advantages? is it a big difference compared to single 24V supply? 

 

[gyraf]

All the old Neumann desks ran (and runs) on single +24v rails.

If you want/need more output level and/or headroom, use a 1:2 (150:600 Ohms) output transformer - this will give you the same as moving to +/-24V, and will probably be better for your opamp.

Jakob E.

 

[rafafredd]

For gain control, put a 20k pot in series with that 43R resistor.

The main reason for the dual supply is to get rid of the two large electrolytics in the circuit. If you go the single supply way, make sure to use good quality ones there. I use output lytic cap with at least 470uF for a 1:2 (150:600) output transformer. Sometimes 1000uF, if available.

 

[dmp]

So, can the OA10s go from 0v / +24v supply to -24v / +24v safely?
I have some OA10s that I'd like to try out in my 990 preamp if so.

The servo 990 preamp circuit (aka JH M1) is in this pdf:
www.johnhardyco.com/pdf/990.pdf

 

[rafafredd]

Hey isophase! Right there it is. Circuit on figure 1 is exactly what I use, with the optional bias compensation network to null DC on the output. Seems enough, so I never use a servo.

 

[isophase]

thanks rafafredd,
i will have a look/study the circuit and try to come up with the best solution/compromise using the stuff i already have on hand to build this the right way.
thanks to all for your comments and input! can't wait to start this project!

 

[rafafredd]

So, can the OA10s go from 0v / +24v supply to -24v / +24v safely?

I am not sure. I would have to check my files at home. But if I recall, yes, it can take +-24v safely. But do not try without making sure.

 

[[silent:arts]]

So, can the OA10s go from 0v / +24v supply to -24v / +24v safely?

no. the power supply voltage range is +/- 2,5 to +/-22 V
speaking for the original, taken out of the original datasheet.

might be different with the clones