DISA Mic-limiter mod. Search for advice.

[erik_DK]

I have this mic-limiter, I want to modify to work as a normal mic preamp. It is made either by DISA or as an in-house production at DR - Danish National Broadcast.

Very high quality. JS-transformers in and out (and also for PSU). Made around 1965.
Getting your hands on a manual for this one is impossible, so first I did some reverse-engineering, and made up some schematics. I think they are correct.

Turning off the limiter is easy, I just removed the FET-transistor T6. So now I have an amp with a very high gain, which must be lowered to get some head-room. Putting in a resistor in place of the FET in parallel with R23 should do the job, but I wonder if there are other, better ways to do this.

I don’t know how to calculate the gain of the transistor stages before and after the limiter circuit, and could need some advice, if it is a good idea to lower the gain of one or both of these, and how to do it.

Also I don’t quite understand the meaning of the diodes D4 and D5, but I do understand (or I guess) they have something to do with the limiters control voltage.

Would it be possible to remove all components around the limiter except C8, R17and C12, and then rewire R23 (maybe a lower value) to go directly to 0V, and then finally rewire the output transformer for normal 1:1 or maybe 2:1 ratio? A ratio of 2:1 would lower the over-all gain of the pre amp and also lower the output impedance, but that is no use, if I am out of head-room in the stage(s) before.

I will post som pictures of the amp later. Almost same looks as the V72 of that time.

 

[gyraf]

I uploaded the official version here: DR - Danmarks Radio

The ratio of R7/R8 (1K/22R) sets gain in the first stage - perhaps you would want e.g. three steps here. Better keep total resistance approximately the same.

Why not keep the limiter/compressor? It sounds nice, iirc - have one somewhere around here. You could perhaps add threshold control, attenuator between D4/D5 and base of T4? C10 is your timing capacitor. A pot (100K?) in series with D2 would be variable attack time, a 1M pot parallel with the timing cap would then be release time..


(..you're in Århus? You should stop by my office in Jægergårdsgade one day? )

/Jakob E.

edit: threshold would probably be better placed before the two D4/D5 diodes - above would more be like a ratio control. Attenuate with a double e.g. 2x10K linear pot before diodes, referring to the related transformer centertap.

 

[erik_DK]

Hej Jakob

I will stop by in the near future. It was me who visited you and lend you those NTP-manuals some weeks ago.

Kind regards Erik

 

[gyraf]

Aah - good! Must admit that I haven't had time to scan just yet, they're still on my desk here..

/Jakob E.

 

[My3gger]

Your writing in tech docs explains successful Denmark's pro audio scene, i find it very interesting. In some ways ex Yugoslavia similar approach, the problem was lack of high quality materials and private money to buy it. Only army could afford anything..
Is the government still helping private pro audio sector or you are on your own?

 

[gyraf]

Yes, DR was a fantastic deal for the local pro audio business - unfortunately I joined just too late to take part in all that: The big privatizations of the 90'es left all these grand old monopolys with very little buying power - so we're basically at our own now..

/Jakob E.

 

[erik_DK]

Aah - good! Must admit that I haven't had time to scan just yet, they're still on my desk here..

/Jakob E.

Thats ok. I will stop by anyway. See you Jakob.

/Erik

 

[RuudNL]

Are C5 and C6 intended to restrict the frequency response?

 

[erik_DK]

Are C5 and C6 intended to restrict the frequency response?

You mean C5 rolling off the low frequencies and C6 filtering out some of the highs? I had that thought, when I made the schematic. I will try to make a measurement, but I don't have all the equipment (and to some degree knowledge) to do that. Listening to the amp is not enough I know, but to me it does not sound restricted.

 

[erik_DK]

And here are some pictures:

 

[erik_DK]

After working on other projects I recently returned to this modification. In the meantime I got tree more of these mic-limiters, so it has been possible to do some A-B testing in the process.

I also have two more modern - or should I say "not so old" - versions in Danner cassettes. Exactly the same schematic just without the power section.

This is what I have done so far:

• Removed all components that had something to do with the limiter
• Rewired the output transformer - primary now in parallel.
• Removed C6 and C8 and made C5 47uF to get a broader frequency response
• Removed R9 so that all signal damping is now done with the potentiometer.
• Changed R30 to 33R to lower the gain of the second stage
• Removed the on-board power transformer, that caused a little hum, so the modules will now run from an external PSU
• In one of the modules I had to change T1, so I put in a BC109 just to see what happened. Works fine.
• Changed the primary of the input transformer to parallel to get some "free" gain. It did not work as expected, so I changed it back. My guess is that the input impedance became to low??
• Removed L1 and C22 - some of the modules had them and others didn´t. And there is no audible difference. As far as I can calculate, it was a 29 kHz lo-pass filter. Could it be in there to avoid some high frequency instability?
• Made a complete recap - some of the almost 60 year old electrolytics were leaking.

The modules now work very well, giving a nice warm sound and plenty of gain for most microphones.

Future experiments could be replacing R30 and R31 with a stepped attenuator to get variable gain in the second stage. Or maybe even an attenuator with 2 decks so either R10 (the potentiometer) or the ratio of R7 and R8 are also included in the control.

You can se the details in the attached files.

Please share your thoughts.

/Erik