OPEN SOURCE DIY Mic Project - ORS 87 - Stripped Down u87

[pasarski]

Hello, I forgot to order the LSK189-TO-92-3L-BK . Can I also take another one or will there be problems? I still have this one here bc239c https://www.mos-electronic-shop.de/239c-p-1791.html
BC 172
https://www.radiomuseum.org/tubes/tube_bc172.html?language_id=1

Those are BJTs. You definitely need a JFET.

 

[mic test]

OK, thanks

 

[cello4u]

Decided to not wait on directions and loaded up the KSA 1.1 board. Still unclear on the Biasing, but it appears to be a mic. This was my fifth try.

 

[Khron]

Still unclear on the Biasing

Same as every other U87 clone.

Apply 1kHz signal between "Cal" and ground, monitor the output signal on an oscilloscope and/or spectrum analyzer, turn bias for minimum distortion / harmonics, increase input signal, rinse & repeat.

 

[Voyager10]

If I may pick a nit, there's nothing to be gained from increasing the signal level and repeating. The bias level which minimises distortion is the same, over a wide range of levels. This post has measurements made with the same ORS87 PCB which illustrates the point.

If you have any half-way decent audio interface, Room EQ Wizard is great free software which gives distortion readings directly, and is much more sensitive than a 'scope. I often use an iPhone with a signal generator app as a signal source, it gives output in the 0-400mV range which is ideal for this application.

If you just want something roughly right, set the voltage on Q1's drain to 10-12V. All the FETs I've measured have an optimum somewhere near this.

 

[Khron]

If you have any half-way decent audio interface, Room EQ Wizard is great free software which gives distortion readings directly, and is much more sensitive than a 'scope.

By "oscilloscope and/or spectrum analyzer" i meant / implied that their virtual versions should work just as well, be they within REW, or plugins in one's DAW of choice.

If you just want something roughly right, set the voltage on Q1's drain to 10-12V. All the FETs I've measured have an optimum somewhere near this.

That presumes a 24V supply; not necessarily the case in "my" ORS87-plus.

 

[Voyager10]

I've been using a 33V Zener, as per the schematic attached to the PCBWay project - 10-12V corresponds to a drain current of 0.4mA or so.

 

[circuitfreak]

Well, after my 4th or 5th attempt, using the KSA board, I now have a working microphone. I know I'm gonna get kicked for this, but, I cannot seem to understand the directions for biasing...I can see on the scheme that signal injection goes between R15 and R9-but i dont see any connection on the board. Having said that, the mic seems to be very low self noise. The sound quality seems to be good, using a BM800 donor body and capsule. However, I have to crank up the input gain to get decent levels-altho, that may be due to the trashy capsule. If any of you would, please take a listen to the attached sound file for critique. Thanks!
View attachment ORS:KSA.wav

 

[circuitfreak]

*edit-R15 and R9-I dont see any connection between those two points on the board, using continuity on my dmm...

 

[pasarski]

Well, after my 4th or 5th attempt, using the KSA board, I now have a working microphone. I know I'm gonna get kicked for this, but, I cannot seem to understand the directions for biasing...I can see on the scheme that signal injection goes between R15 and R9-but i dont see any connection on the board. Having said that, the mic seems to be very low self noise. The sound quality seems to be good, using a BM800 donor body and capsule. However, I have to crank up the input gain to get decent levels-altho, that may be due to the trashy capsule. If any of you would, please take a listen to the attached sound file for critique. Thanks!
View attachment 137340

Sounds splendid Like a U87 actually. I don't know the board, but just stick it in there, stuck it, solder it or whatever is easiest?

U87 is not high gain . S/N ratio sounds alright, can't hear hiss. Sum brum though. Maybe not the mic but something brumming at your place?

 

[Khron]

using the KSA board

Which version, "1.0" (first), or "1.1" (more recent minor update)?

 

[circuitfreak]

Sounds splendid Like a U87 actually. I don't know the board, but just stick it in there, stuck it, solder it or whatever is easiest?

U87 is not high gain . S/N ratio sounds alright, can't hear hiss. Sum brum though. Maybe not the mic but something brumming at your place?

Thank you. Yeah, I dont have the best place for sound checks, thats my AC...

 

[circuitfreak]

Which version, "1.0" (first), or "1.1" (more recent minor update)?

I don't see anything on the board that specifies, but, there is a project number W536781AS3P2, from PCBWay

 

[Khron]

I don't see anything on the board that specifies, but, there is a project number W536781AS3P2, from PCBWay

You can compare how your board versus the render currently visible on the project page, surely..?

Or is there a "CAL" wire-pad on the board you have? A photo might help...

 

[circuitfreak]

It's the 1.0

 

[Khron]

It's the 1.0

There we go then - the "calibration" input was not yet physically implemented there. But it can be easily added - desolder the ground end of that 10k resistor (R15 on that board, i guess), and add a 560 or 620 ohm resistor in series with it. The joint between those two is now the "cal" input.

The above-mentioned procedure applies:

Same as every other U87 clone.

Apply 1kHz signal between "Cal" and ground, monitor the output signal on an oscilloscope and/or spectrum analyzer, turn bias for minimum distortion / harmonics, increase input signal, rinse & repeat.

Or you can take a bit of a shortcut too:

If you just want something roughly right, set the voltage on Q1's drain to 10-12V. All the FETs I've measured have an optimum somewhere near this.

 

[circuitfreak]

There we go then - the "calibration" input was not yet physically implemented there. But it can be easily added - desolder the ground end of that 10k resistor (R15 on that board, i guess), and add a 560 or 620 ohm resistor in series with it. The joint between those two is now the "cal" input.

The above-mentioned procedure applies:


Or you can take a bit of a shortcut too:

Thank you! That’s on me for not realizing there had been an update. My first clue was the cal pad is on the updated schematic, which, I now see, means an updated board…
Thanks much for the clarification!

 

[elskardio]

I'm a bit puzzled...
11 years ago, when I built my first pair of U87 mics, everyone was all about symmetrical clipping as the biasing method. But now, it seems like everyone is biasing for minimum THD. What's up with that sudden change in approach?

 

[pasarski]

I'm a bit puzzled...
11 years ago, when I built my first pair of U87 mics, everyone was all about symmetrical clipping as the biasing method. But now, it seems like everyone is biasing for minimum THD. What's up with that sudden change in approach?

I guess a certain mic tech that was using the symmetrical clipping method fell out of favor or something . I don’t see nothing wrong in either approach. It’s a matter of taste and specific requirements imo. And what sounds best in a given circuit.

It’s even more a matter of taste in tube circuits I think. Might be baloney, but I think I hear a compressor doing different things to the signal depending on the bias. There’s a certain mic manufacturer that me and my friends think never sounds okay when compressed hard. Don’t know if it’s a biasing thing or not, but I remember someone suggesting it some years ago.

I also remember Oliver Archut writing that biasing a tube mic is mostly about tone, but I can’t find the post anymore.

Sorry for bringing up tubes, maybe a bit OT.

 

[OneRoomStudio]

The goal of both methods is to bias for low distortion. Symmetrical clipping (in theory) means that the circuit is tuned to have maximum headroom before clipping on both positive and negative swings.

I think 11(or more) years ago, it was less common to have accurate spectral analysis in plugins, and hardware spectral analyzers are crazy expensive. A lot of people have oscilloscopes though, so looking for symmetrical clipping made sense. My problem with that approach has always been that it's pretty hard to judge where "symmetrical" is. You're trying to decide which curve is less "curvy" than the other - not very precise.

Using cheap/free spectral analyzer plugins, it's easy to adjust for the lowest 2nd harmonic, and much more accurate.

There's an entirely different topic of whether these mic circuit sound "best" at the lowest THD setting, or if there is some "magic" amount of THD you want to leave in. I'm not going to touch that topic. All I'll say is that I like to start with the drain voltage at around ½ of the voltage before the drain resistor (at the junction of R12 and R14 in my original schematic), and adjust for lowest 2nd harmonic from there. YMMV.