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

[Wordsushi]

@JeffKall, you're welcome!

Okay, so I have a couple of ORS builds that are nearly identical with 797 capsules, but one has a 3u GZT 87 transformer and the other has the $18 Aliexpress T-13 in it, and the one with the Ali transformer sounds closer to my 87. But, what stuck out, even when the one with the 3u in it had a different capsule was what I thought was a slight notch right about where the upper mids meet the high end. Changing the capsule didn't change the notch so I tried a couple of tests.

First was a 1k tone, fed from my phone, recorded in my booth.

(ORS 87 with GZT-87 transformer (797 Capsule))


(ORS 87 with $18 AliExpress T-13 transformer (797 Capsule))


Notice how the Ali T-13 seems to go right up to 20k.

Then, WHITE NOISE test.

3u GZT-87

Ali T-13


And there's the notch at 6k that I had thought I had heard.
Nearly identical builds in same HL-77 bodies, with 2n38a9's from the same batch of FETs I bought from Small Dog years ago.

Now, I'm not ruling out that it may not be the transformer, but since that notch was there with the previous capsule (3u K87), the transformer is my best guess.

And despite my initial disdain for the Ali T-13, I have definitely changed my mind about it given the results I've gotten from my two separate builds with them.

EDIT: Additional thought. I have 6 fully functional ORS 87 builds. I only have one with the GZT transformer and it's the only one that has that 6k notch in the sound.

 

[OneRoomStudio]

1. Microphone Capsule (MK1):
   • The microphone capsule is represented by MK1. It is a condenser type that requires a polarization voltage to operate.
     ...so far so good...
2. Bias Resistors (R8, R9):
   • R8 (1MΩ) and R9 (7.5kΩ) provide the necessary biasing for the microphone capsule. The large value of R8 ensures a high impedance necessary for the condenser microphone.
     Well, strong start anyway. R9 is not a "bias resistor," as it comes after C5, which, as it mentioned in the next step, blocks DC.
3. Coupling Capacitors (C5, C6):
   • C5 (33nF) blocks DC voltage while allowing audio signals to pass.
   • C6 (120pF) is likely for high-frequency filtering, providing stability and preventing oscillations.
     Calling these "coupling capacitors" is dubious, but it gets the purposes correct (although misses fully describing the function of C5)
4. Field-Effect Transistor (Q1, 2N3819):
   • Q1 is a JFET (2N3819) that acts as an impedance converter, converting the high impedance of the condenser capsule to a lower impedance more suitable for further amplification.
     This is true.
5. Bias Network for Q1 (R7, R10, C8):
   • R7 (1GΩ) sets the gate bias for the JFET.
   • R10 (10kΩ) and C8 (20µF) are part of the source resistor and bypass capacitor network, stabilizing the operating point of the JFET.
     This is also true.
6. Transformer (T1):
   • T1 is a step-down transformer with a ratio of 9.5:1. This matches the high output impedance of the microphone amplifier to the lower input impedance of professional audio equipment.
   • The primary side of the transformer is connected to the drain of Q1, and the secondary side provides a balanced output.
     Also true.
7. Output Coupling and Filtering (C7, C10, C12, C13):
   • C7 (1µF) and C10 (47µF) block any DC component from the output signal.
     While C7 is an output coupling capacitor, C10 is not. C10 is a filtering capacitor for the power going to the JFET.
   • C12 (10µF) and C13 (10nF) further filter the signal to remove noise and unwanted frequencies.
     This is correct.
8. Feedback Network (R13, R15, C9, C11):
   • R13 (330kΩ), R15 (300kΩ), C9 (470nF), and C11 (4.7µF) form a feedback network that stabilizes the gain and frequency response of the circuit.
     Nope. It did get 1 out of 4 though, I guess. C9 is part of the feedback loop. R13 and R15, while *technically* forming a slight feedback path are actually there to supply the polarization voltage to the capsule from phantom power, and C11 is a filter capacitor on the power.
9. Load Resistors (R12, R14, R16, R18, R19):
   • R12 (47kΩ) sets the load for the JFET.
     Yup
   • R14 (47kΩ), R16 (56kΩ), and R18/R19 (2.2kΩ each) set the correct loading and impedance matching for the transformer and output stage.
     Nope again. R18/19 are there to deliver phantom power from the balanced feed. R16 and R14 are dropping resistors used with the zener to provide voltage to the JFET.
10. Zener Diode (GR1, 1N4744A):
    • GR1 (1N4744A) is a Zener diode used for voltage regulation, ensuring a stable operation voltage for the circuit components.
      This is true.
11. XLR Output (J1):
    • The XLR connector provides a balanced output, which is standard for professional audio equipment, offering better noise immunity and signal integrity.
      This is also true.

If this were a quiz, it would get a "C." Maybe a C-

 

[micolas]

If this were a quiz, it would get a "C." Maybe a C-

Yes, I had fun
Quite a lot of aberrations gathered from the forums.

 

[Wordsushi]

If this were a quiz, it would get a "C." Maybe a C-

Nice analysis.

 

[Wordsushi]

Today's experiment: Trying an Apex 460 transformer, which @RuudNL confirmed was 10:1.
I can report back that:
a) It works great in this circuit
b) It sounds good.

Wish I had more than just one of these, I'd use them for this project.

If anyone has a stock transformer pulled from an Apex 460. It connects thusly.
Red = P-
White = P+
Black = S-
Green = S+

 

[elskardio]

b) It sounds good.

I’m curious how it compares to the GZT?
I have 2 apex 460 transformers sitting in a box…

 

[kingkorg]

@JeffKall, you're welcome!

Okay, so I have a couple of ORS builds that are nearly identical with 797 capsules, but one has a 3u GZT 87 transformer and the other has the $18 Aliexpress T-13 in it, and the one with the Ali transformer sounds closer to my 87. But, what stuck out, even when the one with the 3u in it had a different capsule was what I thought was a slight notch right about where the upper mids meet the high end. Changing the capsule didn't change the notch so I tried a couple of tests.

First was a 1k tone, fed from my phone, recorded in my booth.

(ORS 87 with GZT-87 transformer (797 Capsule))
View attachment 129060

(ORS 87 with $18 AliExpress T-13 transformer (797 Capsule))
View attachment 129061

Notice how the Ali T-13 seems to go right up to 20k.

Then, WHITE NOISE test.

3u GZT-87View attachment 129062

Ali T-13
View attachment 129064

And there's the notch at 6k that I had thought I had heard.
Nearly identical builds in same HL-77 bodies, with 2n38a9's from the same batch of FETs I bought from Small Dog years ago.

Now, I'm not ruling out that it may not be the transformer, but since that notch was there with the previous capsule (3u K87), the transformer is my best guess.

And despite my initial disdain for the Ali T-13, I have definitely changed my mind about it given the results I've gotten from my two separate builds with them.

EDIT: Additional thought. I have 6 fully functional ORS 87 builds. I only have one with the GZT transformer and it's the only one that has that 6k notch in the sound.

Sorry to be a party pooper, but in this kind of tests most of the THD comes from the speaker. That's why we inject the signal directly to the circuit, avoiding speaker and capsule as variable for THD measurements.

 

[micolas]

@Wordsushi
In the YT test of the various ORS87 variants, the green microphone sounds best to my ears ("Steve"...I think Stevie Wonder would sound great on it and he would love it)
The body and headbasket with the correct shape and size make an important positive contribution to the overall sound equation.

 

[elskardio]

On the ORS schematic, Where do you guys inject your signal to bias the fet?

 

[micolas]

On the ORS schematic, Where do you guys inject your signal to bias the fet?

 

[Wordsushi]

I’m curious how it compares to the GZT?
I have 2 apex 460 transformers sitting in a box…

Haven't really done a proper evaluation yet and it's in an ORS87 that has different parts than any other one I have, but at first listen, it sounds larger than the 3u. May have to retune C5 on this one. May be more like the UTM than the 3u.

@micolas Steve seems to be a popular fellow in this group. This is actually the same mic from the above posted transformer comparison and the mic that had the 6k notch. Perhaps that notch is working in my favor.

@kingkorg Good point. I hadn't even thought of that. Thank you for brining up that point. At the very least it gives me a comparison between the two, using the same compromised source. I was primarily interested in seeing how they differ and now I can say for certain they are not the same transformer.

 

[anvl]

My first prototype (using Ali capsule $20 aprox + Ali T-13 transformer). I'll try to upload some sound samples.
I tried to follow the complete schem (not really the ORS87). 150nF for low end, 220pF for high end.

 

[micolas]

Haven't really done a proper evaluation yet and it's in an ORS87 that has different parts than any other one I have, but at first listen, it sounds larger than the 3u. May have to retune C5 on this one. May be more like the UTM than the 3u.

@micolas Steve seems to be a popular fellow in this group. This is actually the same mic from the above posted transformer comparison and the mic that had the 6k notch. Perhaps that notch is working in my favor.

Yes, it is possible that the pit at 6kHz removes some unpleasant resonances, making the voice smoother, velvety.
When you built the circuit Fuchs for the first time you were very excited, delighted with the sound obtained through the gzt-87 transformer, although it has been quite a while I remember perfectly.
Like me and many others. I still think it's an exceptional transformer, sorry Guosheng doesn't ship to my country.
I'm curious if you could test the Neutrik nte10/3 and UTM0857 in the Steve microphone instead of the gzt-87.

 

[Wordsushi]

My first prototype (using Ali capsule $20 aprox + Ali T-13 transformer). I'll try to upload some sound samples.
I tried to follow the complete schem (not really the ORS87). 150nF for low end, 220pF for high end.

View attachment 129157

That is fantastic! Bravo!

@micolas, Steve is done, as far as I'm concerned. I'm not planning on tinkering with Steve. The UTM0587 I have is also in a completed mic, but I do want to build a "premium" version of the ORS87 at some point with a UTM and Arienne's capsule.

The 3u Transformer in my "Holy Fuchs" mic sounds great, but the Fuchs build with the UTM in it sounds closer to the real deal, so now I have to go back at some point and listen to the difference between them and see if that same notch is there.

The 6k notch hits right at a spot where I can really hear it with my voice, so it may indeed be helping. To me, it's like looking at a fresh tub of ice cream with one big scoop taken out, but that notch may make the lower mid more noticeable, thus making it seem warmer.

 

[elskardio]

I made a small PCB for this project. The mic works... but I can't bias the fet with my scope!

I'm feeding the tone thru a 47pF cap directly at the Gate. My scope is connected to the Drain. I have nothing on the scope...

Any hint?

Here's my schematic.

 

[Khron]

I made a small PCB for this project. The mic works... but I can't bias the fet with my scope!

I'm feeding the tone thru a 47pF cap directly at the Gate. My scope is connected to the Drain. I have nothing on the scope...

Any hint?

Here's my schematic.

View attachment 129167

DC voltages all good? Where are you powering the mic from? JFET still in one piece?

 

[micolas]

@micolas, Steve is done, as far as I'm concerned. I'm not planning on tinkering with Steve. The UTM0587 I have is also in a completed mic, but I do want to build a "premium" version of the ORS87 at some point with a UTM and Arienne's capsule.

The 3u Transformer in my "Holy Fuchs" mic sounds great, but the Fuchs build with the UTM in it sounds closer to the real deal, so now I have to go back at some point and listen to the difference between them and see if that same notch is there.

The 6k notch hits right at a spot where I can really hear it with my voice, so it may indeed be helping. To me, it's like looking at a fresh tub of ice cream with one big scoop taken out, but that notch may make the lower mid more noticeable, thus making it seem warmer.

Thank you for the detailed information, very useful.
As for the microphones you built in the smaller bodies, (no. 3,4 and 5) they sound a little thinner, without bass.
I suggest you try to replace the capacitor C7 1uF with an electrolytic one of 2.2uF and C5 33nF with one of 39nF.
With these values I got a fuller, rounder sound with the Neutrik transformer, capsule 797, small body.

 

[Wordsushi]

Didn't really pick it up at my desk at first listen yesterday with ambient noise in the background, but now that I've had a chance to get it in the booth, it seems swapping in the Apex 460 transformer really dulled everything between 2-6k quite a bit vs the Neutrik that was in there. Some nice warmth in the low end tickled my ear, but this one isn't going to work out, sorry to say....
Next...

 

[elskardio]

DC voltages all good? Where are you powering the mic from? JFET still in one piece?

I’m powering the mic with a small Motu M4.
I can adjust to get 11.5v at the drain… but can’t seem to get a signal to my scope.
Pretty strange! I built multiple U87 and never had this problem before.

 

[Khron]

I’m powering the mic with a small Motu M4.
I can adjust to get 11.5v at the drain… but can’t seem to get a signal to my scope.
Pretty strange! I built multiple U87 and never had this problem before.

Scope as in a physical oscilloscope? Why not just use a scope plugin in your DAW?