Converting a D-87 to ORS87

[joulupukki]

I've got a D-87 (Dany Bouchard) that I've built into an Aurycle A460 mic body. I originally had it built using an Arienne Audio K87 capsule but I've since swapped it out for a JLI TSC-2 capsule (because I used the Arienne for my Fuchs U87a). With both cases I've been underwhelmed by the lack of sensitivity of the D-87 and would like to change it if possible so it's closer to what the Fuchs U87a would be (and my other condenser mics).

As is, the -10dB pad and low cut filter switches are functional.

I'm currently using a 33V Zener with a 47K at R17.

It's been suggested that the ORS87 would yield better sensitivity so I wouldn't have to turn up my preamp so high. I'm starting this thread so I don't pollute the D87 build thread anymore. Has anyone else done this conversion with satisfying results? As pointed out by @micolas I think it's just going to be a matter of tracing both circuits and removing what no longer belongs on my existing circuit board. Right?

 

[Khron]

You've been warned you'd be giving up the omni & fig-8 patterns, right?

 

[joulupukki]

You've been warned you'd be giving up the omni & fig-8 patterns, right?

When I switched from the AA K87 to the JLI TSC-2 I already gave up one of those. I've never used omni or fig-8 with this mic anyway. I've got a modded AKG P420 and the Fuchs U87a for that purpose if needed.

These are the U87 schematics I've been referencing with the D-87 build:

http://recordinghacks.com/images/mic_extras/neumann/U87-schematic-1972.png
http://recordinghacks.com/images/mic_extras/neumann/U87-schematic.png

 

[joulupukki]

If I’m looking at the ORS87 schematic right, I think if I were to make the following changes, that’d change it to where it’d be in the right territory. Right?

Is the main reason for the increase in sensitivity the fact that C4/470pF is removed in the ORS87?

 

[Khron]

I think if I were to make the following changes, that’d change it to where it’d be in the right territory. Right

If you replaced C1 with a jumper, you'd hard-wire the mic to omni.

C4 can indeed be bypassed; might just be simpler to just move the capsule wire from the turret where C4/R6 were joined, to where R7 & the JFET gate are joined.

 

[micolas]

Is the main reason for the increase in sensitivity the fact that C4/470pF is removed in the ORS87?
View attachment 144011

Not.
C4=470pF is just a coupling capacitor, it can only affect the passband (low frequencies) but it is big enough for the sound of the U87.
Bypassing C4 causes the microphone phase to be reversed so you will need to reverse the transformer primary or secondary wires or reverse pins 2 and 3 on the XLR output.

C3=10pf greatly reduces the sensitivity of U87i.So you have to remove it from the circuit.
I also think that it is possible to affect the shape of the deemphasis curve.

Edit:
Some have replaced C4 with 1000pF and claim that the microphone sounds with more bass. I didn't hear any differences

 

[micolas]

If you replaced C1 with a jumper, you'd hard-wire the mic to omni.

C4 can indeed be bypassed; might just be simpler to just move the capsule wire from the turret where C4/R6 were joined, to where R7 & the JFET gate are joined.

Hey @joulupukki

@Khron is absolutely right. You can very easily get an ORS87 that only lacks
figure 8.

 

[joulupukki]

Ah, I see now how the back diaphragm would be connected by jumpering C1. And, the JLI TSC-2, if I’m remembering correctly, has a shared backplate wire.

So doing something like this would be more correct for the ORS87 wiring?

Not.
C4=470pF is just a coupling capacitor, it can only affect the passband (low frequencies) but it is big enough for the sound of the U87.
Bypassing C4 causes the microphone phase to be reversed so you will need to reverse the transformer primary or secondary wires or reverse pins 2 and 3 on the XLR output.

Ah, so I could just leave it in.

C3=10pf greatly reduces the sensitivity of U87i.So you have to remove it from the circuit.
I also think that it is possible to affect the shape of the deemphasis curve.

If the basic ORS87 uses 10pF for C3 in the circuit and it has greater sensitivity than a normal U87/D87, what main difference in the ORS87 contributes to it being more sensitive? If there’s some other tweak I could make to the D-87 circuit that bumps the sensitivity up without all the other changes, I’d be game to try that. I mean, I could take out C10 and just see how that sounds or potentially lower it to 1/2.2pF.

@Khron is absolutely right. You can very easily get an ORS87 that only lacks
figure 8.

Because of the shared backplate on the JLI TSC-2 I think that may be what I already have.

 

[micolas]

View attachment 144026

Ah, so I could just leave it in.

If the basic ORS87 uses 10pF for C3 in the circuit and it has greater sensitivity than a normal U87/D87, what main difference in the ORS87 contributes to it being more sensitive?

In ORS87, the 10pF capacitor is not installed at all.

The scheme in blue was just a computer simulation. I posted it just to see the values with the 33v Zener.

The more realistic schematic is the one in the picture here, but it was originally made with a 24v Zener.
***
So we have to put:
Zener GR1 = 33v
R16 less than 47Kohm, it is tested
(a Zener to stabilize effectively needs to be applied a few extra volts)

Don't use that capacitor (10pF)

 

[joulupukki]

Ah, that makes more sense. Hopefully I'll get time to try it out today.

 

[micolas]

Ah, that makes more sense. Hopefully I'll get time to try it out today.

I finally overcame the language barriers

 

[jp8]

Quickly simulated the ORS87 with and without 10pF feedback capacitor and a 9.5:1 transformer with a 3k load.

JFET is BF256B from LTspice lib
With 10pF: Gain = -8.2 dB @ 1kHz
Without 10pF: Gain = 6.3 dB @ 1 kHz.

A major difference....

A feedback capacitor will definitely affect frequency response, as explained in my KM84 description. More LF and RF roll-off with lower values. And distortion and max SPL are affected too, of course.

 

[joulupukki]

Quickly simulated the ORS87 with and without 10pF feedback capacitor and a 9.5:1 transformer with a 3k load.

JFET is BF256B from LTspice lib
With 10pF: Gain = -8.2 dB @ 1kHz
Without 10pF: Gain = 6.3 dB @ 1 kHz.

A major difference....

A feedback capacitor will definitely affect frequency response, as explained in my KM84 description. More LF and RF roll-off with lower values. And distortion and max SPL are affected too, of course.

Interesting findings @jp8. Would you mind sharing the LTspice file if it’s easy to do?

 

[OneRoomStudio]

C1/C3 (the 10pf cap) should NOT be in the circuit if you want more gain. The schematic of mine that you edited is not the right one - we discussed the effects of that cap at length in the thread (which is why there were schematics with it present), but it was always made clear that including that cap will DECREASE your overall gain significantly. Reference this one as the "ORS87" (note: I'm not the one who decided to call this the "ORS87").

 

[jp8]

Interesting findings @jp8. Would you mind sharing the LTspice file if it’s easy to do?

Sure, no problem. The attached zip file contains the ORS87-Plus circuit, modelled after @Khron 's circuit on PCBWay. I have bought this PCB, btw, but haven't gotten around to actually building it. The attached LTspice circuit is slightly different from the one I refer to in post #12, but that's because I brought it back to Khron's circuit. So you will find different gains when you drop a 10pF feedback capacitor into the attached circuit, and because I changed the transformer model, which I brought closer to an AMI T-13. Mind you, this is just a basic transformer model, and simulation results will differ from actual performance. But whether it's a basic model or more sophisticated, it will help you explore how the circuit works and what effect certain changes have on e.g. frequency response, gain, THD and noise. Let me give an example using the attached circuit.

Khron's circuit has a feedback capacitor installed (C4). Let's explore the effects of changing its value. First with the 3.3pF from Khron's circuit:




The simulation predicts a 4.6 dB resonant peak at 20Hz, determined by C7 and the transformer inductance. This peak may, or may not be intended, I don't know. And this may, or may not be acceptable or objectionable. That's up to you to decide. Anyway, the sim shows that this peaking and low-end roll-off is at least a point of attention, and you want to check the actual circuit for its behavior.

Now what if we remove C4?


Gain goes up by more than 5 dB and because a smaller (or no) capacitor also means that the output impedance of the JFET rises, the resonant peak will be much less pronounced.

As I mentioned earlier, the feedback capacitor also changes THD. Without showing the sim results, the circuit with 3.3pF feedback capacitor has 9.2dB lower THD with a 100mVp/1kHz signal from the capsule. So more gain does not come for free: it changes the FR and increases distortion. This may be exactly what you want, but you need to take a holistic view of the circuit and realize that any change will not only have the desired effect on e.g. Gain or THD, but also on other properties of the circuit. The whole circuit is a balancing act between Gain, SNR, FR, THD, and max SPL. Of course, everyone is free to tip the balance more to one side or the other, but you can't just say that one must leave out that 10pF and assume that there are no further consequences. Those Neumann guys put it in there for a reason, and I assume they weren't stupid.

Happy experimenting!

 

[joulupukki]

Thank you @jp8, you are a wizard! I love learning more about all of this and I will definitely do some experimenting with the circuit in LTspice.

C1/C3 (the 10pf cap) should NOT be in the circuit if you want more gain. The schematic of mine that you edited is not the right one - we discussed the effects of that cap at length in the thread (which is why there were schematics with it present), but it was always made clear that including that cap will DECREASE your overall gain significantly. Reference this one as the "ORS87" (note: I'm not the one who decided to call this the "ORS87").
View attachment 144078

Got it, thanks for the clarification. Kudos to @Wordsushi for coining the build.

Ok, so I've made the following mods to bring my D87 build to mimic (or at least get close to) the ORS87 with these changes:

Here's before and after shots. The removals were pretty easy since they are only on one of the two circuit boards:



I haven't yet re-biased the ORS87 but that'll be next. The gain/sensitivity of the mic with the ORS87 circuit is much better. The -10dB and bass rolloff switches still work (since I left those components in). I haven't compared the frequency responses yet in REW but will give that a shot later after I re-bias it.

Here's a quick audio sample featuring this ORS87, my Fuchs U87A build, and my modded AKG P420:
View attachment ors87-vs-fuchsU87-vs-AKGP420Modded.wav

Components of these mics:
ORS87: JLI TSC-2 capsule with 3U Audio GZT-87 Transformer in an Aurycle A460 donor body
Fuchs U87A: Arienne Audio K87 capsule (v1) with 3U Audio GZT-87 Transformer in the AliExpress HL-95 mic body
AKG P420: Modded with a 470pF MLCC tacked on to the existing 220pF surface mount cap (690pF total)

I'm mostly interested in using this mic potentially as a vocal mic for music, but it'll be interesting to see how it does with acoustic instruments and other sources.

Thanks to all for helping me think through the change and teaching me more about how the circuit works!

 

[micolas]

Of course, everyone is free to tip the balance more to one side or the other, but you can't just say that one must leave out that 10pF and assume that there are no further consequences. Those Neumann guys put it in there for a reason, and I assume they weren't stupid.

The Neumann guys are not stupid. They suddenly became stupid when they ditched the 10pF capacitor, stopped making the U87i and designed the U87ai which is still being built today and is preferred by many musicians and sound engineers in many scenarios today's modern sounding music

 

[micolas]

Thank you @jp8, you are a wizard! I love learning more about all of this and I will definitely do some experimenting with the circuit in LTspice.


Got it, thanks for the clarification. Kudos to @Wordsushi for coining the build.

Ok, so I've made the following mods to bring my D87 build to mimic (or at least get close to) the ORS87 with these changes:
View attachment 144166
Here's before and after shots. The removals were pretty easy since they are only on one of the two circuit boards:
View attachment 144167
View attachment 144168

I haven't yet re-biased the ORS87 but that'll be next. The gain/sensitivity of the mic with the ORS87 circuit is much better. The -10dB and bass rolloff switches still work (since I left those components in). I haven't compared the frequency responses yet in REW but will give that a shot later after I re-bias it.

Here's a quick audio sample featuring this ORS87, my Fuchs U87A build, and my modded AKG P420:
View attachment 144169

Components of these mics:
ORS87: JLI TSC-2 capsule with 3U Audio GZT-87 Transformer in an Aurycle A460 donor body
Fuchs U87A: Arienne Audio K87 capsule (v1) with 3U Audio GZT-87 Transformer in the AliExpress HL-95 mic body
AKG P420: Modded with a 470pF MLCC tacked on to the existing 220pF surface mount cap (690pF total)

I'm mostly interested in using this mic potentially as a vocal mic for music, but it'll be interesting to see how it does with acoustic instruments and other sources.

Thanks to all for helping me think through the change and teaching me more about how the circuit works!

@joulupukki
Did you get what you wanted?
(Sounds great by the way!)
Congratulations!

 

[joulupukki]

@joulupukki
Did you get what you wanted?
(Sounds great by the way!)
Congratulations!

Thanks @micolas. Yeah, I think it's to a place where I really like it. Here's a quick clip comparing it with my Fuchs 87 after biasing. It seems a LOT more usable now. It'd probably do really well as a main center mic in my bluegrass band, though I might prefer just taking the AKG P420 since that thing is built like a tank and was cheap on the used market. In any case, not too shabby for a relatively simple circuit.

View attachment Post-Bias-ors87-vs-fuchsU87-vs-AKGP420Modded.wav

 

[micolas]

Thanks @micolas. Yeah, I think it's to a place where I really like it. Here's a quick clip comparing it with my Fuchs 87 after biasing. It seems a LOT more usable now. It'd probably do really well as a main center mic in my bluegrass band, though I might prefer just taking the AKG P420 since that thing is built like a tank and was cheap on the used market. In any case, not too shabby for a relatively simple circuit.

View attachment 144174

They all sound wonderful.
You now have 3 U87-style mics with different flavors, each usable for different sources, different situations.
You did a good job, I like it. That's DIY!
I think I have over 30 microphones with 87 and 67 style sound. I like them all and keep them.