Neumann Vintage U87 Clone : Build Thread.

[Moby]

Hey guys, does anybody use Moby's Bv.13 in the D-87M blue?
View attachment 112372
Maybe someone can help me with wiring.

View attachment 112371​

This is from Moby's Site.

is it correct that i have to solder the 2 white one in the PRI together and in the SEC green and yellow?

Then connect it to the PCB:

black --> P+
red --> P-
white --> S+
blue --> S-

is that correct?
@Moby ?
Thanks in Advance!

Damian

Yes, correct.

 

[paolo.mangiagalli]

Thanks to this fantastic community, I just finished build for my son couple of U87, using SYT bodies (just got them before Jenny's move), Dany's blue PCM, Cinemag 13113, one with a SYT supplied capsule and one with Soliloqueen Capsule. They both work out of the bench, apart the fig.8 of the chineses capsule, probably due to lack of bacplate insulation. I am not trained to appreciate differences, will let this to my son who is completing his training as sound engineer and has planned an A/B/C with a Neumann u87. Again thanks to all !

Now my question. I have assembled the PCB with a low IDSS 2N3819 (8mA), orienting the FET as in the serigraphy, also looking at the build example on Dany's site. By looking at the datasheet IC pinout and the schematics of the board, it seems to me that the 2N3819 is wired with source and drain reversed. Option 1: I read wrongly the schematics Option 2: the circuit works also with the reverse wiring. From my crude knowledge of JFET manufacturing, this can also be the case; the only real impacting change from wiring a 2N3819 in reverse should be the differences in Gate-Source vs Gate-Drain capacitance. I ahve also read somewhere that for the 2N3819 the CGD is higher that GGS, contrary to most JFETs, but I couldn't confirm. Any thougths appreciated.

 

[abbey road d enfer]

From my crude knowledge of JFET manufacturing, this can also be the case; the only real impacting change from wiring a 2N3819 in reverse should be the differences in Gate-Source vs Gate-Drain capacitance.

That's typically the case with many "simple" FET's (that don't include protection diodes or other additions).

I ahve also read somewhere that for the 2N3819 the CGD is higher that GGS, contrary to most JFETs, but I couldn't confirm. Any thougths appreciated.

All datasheets mentioned 8pF "input" capacitance and 4pF reverse capacitance. Wiring it in reverse would almost always result in lower "gain" (sensitivity) and degraded Signal-to-Noise ratio..
You may be able to make a direct comparison, using a tone generator and a capacitor replacing the capsule and measure the resulting gain.

 

[joulupukki]

Newbie here and just a little bit confused. I’ve got an Aurycle A460 I’d love to convert to this U87 build. But, in all the pictures I’ve seen so far of the taper builds, I don’t see where the transformer would actually fit. I’ve opened up my A460 and it doesn’t seem like there’s enough room inside for a transformer. What am I missing?

 

[Horizoneer]

Newbie here and just a little bit confused. I’ve got an Aurycle A460 I’d love to convert to this U87 build. But, in all the pictures I’ve seen so far of the taper builds, I don’t see where the transformer would actually fit. I’ve opened up my A460 and it doesn’t seem like there’s enough room inside for a transformer. What am I missing?

I don't know the A460, but doesn't it have a metal bell cover down at the XLR end that the transformer sits inside?

 

[joulupukki]

I don't know the A460, but doesn't it have a metal bell cover down at the XLR end that the transformer sits inside?

Ah ha, look at that! Duh. I guess I didn't think to disassemble it and find out. There it is. Thank you.

 

[OneRoomStudio]

Does anyone happen to know if either the D-87L or D-87M boards would fit in an M-audio Nova body (without the switches)?

 

[joulupukki]

If I’m building this project into an Aurycle A460 donor body, do I need to order the switch PCB or can I just connect the existing wires from the pad and cut switch to the taper PCB? Looks like from previous reports I would need to switch the leads of the low cut switch to have it work properly.

For the pattern selector, I’m guessing I could just use some internal jumper posts and unscrew the body to switch when/if needed, right?

 

[mrgrooves666]

If I’m building this project into an Aurycle A460 donor body, do I need to order the switch PCB or can I just connect the existing wires from the pad and cut switch to the taper PCB? Looks like from previous reports I would need to switch the leads of the low cut switch to have it work properly.

For the pattern selector, I’m guessing I could just use some internal jumper posts and unscrew the body to switch when/if needed, right?

Yes and yes. No need for switch pcb

 

[joulupukki]

I'm just curious if anyone knows where the PCBs ship from? I placed an order last Thursday for a PCB and caps I could use to build the U87 into an Aurycle A460 body I've got but haven't heard anything. Just wondering what I should expect there, if anyone knows.

 

[fsalter]

"I'm just curious if anyone knows where the PCBs ship from?"
Dany's business is in Montreal, Canada. I know him well and could ask him what's the hold up. But no worries, he may be busy. I know he has a teaching job too.

 

[joulupukki]

I've almost got all of the parts needed for this build. With the Aurycle A460 donor body, do I need anything in particular to mount the K87 (an Arienne Audio K87) capsule or is the stock A460 mount going to work? Seems like the stock capsule is smaller.

 

[joulupukki]

I've almost got all of the parts needed for this build. With the Aurycle A460 donor body, do I need anything in particular to mount the K87 (an Arienne Audio K87) capsule or is the stock A460 mount going to work? Seems like the stock capsule is smaller.

I discovered the answer to my own question by just diving in and replacing them. The pedestal and saddle are a flexible one-piece material and the two screws holding the old capsule in fit perfectly into the Arienne Audio K-87 capsule I’ve got. The K-87 capsule doesn’t go all the way down to the bottom of the saddle, but it seems to be held in place nicely enough. Just waiting for the PCB to arrive now (will hopefully be tomorrow).

 

[joulupukki]

Finally have it built! Like others, initial thoughts are that it has lower output than I had hoped. I'm curious if a different choice of FET (with lower/higher Idss/Vg values) would make much difference in that? Self noise seems fairly quiet. It has more bass response and a little less treble than I had expected, but the only thing I've got to compare it with is the Townsend Labs Sphere L22 and its mic models (and some KM-84 clones I've built). Compared with the KM-84 clones the lack of higher frequencies is definitely noticeable. The low cut filter switch seems to work pretty well.

I biased using a trim pot and then combined two 6.8K resistors in parallel and ended up close to the 13.66K reading on the trim pot. I'm using a 2N3819 from Mouser with Idss of 10.7mA and 3.53V Vgs (off). I installed it into the Aurycle A460 mic body (since I already had one of them collecting dust). It's using the Arienne Audio K-87 capsule and a 3U GZT-87 transformer.

I made the attached sample recording with no EQ or post-processing through an UA Apollo Twin | X and the Neve 1073 pre-amp emulation set to -50dB.




View attachment diy-u87-mic-voiceover-test.wav

 

[soliloqueen]

I discovered the answer to my own question by just diving in and replacing them. The pedestal and saddle are a flexible one-piece material and the two screws holding the old capsule in fit perfectly into the Arienne Audio K-87 capsule I’ve got. The K-87 capsule doesn’t go all the way down to the bottom of the saddle, but it seems to be held in place nicely enough. Just waiting for the PCB to arrive now (will hopefully be tomorrow).
View attachment 126005
View attachment 126006

i put extra holes in the backplates of my k87s so that they can be mounted at 90 degrees all front, at 90 degrees front and back, at 60 degrees all front, and at 60 degrees front and back, all while being able to mount a cover in at least one of those arrangements and have it be at the top of the unit. it adds a fair chunk of change to side drill and tap that many holes per plate, so it's nice to see it pay off for someone

 

[Horizoneer]

I think the idss makes a difference to output level, based on nothing more than memories of reading through this thread when I built my pair a year ago. I lucked out, a friend found some old FETs in a market, I was able to pick two from a batch of 12, at 3 & 4ma idss. The 3ma one could be biased slightly hotter before the onset of distortion.

 

[joulupukki]

Nice. I'm gonna remove the 2N3819 and switch it out for a J305 that has an Idss of 6mA just to see if there's any difference. I read a few things about this stuff in this long thread, but not sure all of it makes sense to me yet. MP uses an Onsemi J305 in their SDC-84 build and so it may be worth a shot here too.

 

[joulupukki]

I don't know if my ears are just playing tricks on me, but I think I may like the sound of the J305 JFET in this mic better than the 2N3819 that I had in there before in the previous recording. I could be mistaken, but with the J305, the mic seems to be a little less boomy and also has a little more higher end definition. But maybe it's all just in my head.

With the 2N3819 I had in there, I chained two resistors together for R11 with a total value of 13.66K. With the J305, I used a couple resistors to reach 7.88K. The J305 has a bit different pinout (Pin 1 is the gate) so it's installed 180° compared to the 2N3819.



Here's a sample recording with the J305:

View attachment diy-u87-mic-with-j305-voiceover-test.wav

Edit:

Yep, definitely it seemed a bit more woofy (more bass response) with the 2N3819 I was using. I'm not sure if that's good or bad, but right now I'm leaning to keeping the J305 in there. Here's a music sample with no EQ and only slight reverb. I recorded the lead vocal and guitar using the U87 on my voice and my DIY KM-84 on my guitar at the 12th fret. The tenor and baritone vocals I added after-the fact using the U87. The upright bass is also recorded with the U87 after-the-fact about 8 inches from the bridge. Small reverb was added to everything but the bass. The guitar has a little bit of a stereo effect on it just to add space. But, it seems like the layers stack fairly well with no hyped high end. Definitely with some targeted boosts in mixing it should do pretty well!

View attachment DIY U87 Singing Test J305.wav

Another note...

Comparing the current config (J305), a recording I made with the 2N3819, and with the Townsend Labs Sphere L22 it seems like the 2N3815 is maybe more "true" to the L22's vintage U87 mic model because it has more low end content. I'm not sure long-term which I'll keep, but the low-end content is likely stuff I'd roll off when mixing anyway. I'm interested to hear what others' experience has been.

 

[MicMaven]

Comparing the current config (J305), a recording I made with the 2N3819, and with the Townsend Labs Sphere L22 it seems like

Um ... er ... ahem ... BOY! Sheesh! This one is really tough, but considering you are crazy enough to ask MY opinion, ... While MY ears LIKE BOTH iterations, A LOT, I find the first recording just the slightest little bit more raw, clear, and almost scary realistic, as if you are speaking right into my ear (which is really good,) and I find the second recording ever so slightly more clinical, smooth and mature, but sounding more like a recording. I do not discern more or less bass in either, the difference I hear is one has slightly more immediacy and presence, in a very real sounding way. While I suspect real audio engineers will prefer version 2, I would keep version 1.

Very subtle difference to my ears. OK, I am just a rookie and probably missing something obvious, so sue me, but that is just MY take. (He typed as he donned his flameproof Teflon(R) suit!) James

 

[joulupukki]

Um ... er ... ahem ... BOY! Sheesh! This one is really tough, but considering you are crazy enough to ask MY opinion, ... While MY ears LIKE BOTH iterations, A LOT, I find the first recording just the slightest little bit more raw, clear, and almost scary realistic, as if you are speaking right into my ear (which is really good,) and I find the second recording ever so slightly more clinical, smooth and mature, but sounding more like a recording. I do not discern more or less bass in either, the difference I hear is one has slightly more immediacy and presence, in a very real sounding way. While I suspect real audio engineers will prefer version 2, I would keep version 1.

Very subtle difference to my ears. OK, I am just a rookie and probably missing something obvious, so sue me, but that is just MY take. (He typed as he donned his flameproof Teflon(R) suit!) James

Hehe. Great descriptions of the two recordings. I cranked up the volume and listened to them more critically and your description is spot on – #2 definitely sounds more “recorded” and it’s definitely different.

I think what I’m going to do next is use a smaller trim pot that can stay inside the mic body and just leave it in there permanently. Then, I’m going to try one of the other 2N3819 JFET that has a lower Idss. Something in-between these two sounds would be perfect. Just a tiny bit less low-end than the #1 sample would be ideal for my needs. I won’t be able to try it for a few days, but I’ll report back when i do.