New u87 Body

[micolas]

@joulupukki
You also need to buy more 2N3819 because jFETs have large parameter dispersion and you will need to select an appropriate one for the circuit
And the 2.2kohm resistors must be matched with high precision (personally I pair them below 0.5%), so more must be bought.

But let's not deviate from the topic, you can find out all the details on the original topic of Michael Fuchs.
Let's move the U87 rev Ai discussions there.

 

[Tronfxdx]

Where’s a good place in the US to get the Styro caps? It looks like from the BOM that I’d need:

1x 160pF 500V 5%
1x 270pF 300V 5%
2x 470pF 100V 2.5%

Where’s a good place in the US to get the Styro caps? It looks like from the BOM that I’d need:

1x 160pF 500V 5%
1x 270pF 300V 5%
2x 470pF 100V 2.5%

I have been buying them from:https://www.justradios.com/polystyrene.html Shipping is pretty fast and they seem to be good qualtiy.

 

[mrgrooves666]

I recently bought 5 of them HL-77 bodies. I like the plain look of the HL-77 body, and of course, the price, and made holes myself for regular switches with a dremmel.. 2 for a pair of mihi's 87's, and 2 for a pair of Rogs's OPIC multipattern, with Ari's capsules (F47 in the case of the OPIC pair). I have them all done just waiting on ARI's capsules, hopefully soon.
With the fifth body I built an OPA Alice with KK's V1 electret capsule. Very nice mic btw! Yes the bodies, they ring like mofos though, even with 3d printed mounts I made on my own, with the same concept of the rubber rings. I managed to reduce the ringing quite a bit by putting a very thin layer of this stuff, as suggested somewhere else in the forums, squeezed in the inside of the body tube:
https://www.amazon.com/-/es/Skywalk...28-8570-f402ea83e4a2&pd_rd_i=B00A7AXBSQ&psc=1
It is somewhat tricky, because its sticky, so I applied a coat of paint afterwards on the inside over it, to get rid of its stickyness after installing it. The ringing reduced quite significantly to where now is almost more of a solid thump.

 

[mrgrooves666]

 

[mrgrooves666]

Ought to mention, the metal rails are touching this stuff, particularly on the bottom, where the tube slims a bit. I think this might be contributing to deaden the ringing as well, so it might not be just the basket.

 

[Wordsushi]

Ok, this is worth noting. I have a pair of HL-77 bodies, the ones that seem to have that incurable 100hz resonance when you thump the side. The resonance got irritating after a while so I transplanted everything from both those bodies into the superior HL-95 body. I then "retired" the HL-77 bodies to the experimental "Microphone Island of Dr Moreau" pile of bodies to use just for unserious projects and prototypes, etc.
I then used the HL-77 to house an ORS 87 PCB and the complete build of that mic sounds pretty good to me.
Because... and I can't explain at what point this happened... the resonance disappeared.
Being so dang excited about the ORS 87, I had forgotten about the HL-77's "issue" during the build.
This is my theory.
I think it's the single PCB.
That, and the capsule mount, are the only differences between the two builds this body has housed.
The Fuchs build had twin boards, mounted back to back, a transformer zip tied to the rails, a capsule and mount.
The ORS has one board, and all the rest of the same stuff.
Except for the mount and saddle. This new mount is one I 3d printed to be a little higher than the previous one.
The previous mount screwed into the capsule deck with two screws, mounted north to south from underneath.
This mount screws in at 4 points, into the top of the capsule deck.
So it's either the new mount or just using a single board instead of two.

 

[micolas]

Ok, this is worth noting. I have a pair of HL-77 bodies, the ones that seem to have that incurable 100hz resonance when you thump the side. The resonance got irritating after a while so I transplanted everything from both those bodies into the superior HL-95 body. I then "retired" the HL-77 bodies to the experimental "Microphone Island of Dr Moreau" pile of bodies to use just for unserious projects and prototypes, etc.
I then used the HL-77 to house an ORS 87 PCB and the complete build of that mic sounds pretty good to me.
Because... and I can't explain at what point this happened... the resonance disappeared.
Being so dang excited about the ORS 87, I had forgotten about the HL-77's "issue" during the build.
This is my theory.
I think it's the single PCB.
That, and the capsule mount, are the only differences between the two builds this body has housed.
The Fuchs build had twin boards, mounted back to back, a transformer zip tied to the rails, a capsule and mount.
The ORS has one board, and all the rest of the same stuff.
Except for the mount and saddle. This new mount is one I 3d printed to be a little higher than the previous one.
The previous mount screwed into the capsule deck with two screws, mounted north to south from underneath.
This mount screws in at 4 points, into the top of the capsule deck.
So it's either the new mount or just using a single board instead of two.

All parts in the microphone contribute to vibrations, resonances, etc. so to the final sound. Possibly the new capsule mount has a important influence.
Can you post pictures of it and the positioning of the capsule in the headbasket?

 

[abbey road d enfer]

And the 2.2kohm resistors must be matched with high precision (personally I pair them below 0.5%), so more must be bought.

It's a very common misconception.
The relative value of these resistors just matters for the level of each leg (knowing they're of opposite polarity). But it has no major consequence for performance. Even if one leg had no signal, the connection would still be balanced, because balance is related to impedance not level.
The absolute value of the 2.2k resistors relates to the output impedance of each leg. Many think that making these resistors equal will provide matched impedance on each leg.
Well, that isn't true.
Since one leg comes from the source and the other from the drain, the impedances are different.
On the drain side it's very close to 2.2k, but on the source side it's 2.2k in parallels with 1/Gm, which results in about 400-500 ohms.
As a result, the output impedance at the emitters of the followers is slightly different, with one leg at 20 ohms and the other at about 13 ohms. That limits CMRR to about 60dB at best.
For the same reason, matching the transistors Hfe (Beta) is wasted effort. As it has already been mentioned, matching Vbe is more important, specifically in the case of driving transformer inputs.

 

[micolas]

It's a very common misconception.
The relative value of these resistors just matters for the level of each leg (knowing they're of opposite polarity). But it has no major consequence for performance. Even if one leg had no signal, the connection would still be balanced, because balance is related to impedance not level.
The absolute value of the 2.2k resistors relates to the output impedance of each leg. Many think that making these resistors equal will provide matched impedance on each leg.
Well, that isn't true.
Since one leg comes from the source and the other from the drain, the impedances are different.
On the drain side it's very close to 2.2k, but on the source side it's 2.2k in parallels with 1/Gm, which results in about 400-500 ohms.
As a result, the output impedance at the emitters of the followers is slightly different, with one leg at 20 ohms and the other at about 13 ohms. That limits CMRR to about 60dB at best.
For the same reason, matching the transistors Hfe (Beta) is wasted effort. As it has already been mentioned, matching Vbe is more important, specifically in the case of driving transformer inputs.

However, Neumann still pairs them quite strictly

 

[abbey road d enfer]

However, Neumann still pairs them quite strictly

If I'm not mistaken, it's the U87. It's a very different schematic. There are no emitter-follower transuistors, and there is a transformer.
These resistors must be equal because they need to draw the same current from each leg, in order to not magnetize the transformer.
My comment was in regard to the Schoeps schematic and its derivatives.

 

[micolas]

If I'm not mistaken, it's the U87. It's a very different schematic. There are no emitter-follower transuistors, and there is a transformer.
These resistors must be equal because they need to draw the same current from each leg, in order to not magnetize the transformer.
My comment was in regard to the Schoeps schematic and its derivatives.

I'm sorry for the misunderstandings
I was referring to the U87 topology, that's what the thread was about

 

[Wordsushi]

All parts in the microphone contribute to vibrations, resonances, etc. so to the final sound. Possibly the new capsule mount has a important influence.
Can you post pictures of it and the positioning of the capsule in the headbasket?

In retrospect, I don't think it's the mount. I forgot I had swapped this mount out for one with a wider saddle when I put the 797 capsule in this past weekend and the resonance issue was not present before that. I found a contemporaneous note on my phone going back two weeks that would indicate the low end sounded "pretty clean" to me at that time, so I'm going to assume I would have noticed and noted the resonance had it been there.

So, as unlikely as it is, I'm going with my theory that using just one board instead of two has dampened the resonance. Same thing in my 2nd HL-77 body. I don't know why this would be the case. It doesn't really make sense to me, yet. But, at this point I'm just going to accept the results and put the HL-77 back on my list of bodies I'd buy, if only for use with the single board ORS87.

 

[micolas]

In retrospect, I don't think it's the mount. I forgot I had swapped this mount out for one with a wider saddle when I put the 797 capsule in this past weekend and the resonance issue was not present before that. I found a contemporaneous note on my phone going back two weeks that would indicate the low end sounded "pretty clean" to me at that time, so I'm going to assume I would have noticed and noted the resonance had it been there.

So, as unlikely as it is, I'm going with my theory that using just one board instead of two has dampened the resonance. Same thing in my 2nd HL-77 body. I don't know why this would be the case. It doesn't really make sense to me, yet. But, at this point I'm just going to accept the results and put the HL-77 back on my list of bodies I'd buy, if only for use with the single board ORS87.

Probably for VO where you don't have to put up with crazy SPL on low frequencies, the HL-77 headbasket will work ok.
In addition, it has the right shape and is a little cheaper (although they are getting more expensive all the time, and China complains that it has deflation and prices drop dramatically)

 

[soliloqueen]

It's funny, you can kind of tell the "lineage" of a capsule by looks and part choices. 797, 3u and in-house rode look super similar, due to where the founders used to work or what teams used to work together in the 90s and 2000s. So do takstar, aym and shanghai, for similar reasons (though they look less similar than the above three, which can be nearly indistinguishable depending on the model)

 

[thomasdf]

Hello everyone, about a year ago I bought two of these bodies, and now I am working on the circuit. I plan to go as close as possible to 1:1 u87i , the old version. I know it wont be 100% accurate, but I am looking for something with that sonic signature, and colorful sounding, if it means anything.

The body does not have holes for switches and I think toggles are ugly so I was think of making it cardio only, no low cut, a possibly a pad button, in the least ugly way I can make it.

Here are the schematics as of now :

edit : schematic updated / corrected, see post below

It's 1:1 the original schematics, minus the VU / battery part, that I replaced with info taken from some schematics found on the forum (Chinese simplified U87, IIRC)
Does everything look all right to you ?

You'll notice I added the capacitors and inductors on the output as an option, I am curious to ear to what extent they change the sound. Same goes for the tantalum, I'd need to check which ones were tants in the original, and I am curious how much it contributed to the tone.
I have also removed the "cardio" unlinked pad for now on.

The body being quite narrow, I think I will need to make two PCBs and split the components on both.

I plan to upload the gerber file for anyone here to use, when tested. There won't be a Switch PCB, but all functions of the original will be there : pattern, low cut, pad.

I want to use Ari's K87 or cardio only 87 when ready, and I haven't made my mind up yet re the output TX.

Thanks

 

[Khron]

so I was think of making it cardio only

So basically, ORS87 then?

 

[thomasdf]

I am not familiar with the ORS87, I'll look it up. Is it a non ai cardio only accurate repro ?

My thought was also to share the design with the community here, and let some options open for let's say a full spec U87i.

I now there is a lot of work involved (I am doing it right now haha) but paying around 40 euros for a non proprietary design, two times, makes my wallet and brain hurt. My time is free, and then JLC PCB or any other PCB maker can make some rather cheap and beautiful boards.

Edit :

So basically, ORS87 then?

I realize now this is where I got the 56K resistor R17 on my schematic.
I still think there may be a need or interest for a full spec PCB. I can always leave out component and options for my no LC no pattern build. It also allows to add these back in in the future, if needed. Or if I decide I can live with a pair or 87 bearing little antenna-like switches around the headbasket haha

 

[Drosselmeier]

 

[Khron]

I now there is a lot of work involved (I am doing it right now haha) but paying around 40 euros for a non proprietary design, two times, makes my wallet and brain hurt. My time is free, and then JLC PCB or any other PCB maker can make some rather cheap and beautiful boards.

https://www.pcbway.com/project/shareproject/ORS87_Plus_Open_source_U87_mic_project_2922baaf.html

 

[thomasdf]

@Drosselmeier thank for your sharp eye ! Here is a corrected version
@Khron thanks a lot for the link. I think I do not understand something because to me the schematics are quite different. Different cap values, different resistors value... What am I missing ? Is the ORS87 based on the U87i ?