Open source project: "Generic" Shoeps-style microphone PCBs on Osh Park

[midwayfair]

A forum member recently posted a WTB for discarded PCBs from a fairly common MXL Schoeps-style mic.

As MXL and other manufacturers selling basically the same product move to SMD, it's harder and harder to find a mic that can just be straight-up modded, and for the most part I don't think the PCB guys are going to be very interested in making a project for sale that's a copy of a dirt-cheap mic.

So I've decided to just make this an open source project for a couple of these microphones.

I've started with the Schoeps style circuit, fitted to the body that's used in the V63/V67 (and the numerous similar mics by Thomann etc), Aurycle (IMO the best completely blank bodies you can get for this size - unpainted but heavy brass), BM800, etc.

******These documents may change until they are verified by both me and at least one other user!******

Schematic (the topology works with the stock MXL etc circuit but it has been reverted to original parts or modified where it's beneficial):

PCB:

You will need a capsule polarization circuit for this project.

I started another thread for those here: https://groupdiy.com/index.php?topic=65174.msg825897

OSH Park Shares:
https://oshpark.com/projects/Uvt6XWLy (Main PCB)
https://oshpark.com/profiles/midwayfair (all my Osh Park shares in case I forget to update the other links)

And Mouser BOM: https://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=ef408f27d6

Oscillator PCBs on OSH Park:
https://oshpark.com/projects/PdZO7cBb (Single-output voltage oscillator PCB)
https://oshpark.com/shared_projects/Hf8U2pxB (dual-voltage)

This project is CREATIVE COMMONS, SHARE ALIKE.

Once a project is verified, if there is any need for it, I can make the Eagle files publicly available (though most people who mess around with Eagle could re-draw these in a matter of hours by just copying the pictures).

 

[Khron]

I see hole markers on the boards - they're definitely too close together on the audio board.

 

[midwayfair]

I see hole markers on the boards - they're definitely too close together on the audio board.

I thought so too, but it was Eagle putting a "halo" around the holes. This is what it looks like in the Osh render -- there's a good mm between the holes, and you're only using one or the other:

https://644db4de3505c40a0444-327723bce298e3ff5813fb42baeefbaa.ssl.cf1.rackcdn.com/112aa3e574fbfd3963450ec76bf7459e.png

Oh! I forgot to mention -- if anyone knows the screw type that's usually used to hold in capsules, I'd like to add those to the mouser BOM if they have them. I like that they have a "washer" attached to the screw itself on the top.

 

[Khron]

I dunno, that still looks a bit "too close for comfort", imho. Surely there should be room inside MXLs and stuff, to offset those holes vertically by another 2-3mm, isn't there?

 

[midwayfair]

I dunno, that still looks a bit "too close for comfort", imho. Surely there should be room inside MXLs and stuff, to offset those holes vertically by another 2-3mm, isn't there?

I'm not really sure. I don't have an example of the one with the closer spacing (which I got off the graphic from the other thread). The holes closer to the top are the exact positioning of the Aurycle boards, so I know they're good.

You're only going to use one of the holes at a time, so are you worried about a fab house messing up the drills? The holes are farther apart as the closest actual component drills.

 

[Wordsushi]

I think the screws that hold the pcb to the rail are M1.6 x 5mm.

 

[midwayfair]

I think the screws that hold the pcb to the rail are M1.6 x 5mm.

I think those are the screw size for screwing the rail to the metal plate that the capsule sits on. The ones that go into the rail are only like 1mm. They almost fit in a PCB drill. They're also the same size as what goes in the capsule itself, which have a wider head that looks like a washer is pre-attached. I don't think I could get a good picture of them. I specifically want to find the ones that go in the capsule because I think they're better at holding the PCBs in place.

 

[Wordsushi]

On the Micparts kits, they send out M1.6 screws to attach the PCB to the rails. Those smaller screws that are used to attach the capsule to the mount are too thin and don't catch in the pre-tapped rail holes.  M2's are the size that screw the rails to the capsule mount plate and also the headbasket to said mount plate.

*I'm not seeing the image on my end, so here's the link

https://www.dropbox.com/s/nfvupbosp7lew46/IMG_2902.JPG?dl=0

 

[midwayfair]

On the Micparts kits, they send out M1.6 screws to attach the PCB to the rails. Those smaller screws that are used to attach the capsule to the mount are too thin and don't catch in the pre-tapped rail holes.  M2's are the size that screw the rails to the capsule mount plate and also the headbasket to said mount plate.

*I'm not seeing the image on my end, so here's the link

https://www.dropbox.com/s/nfvupbosp7lew46/IMG_2902.JPG?dl=0

hm. okay. I'll change the size of the holes and re-upload in a moment.

Mouser doesn't seem to carry them (or if they do, the measurements for their screws and fasteners section is too arcane for me). Anyway, to the hardware store go I.

 

[homero.leal]

Looks great Midwayfair... congrats and big thanks!!!

Quick question...  are you using a 3.3v zener diode and 3 leds, instead of a 12V zener?

Is this for reduced noise?

Kind regards!

HL

 

[midwayfair]

Quick question...  are you using a 3.3v zener diode and 3 leds, instead of a 12V zener?

Is this for reduced noise?

Yes. The Blue LEDs I have measure 2.8V and the white LEDs I have measure almost 3V exactly.  [EDIT: These are bad measurements according to the datasheets. Should be 3.2V for white. I've included a 2V4 on the mouser BOM for people who want to hit 12V exactly, but I think 3V3 is a more common in-the-bin part.] Three of them plus a 3.3V is close enough to 12V for the circuit to work. (People can get closer if they want.) Here's my note from the thread this carried over from:

I made an important change in the main schematic which we didn't discuss yet: I made the Zener regulator from a lower value zener and white LEDs. I had found this thread a while back when I was looking for a solution to Zener noise. Since we only need to hit 12V or less on this regulator, this seemed like a good opportunity to use it (I can't very well do so if I need a 30V regulator, for instance). Zener diode noise is extremely well discussed in many, many, many places if anyone needs to look up more info. The 6k2 should be sufficient current limiting, but you might want to take a black marker or paint over the LEDs to blot them out so you can actually see the circuit ... them things are bright. If anyone has further suggestions about this, let me know before I call this one ready to prototype.

Some other notes:

4) I'm still not 100% sure about the trimpot value in the oscillator in this mic (I've only done voltage oscillators this month for the first time, and it was all in u87 variants), but I bumped it up to 10K.

5) I put pads on the board for the capsule, but I  would just do it deadbug stye.

6) I added a high pass filter that has worked well for me in the past. Yes it changes the bias slightly, but they're matched so you should be fine. I cannot measure or hear the noise increase on my equipment (but I assume it's there). It's 150Hz but if that's too much for you just increase the size of the resistor.

7) I added a pad. I opted not to do it like the original because it would have involved an extra wire run. If you're going to use it, you can route the capsule backplate connection through the switch on the way to the PCB. It's only connected to the PCB when you've got the pad in, so the PCB resistance (which I think is about 300M for FR4) loss shouldn't be a problem.

8) As suggested, I left out the input capacitor. If it's my mic, though, I'm probably ditching the trimpot, grounding the 1G, and selecting R10 for the bias. I couldn't fit both trims and it made for a MUCH better layout to just omit the cap completely.

9) Lastly, I've opted to omit the corrective EQ for the capsule. Here's my thinking: Anyone who is grabbing one of these PCBs probably isn't satisfied with the 25mm 67-style LDCs and the capsule should be the thing they replace before touching the circuitboard. There are many better options for capsules out there. If you want to get super cheap about it, you can even just omit the oscillator completely and this works as an Alice and the transsound large electrets don't usually need EQ correction either, and frankly I think they sound awesome compared to the stock capsules that come in the 990 etc. (Though I think that the larger 67-style one that comes with the Aurylce FET kits sounds better than both and I like that in my modded 990 over the other Kx7 capsules I tried.)

Cheers.

 

[homero.leal]

Thank you Midwayfair, great reference!

Quick question again...  ;D Why not 4 x3v zeners? On the thread, RuddNL said he used 3x2.7v zeners.
Samuel measured 3.3v zener noise at just 5 or 7nV/rtHz.

Just trying to understand the rationale.

Thanks again, and kind regards!

HL

 

[midwayfair]

Thank you Midwayfair, great reference!

Quick question again...  ;D Why not 4 x3v zeners? On the thread, RuddNL said he used 3x2.7v zeners.
Samuel measured 3.3v zener noise at just 5 or 7nV/rtHz.

Just trying to understand the rationale.

Thanks again, and kind regards!

HL

The LEDs make no noise at all -- plus the Zeners are twice as much PCB space. It looks like there's a lot of space on the PCB, but there really isn't because I needed to make really sure that the parts were nowhere near the rails, which can be kind of bendy in these mic bodies. I could maybe cut it a bit closer but why risk it. Or do standing resistors (yuck).

Also, while the LEDs in the mouser cart are 50c each, that's just at mouser and I expect people will know cheaper ways to get them or already have them in the bin along with most of the other components. They're like 3 or 4 cents at Tayda, for instance.

 

[homero.leal]

Perfect... and understood! Thank you Midwayfair..,

Sorry if asking this but, would you consider two versions of the main board, one with Ricardo's approach to HFEQ, and the other one with Henry's approach, as it is now?

This is cause I have read that resistors between 0.1uf  caps and PNPs will induce some noise. We don't want to lose what we gain with using the leds. Just a suggestion...

 

[midwayfair]

This is cause I have read that resistors between 0.1uf  caps and PNPs will induce some noise. We don't want to lose what we gain with using the leds. Just a suggestion...

If it's my mic I'm jumpering those resistors and going with the original schematic. I also wanted to ideally keep this to a single layout until it's verified. But here's my thoughts on this:

Audio improv didn't measure an increase in noise with the added resistors. I believe him. I am also pretty sure it's more of a S:N ratio issue than the components themselves adding noise -- they form a voltage divider with the 150K resistors. Even at 10K, which I believe is the common value, you are losing less than a dB of signal.  The signal is actually not perfectly balanced at that point already, because the drain output impedance is higher than the source output impedance (by a couple K at least I think). Ideally you'd pick the resistors and capacitors for the high pass filter to reflect the total resistance including the output impedance of the transistor pin, in which case you're already deviating from the schematic and stock BOM. In any case, if you're worried about the size of the resistors, use very low values and go bigger on C4 and C5! The voltage is very low here and you should have no trouble finding film caps that will fit at almost any capacitance below 470nF (which is waaaaay more than you need).

 

[homero.leal]

Thank you Midwayfair...  appreciate your time and patience!

Board looks great as it is now. For sure they will be  great boards for DIYers.

HL

 

[Wordsushi]

Midwayfair,
Quick noob question about C3/4. Is there a specific reason you went with 100nf as opposed to something larger? I ask because in the Schoeps mods I've done in the past, I've been partial to using 1uf to get more low end for VO.

Thanks for this!

 

[Khron]

The high-pass those form, with those 150k resistors is at 10Hz with 100nF, and 1.1Hz with 1uF. Those are the -3dB points, by the way.

Whatever difference you may have heard must've been "all in your head", no offense

Midwayfair,
Quick noob question about C3/4. Is there a specific reason you went with 100nf as opposed to something larger? I ask because in the Schoeps mods I've done in the past, I've been partial to using 1uf to get more low end for VO.

Thanks for this!

 

[midwayfair]

Midwayfair,
Quick noob question about C3/4. Is there a specific reason you went with 100nf as opposed to something larger? I ask because in the Schoeps mods I've done in the past, I've been partial to using 1uf to get more low end for VO.

Thanks for this!

You mean C2 and C3?

You can use a larger value if you want. There's space on the board for anything up to a 1uF 63V box cap. But it's going to make the LPF less useful (ranging to useless once you're above about 180nF) because you really don't want to go below the 10K with that setup. Khron beat me to it on the calculations.

Capsule problems are far more likely than capacitor size problems.

 

[Wordsushi]

Thank you for helping me better understand it. So much to learn!