Open source project: Little oscillator PCBs for LDCs shared on Osh Park

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

midwayfair

Well-known member
Joined
Apr 7, 2015
Messages
606
Location
Baltimore
I decided to move the oscillators for the "generic" microphone PCBs to their own thread.

These are oscillators to provide an adjustable polarization voltage to condenser microphone capsules. You will need these if you are building one of the projects with no on-board polarization voltage. These are also an "upgrade" of sorts for a mic that uses only the phantom-power voltage for polarizing the capsule, though I think it's rare for a microphone that the PCBs fit to have only one circuitboard.

Project background:
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 sellers on here are going to be very interested in making a project for sale that's a copy of a dirt-cheap mic.

These PCBs are 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 (I believe that's the same body shape and size), etc.

Standard warning:

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

This is a work in progress -- I am going to make the PCB order this weekend, though, so I'll be able to test all the open source project PCBs at once.

Schematics and layouts:

oscillator-for-schoeps-circuits1.png


oscillator-for-schoeps-circuits-pcb.png


This one's for single-pattern mics where you want a single polarization voltage that's higher than what you get from the phantom power alone. Usually it's set to +60V, but you can make it higher if you want.

OshPark share link: https://oshpark.com/projects/4gPWNMAc

Mouser cart: http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=5a0be69ef3

There are some notes on the bill of materials that matter depending on your exact needs. Typically this will be paired with the Schoeps-style transformerless project PCB (https://groupdiy.com/index.php?topic=65033.msg824052), in which case your supply voltage will be 12V, but if you want to use it with one of the transformer-coupled mics, the oscillator connection point source voltage is higher (usually closer to 35V), so you'll need the higher-value trimpot (25K). You can also pick between a multi-turn or single-turn trimmer (which is slightly cheaper).
multipattern-oscillator-schem5.png


multipattern-oscillator-pcb.png


This one's for multipattern mics. It creates adjustable +/- capsule polarization voltages for multipattern condenser microphones. Of course, you could build the multipattern board with only a single voltage if you like. It's identical except that one of the rectifier diodes goes to ground instead of the supply voltage (six of one and a halfdozen of the other).

OshPark share link: https://oshpark.com/shared_projects/4HwGZIpj

Mouser cart: http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=b0c39c326c

Like the single-pattern, there are some notes on parts depending on exactly what mic you want to use it with.

I'll do a full build document once I've had a chance to verify the PCBs. If anyone wants an extra OshPark PCB -- because I won't be building three of each of these! -- let me know.

Thanks to the forum members who helped me out with the oscillators in this thread, the members who have PMed me (you know who you are!), and AudioImprov.

The oscillators are based directly on the one found in the Schoeps microphone and the one found on Audio Improv (which is the same except a slightly different position of the 27pF cap --- and I don't know why it's different in his but it works either way).
 
For the record: this type of oscillator depends kinda crucially on the coupling between the two inductors. All the ones i've seen in factory mics, even the ones using SMD components, have through-hole (axial) inductors, parallel and right up next to each other (and hot-snotted together).

I've found that out "the hard way" - i've been... having a few headaches with my own design (partially down to my selection of SMD inductor and/or part layout)...
 
Khron said:
For the record: this type of oscillator depends kinda crucially on the coupling between the two inductors. All the ones i've seen in factory mics, even the ones using SMD components, have through-hole (axial) inductors, parallel and right up next to each other (and hot-snotted together).

I've found that out "the hard way" - i've been... having a few headaches with my own design (partially down to my selection of SMD inductor and/or part layout)...

Yeah, I've been warned about this. I have an alternate layout (based on the perfboard layout I made) if I get a problem. Do you think I should cancel the Oshpark order I just made?
 
Actually, this turned out to be a really simple fix. PCB images have been updated, I'll update the OP with the Osh Park links in a moment.
 
Suggestion...

Will be possible for you to align 220uf cap, this way?

DCDC_Board.png


Sometimes my favorite 220uf or 330uf caps are to tall, and this could easily allow bending the cap leads, to place it horizontally, like this:

images



Thanks and kind regards!

HL
 
Good idea! At some point I probably had it that way but got overly worried about having nice symmetrical lines.

I will say too that that cap is likely unnecessary with any properly filtered supply voltage -- particularly if there isn't a zener involved anywhere.
 
The Multipattern oscillator is working.

However, as usual, I made a derpy mistake: D1 and D3 are backwards. I actually figured it out by accidentally installing one of them backwards and wondering why hat one worked and the other one didn't!

I will upload the replacements tomorrow, and they'll be marked with the version. In the meantime the layout etc hasn't changed, so if anyone happened to order one of these, just know that you will have to flip those two diodes.

I "psuedoverified" the other oscillator. The layout is extremely similar, so I just installed D1 the slightly different way. It works just like it should. However, I wouldn't blame anyone for waiting until I verify it.
 
Excelent news, Midwayfair!

Thank you  for sharing your great work with the community! Much appreciated!

If you have the time, please don't forget the change in the electrolytic orientation...

Kind regards!!!

HL
 
homero.leal said:
If you have the time, please don't forget the change in the electrolytic orientation...

Already made this change, was just waiting on verifying the boards in case I had to make more changes :)

Thanks again for your help, I really appreciate it.
 
Okay, the multipattern version is finalized. The only change (aside from flipping the two diodes) was that I added a second pad for the +60V voltage ( in case someone doesn't want to wire the front membrane wire directly to the pattern switch for some reason).

Eagle files if you want to play around with it:

https://drive.google.com/open?id=0B5Sa1JDj8POIN1ZNbzZUcFVJU1k
https://drive.google.com/open?id=0B5Sa1JDj8POIWDVJMG5ucHhhelk
 
The eagle-eyed Homer Leal spotted that my ground plane missed C10's grounded side. I added a little fill to correct that problem, so it's updated on the Eagle files and the Osh Park share link. (The pad's right next to another ground connection if anyone needs to fix an existing board.)
 
midwayfair said:
These are oscillators to provide an adjustable polarization voltage to condenser microphone capsules. You will need these if you are building one of the projects with no on-board polarization voltage.

Thanks for the project!  Which projects don't include the on-board polarization?

I built a Schoepps style mod in a V67 and was pleased with the result, even with stock capsule.  Without looking at the circuit (was years ago) I'm guessing it did not include on-board polarization.  (It was only one board.)  What is the advantage of controling / raising the polarization voltage?
 
tommypiper said:
Thanks for the project!  Which projects don't include the on-board polarization?

I built a Schoepps style mod in a V67 and was pleased with the result, even with stock capsule.  Without looking at the circuit (was years ago) I'm guessing it did not include on-board polarization.  (It was only one board.)  What is the advantage of controling / raising the polarization voltage?

The only PCB out of the handful that I put up on Osh Park that has on-board polarization is the transformer-coupled version that specifies onboard polarization. (The other entries on my Osh Park profile specify that they do not so hopefully people won't get confused.)

If you have a total of 4 transistors or more in the mic you modded, it probably has an on-board polarization scheme. It would fit with standing resistors (to which I have a personal aversion :) ).

The reason for the higher polarization is that you can get a slightly better signal to noise ratio. If you increase the polarization on the capsule, there is a larger total voltage difference between the capsule sides, which means that it will make a bigger voltage to be passed onto the FET to be amplified (and impedance converted). So more voltage at the capsule = bigger raw signal. You don't necessarily NEED to hit 60V, especially on a shoeps-style circuit, and there are certainly times that you can make a microphone too loud for the preamp you're pairing it with depending on what you're recording.

p.s. Sorry for the dead schematic and PCB images. I have to migrate them to my Wordpress site because Dropbox doesn't allow me to host them the way I was using it. I'll have it fixed in a bit, I've just been really short of time lately, and I also wanted to find a host that allows me to replace files while keeping the same URL.
 
Take care with the inductors!
I found that sometimes they are wound 'clockwise', but sometimes 'anti-clockwise', causing the oscillator not to work!
(If you have the wrong combination.)

By the way: I think the last filter (1 G.ohm-0.1 uF) might be a bit of an overkill...
(The oscillator works in the MHz range, after the first filter, all the 'ripple' is gone!)
Also it will take much longer to get the final polarisation voltage.  The RC time = 100 seconds with 1 G.ohm/0.1 uF, that is almost 2 minutes!
 
RuudNL said:
Also it will take much longer to get the final polarisation voltage.  The RC time = 100 seconds with 1 G.ohm/0.1 uF, that is almost 2 minutes!

Ah. This explains why the mic I built with the verified board took longer to "warm up" than the previous build.

Would it be better for everyone to use 10nF in place of C4, C5, C8, and C9 as well?

Also, is there any test for which way an inductor faces? Tapped inductors would make it obvious, but is there a way to tell on one of the 7mm inductors?
 
I think in the 'dual polarity' circuit, C7 should be 1 nF. (Not 100 nF)

I found out about the inductors after making a prototype for a microphone manufacturer.
The prototype worked, but the production models had a problem with the polarisation voltage.
It was apparently a 'bad batch' of inductors! (Wound in the opposite direction.)
I made a test rig to test the inductor pairs 'in circuit', to be sure that the combination will work.

I think if you keep the filters as designed now, 1 M.ohm and 10 or 22 nF is more than ample!
(That is: with R3 and R5 also 1 M.ohm.)
 
RuudNL said:
I think in the 'dual polarity' circuit, C7 should be 1 nF. (Not 100 nF)

Oh, shoot, you're absolutely right. I thought I changed that but it looks like I uploaded the wrong schematic picture. Thanks for the catch.

EDIT: Had a look at the guts of the perfboard prototype and it looks like I did in fact use 10nF in that build.
 
If I'm seeing this right, changing the diode positions on the single is all that's required to generate a  negative bias voltage, should you only need the one bias? 
 
****IMPORTANT****

The "BM800" body I had a couple ears ago was probably not actually a "BM800". I've removed it from the list of compatible bodies. I will see about making a version of the project that fits the list analogguru put in the other thread later this year (unless someone else gets to it before me).

emrr said:
If I'm seeing this right, changing the diode positions on the single is all that's required to generate a  negative bias voltage, should you only need the one bias?

Yes, if you want to generate a negative voltage rather than a positive one, you would reverse the diodes. The diodes only see AC voltage and the direction each pair faces changes whether they rectify to a positive or negative voltages.

May I ask what situation that would pop up in? (Remember you can't draw any current from these if you wanted to use them for some other purpose.)
 
Just a couple observations, from personal experience on my oscillator boards:

1) The pF-range cap values aren't that critical, i used two 22pF on mine and the circuit works just fine :)
2) Despite what it says in the text-annotations in the schematic in the first post, that electrolytic doesn't need to be low-ESR - we're dealing with (at best) single-digit mA figures here ;D
3) Those output "coupling cap" values aren't critical either - think of them as DC-blocking caps in an audio circuit. 1nF is more than enough, but larger values won't matter, or harm anything, so there's no real need to worry about 1nF versus 10nF versus 100nF. If one was mass-production-minded, 100nF everywhere (to remove one BOM item and get bulk discounts) would be the way to go :D

PS: RuudNL Yes, the RC time-constant with 1G/100nF would indeed be 100sec, but that's "only" the time the voltage reached 63%; i believe the widely-accepted(?) figure for it reaching 100% is 5 times the RC time-constant, so 500sec  :)
 

Latest posts

Back
Top