Shure KSM27 / KSM141 circuitry

About a year or two ago, i grabbed a KSM27 and a KSM141 from the same seller on eBay, both sold as faulty. The 27 pretty damn noisy, and the 141 with very low output (and a bit on the noisy side).

I must say, i WAS quite surprised by the amount of circuitry crammed into these things. I had lifted the schematic for the 27 sometime last year, and for the 141 these past couple of weeks, and just the other day i realized that they're virtually the same. The only differences are the existence of a DC-DC converter on the back side of the KSM27 board (needed to bias the LDC), and a 3-position pad (vs. 2-pos) on the KSM141 (small-diaphragm electret).

I went through all this "masochistic torment" because my hunch was that the noise was due to the electronics - the capsules look quite ok, and there's no visible mechanical damage on the mic bodies.

Just last night, for the hell of it, i (manually) "copied" the schematic into LTspice, to see if i could somehow make sense of what's going on in there. Let's say that most of the circuitry makes a certain amount of sense, but i'm still utterly baffled by the JFET stage.

I'm attaching a screenshot of the LTspice schematic - if needed / requested, i can  also attach the schematic file.

Anyone out there / around here feeling... intellectually "brave" or adventurous?

Are you certain you've transcribed the circuit correctly?

It looks like a "normal" FET mic input with a constant current source in the source leg, but also somehow connected to the drain leg. All sorts of feedback going on and some EQ stuff.

So, err, yes, somewhat baffling!

Here's some info from shure:

"
In early 2006, a small quantity of KSM27 microphones were manufactured that had one or more faulty 120pF capacitors on the internal circuit board. This problem was not found in final testing because the noise problem only appeared after phantom power has been applied to the mic for a period of time, typically more than 5 minutes.

Unfortunately, the faulty capacitor(s) was not always in the same position on the board and thus finding the root cause was difficult. This capacitor value is used at the following part postions on the KSM27 schematic: C3, C4, C8, C9, C10, C17, C18, C21, C22, C26, C28.

By May 2006, the problem had been resolved. If you have a Shure KSM27 mic that exhibits this problem, please contact Shure Service at 800-516-2525 about having the mic repaired.

Answer ID 3387
"

Check that first, next step could be swapping capsule to the different circuit.

Well, i'd be kinda worried if i had done the same "reading" mistakes when lifting the schematics for both mics, about a year apart, so... i'd say odds are it's quite correct. Hell, even the component designators are the same between the two mics I edited the part names to reflect exactly what's on the boards.

The "group" between R18(10meg) and R12 (15k) is the 3-position (on-off-on switch) high-pass filter - R18 shorted out = flat; R18 not shorted out = -6dB @ 115Hz; R21 shorted out = -18dB @ 80Hz.

Which exactly would be the constant current source, in the source leg? For what it's worth, the JFET's drain is up, and the source is down.

And yes, i had indeed come across that Shure article when i was investigating my 27. I've got a bunch of 120pF caps already, and at the time, i removed all the indicated caps from the 27 board, but i don't recall there being any change. Then again, i guess i could stick in the new ones and see what happens

Yesterday i also removed nearly all the components from the 141 board, to be able to check them all out-of-circuit, and they all measured as expected. But yes, SMD ceramics can and do go bad for no apparent reason - that's how i've revived several laptops in the past

Khron said:

Which exactly would be the constant current source, in the source leg? For what it's worth, the JFET's drain is up, and the source is down.

Click to expand...

Q6 and Q9. The only DC path to ground in the source leg.

That looks more like a CFP / complementary darlington arrangement to me, and it's "mirrored" by Q3 & Q8. As i see / understand it, via the HPF-section, it drives Q1 (top right) which goes to XLR pin 2. The similar arrangement going to Q2 and subsequently pin 3 might be, as far as i can figure, for impedance matching / balancing.

... But i COULD be wrong

Well, i've got ONE piece of good news - i got miffed, and since i had a piece of taped 120pF caps, i decided "screw it, i'll replace all the 120p AND 68p caps".

And lo and behold, the bastard (ie. KSM141) WORKS fine, and is now as quiet as... well... nothing

Do exactly same thing with ksm27 and probably it will work also
Get some C0G/NP0 caps for it!

I just did, but it's starting to seem like the capsule's screwed, unfortunately.

I had bought 50 TDK 120pF/50v 5% C0G caps, just for this purpose. Ah well, still, one out of two ain't (that) bad

Thanks everyone for the ideas / motivation!

Check the capsule in the different circuit

I was looking at some fets, and got intrigued by ifnd89 used in nt3 and this guy., used in ksm44 as well. Were you aware of diodes and resistor in this JFet?

Yes, i was - i recall looking up and finding the datasheet for those JFETs. I guess that saved them a good few cents on the 1G resistors.

I think some similar ones are used in the cheapo BM800 mics (2SK596 or something like that).