Shure KSM9 schematic

[granger.frederic]

Hi
does anybody have a schematic or service manual for the ksm9, or even some measure points...
it need to repair one with a massive gain loss , i suspect a default in the voltages capsules polarisations

best regards
Fred

 

[Khron]

I don't suppose "lifting" / decoding the schematic yourself is an option, is it?

I managed to do that for my KSM141 and KSM27 (which turned out to be surprisingly similar).

 

[granger.frederic]

very difficult because the pcb is multi layered and there's around 100 smd components
could you post yours ?

 

[Khron]

I'd be somewhat surprised if they bothered with anything more than two layers, if i'm honest.

At least one version can be found here: https://groupdiy.com/index.php?topic=61079.msg773749#msg773749

But i think i've drawn up the KSM27 schematic in Eagle as well, i'll have to check...

 

[granger.frederic]

thanks for the link
unfortunately, the ksm9 amp is more complex (switching unit and two capsules...)
it will take me one or two weeks to decode it , i don't have the test points values, and some components have no name  :-\
i'll check the small ceramic caps as you did ...
i doubt it's the capsule...
anyway , quite a bad start  :-\

 

[Khron]

In my case, both mics had some really strong "wind noise", with or without the capsule connected.

 

[granger.frederic]

the PCB has 3 layers...
Same wind noise capsule unplugged , but only high noise floor with capsule.
do you think the problem is  more likely on the amp side or the dc-dc side ?
another weird thing , i can't locate any front jfet, only npn or pnp, thus it's probably inside the capsule assembly.
Or there's no jfet...???
to dismantle it , i think i need a special tool for internal locking nut unscrewing...
any idea to built or replicate the tool ?

 

[Khron]

Well, considering that two out of my two Shure condensers (which happened to share the same circuitry) both had the noise caused by ****** ceramic caps in the amp section... ;D

Granted, i bought them as faulty (got a pretty great deal on'em, too)

If it's anything like the KSM27 / AKG Perception LDC's, i was able to use a pair of big needle-nose pliers to stick into the two "slots" in the locking-nuts on the bottom.

Since the KSM9 body's much thinner, maybe you could try with a beefy pair of tweezers, or something?

 

[granger.frederic]

great , thanks , i'll replace the ceramics
what do you thing about the lack of Jfet ?
do you have an estimation of the standard polarisation voltage for the main capsule ?

 

[Khron]

I found the "full" KSM27 schematic i had lifted (i also "recreated" the PCB, just as a double-check). All those caps between the bases and emitters / collectors of the BJT's were either 68pF or 120pF. As you could read in the thread i linked earlier, i just took the "shotgun approach" and replaced them all with some TDK 120pF - i had ordered about 50 of them, so i had plenty to spare.

If you have access to them, it might be worth measuring the voltages on the connector(?) going to the head-assembly.  It surely MUST have something like a JFET somewhere. I've seen a couple LDC mics where the JFET was on a "daughterboard" along with the high-pass and pad switches, as opposed to the "main signal board".

How many connections are there, over there? I'd expect something in the region of 6-12v going to the drain, a ground pin at least one "audio out" pin, and of course, a / the polarisation voltage for the capsules (or two, in case they have a +/- DC-DC converter).

 

[granger.frederic]

great work ! thanks a lot 
four connectors are going to the capsule in the ksm9
one is grounded , and as i recall there one at 21v and another at 12v (drain) but i'll check the voltages later as i'm not at home
the fourth should be audio out

21v seems low to me for the capsule. 

the capsule assembly has a polar pattern switch (hyper and cardio)
2 capsules in parallel for cardio and opposite phase for hyper
i must try to dismantle
i'll post some pictures if i can access to the capsule's electronics

 

[Khron]

I've read a bit about it over on recordinghacks.com. low or not, it might be worth lifting the schematic of the DC-DC converter - that is, assuming there IS one at all.

There's also a chance the capsule bias voltage is just passively derived from the phantom-power supply, as in cheaper mics (like the sE 2200A, the old one, or the sE X1). In that case, the voltage can easily be "only" in the region of 30v-ish.

That being said, i've revived several laptops where ceramic caps had failed short-circuit. Not sure it's the case here as well, but if one started going leaky (electrically), it could conceivably be dragging the supply voltage down.

But yeah, if at all possible, try to trace the circuit back from that pin you measured the 21v on.

 

[granger.frederic]

thanks Khron for all the advices !
next week , i'll check all the caps, and report back
there's definitely a dc-dc converter
the input dc can be from 11v to 48v
the schematic lifting will be my last option if i can't fix it 
i'll  post it here if i'll do that

 

[Khron]

I might be biased (no pun intended), but i very much do NOT believe in miraculous fixes coming out of nowhere... or out of sheer dumb luck

Like i said, trace back ONLY the bias voltage line / circuit, i assume that should be way simpler than the entire audio circuit (as well).

I don't suppose you've considered sharing some photos of the circuit board, have you?

 

[granger.frederic]

coming from nowhere ???
miraculous ???
i don't thing this applies to me
i'm maybe a bit lazy but there are multiple ways (less time consuming) to test and fix electronic devices , including your way
i will share pictures asap

 

[Khron]

Well, without knowing what the topology looks like, how can you know what the path of that bias voltage is? Or what voltages you should expect along that path...

All i'm saying is, you might not be able to get away without tracing out where that bias voltage comes from (and not necessarily the entire audio circuit)

 

[granger.frederic]

understanding globally in my mind, testing it locally,  represent less efforts than whole writing it accurately : unsoldering all components,tracing the 3 layer pcb,  testing the components, or even guessing for some non marked component values , without knowing if the values are correct (remember that the circuit is out of order)
i add that smds are a pain to rework and ultra fragile !
i've worked on the mic only 1/2 hour , i'm giving me another couple of hours to fix it
first, i don't want to use the brute force , and potentially do more damages,  but if you are an adept , i respect your way to do
maybe at last, but i'm sure i can fix it by my way, like i do all the time ...

 

[Khron]

I didn't remove (or had to) any components from my two mics, but like i said, those were "only" two-layer boards. I have no idea how the inside of the KSM9 looks like, yet.

And you'll only see even-number layers on PCB's

And regarding SMD rework - 5 years ago, when i was troubleshooting a dead LCD TV i bought, i only had my soldering gun, with a loop-tip made out of 1mm diameter copper wire. I used that to remove and resolder various 0603 and 0805 ferrite beads, to isolate where the 3.3v rail was shorted to ground - turned out to be one section of the HDMI receiver chip.

I doubt they would've had much reason to use components smaller than 0603 in a microphone, and just so you know, they're not as fragile as you might think The traces though, if they're thin, and you use too much heat and/or too much force, can be fragile, if you're not careful. Just saying...

 

[granger.frederic]

OK , ksm9 fixed !
3 issues:
-front dual switching diodes smd A2X changed (maybe hot-plugging)
-capsule cleaned with distilled water
-polar switch cleaned

gain is ok and no more noise !

the large electret capsules seem to be very sensitive to moisture in this mic , thus i recommend to put another layer of foam (there is one inside)