AKG C535

[dmp]

I'm trying to fix a AKG c535 microphone that has very low output...
the capsule works fine, I tried it in another mic.
I wrote down the voltages in red on the schematic. It seems the phantom power is dropping way down for some reason, but I don't see any place that is drawing a lot of current??
Also, I don't see how the gate of the FET (q1) could be 0 volts?

 

[pucho812]

how are the electrolytics?   how is the transformer? it's possible you have a faulty part...

 

[dmp]

The transformer is ok. I think that 12v is correct for some condensors, so I started tracing a signal through. I disconnected the capsule and can pass a signal through to the output to my scope when the input signal is at the grid of the FET. Since the capsule works I think the 10G resistor is open so the FET is not biased into conduction. I don't know how to test that theory since a multimeter has such a lower impedance than 10G, can't measure the voltage there. I have no resistors that big.

 

[mad.ax]

The fet's gate may appear to be 0V because you load the 10G too much with your multimeter....

Your test method with the scope looks good. In order to test the 10G resistor, you can make a voltage divider with it, and the highest resistor you have. i.e: 1Meg
Apply 48V,  measure what you get, and do the maths...

Or if you're lazy : http://www.raltron.com/cust/tools/voltage_divider.asp

Axel

 

[abbey road d enfer]

Testing with a multimeter has its limits. If the 10G res was duff, you probably would not measure the correct voltages on the rest of the circuit.
You must trace signal, with a small amp and headphones, check if you have signal when touching base of Q2, then emitter of Q2, base of Q3, then emitter of Q3, then primary of transformer. The culprit is when signal disappears. Signal-tracing 101...
Beware, you must insert a capacitor in your probe, since there is DC on all these points.
I would suspect the lytics first.

 

[dmp]

I stopped tracing with headphones when I had a splitting headache, and switched to a scope  ;D
I think the caps are ok because a signal is passed from the gate of Q1 to the output if the offset is turned up on the signal generator. As I turn down the offset control the signal cuts off as expected (I think) because it drops below the bias of the FET.
If I try to test the 10G resistor with a voltage divider using a 1meg, the output will only be .0001 of the input, and then if the multimeter adds a resistance in parellel, it is even lower...  I think I might just try to find a new one to try switching it.

 

[mad.ax]

Sorry, I used this plan with 1G//1M. Should have done some math before posting... :-X

Anyways, the goal is not to measure 10G with any precision, but to know if it's broken or not, right?

You can use a higher voltage if you have one... Or if you know your voltmeter input resistance, just ditch the 1M...

Say your voltmeter is 20M, with 48V at the input of the 10G, you should read about 96mV... Hopefully, a decent modern DMM should be able to make the difference between 96mV and nothing...

Axel

 

[dmp]

So to test the 10 Gohm resistor... I measured the resistance of my multimeter as 1meg, so I tried 15 volts across the 20 Gohm resistor and the multimeter (in series) so the other leg of the voltage divider is the 1 meg of the multimeter. The multimeter measured ~8mv, so the resistor is not open, and apparently ~2 Gohm actually (if my math is right... 15v*1meg/.008). As I changed the supply voltage from 5-15volts the multimeter changed correctly so it was a definitive test. Thanks for the test idea Axel!
I had myself convinced that the resistor was either bad, or had a leak on capsule side. Both hypothesis disproven. Now I'm stuck as to what's wrong???

 

[pucho812]

if your certain all eletrolytics are good and all the resistors are good and the transformer and capsule work.

then check the trasistors and also check the switch... it's all that's left.

 

[PRR]

It is obviously a "12V inside" mike like many 414a. It works the same on P48 or P24, and slightly lower on P12. I dunno how that capsule can bias; but that's clearly how they made it.

There is 6.1V going into the 10Gig to FET Gate, and 8.4V at FET Source. 2.3V Gate-Source voltage is very reasonable for a JFET. This part works.

You can not direct-connect an audio signal generator to the FET gate.... it needs to sit up near 6V, but your sig-gen against 10Gig will pull the DC down to "0 volts" (millivolts), put FET Source way low, and Q2 will be starved. Here you need a cap. It must be LOW leakage. It must pass audio into 10Gig and about 30pFd. If 0.01uFd works for 1Meg loads, then 1pFd would pass bass, though loaded by the 30pFd(?) capsule. Use 100pFd ceramic or silver-mica.

The "impossible" fact is Q3 5.6V at base and 6.5V at emitter. Even assuming a passive meter (not a VTVM or a DMM), that's not right. Q3 base should be very close to the 6.1V, say 5.8V, and Q3 emitter should be 0.6V lower, near 5.2V.

Device failure is VERY unlikely unless the mike has been plugged into HIGH level line OUTputs. If it is older, the electrolytics 9in and out of Q3) are suspect. Use an AC-coupled listening amp. The sound should be the same on both sides of each cap.

 

[dmp]

Thanks everyone for the suggestions. I learned a little more, but haven't gotten it working.
I used a 100pf mica cap attached to the FET gate and played music through the circuit, into a preamp, and then into headphones. It sounded fine. Then with a signal generator and scope I confirmed that the output signal from the circuit was equal to the input signal attached at the 100pf cap (since the FET is followed by two emitter followers, Vin should = Vout).
It seems every part is confirmed to be working separately, but the mic doesn't work assembled.
I double checked that the AKG capsule works in the other mic (an mxl 603s) and it does. Neither the AKG capsule or MXL capsule work connected to the AKG circuit.
Looking at the MXL circuit, it also appears to be "12V inside" but the capsule sees ~12v on one side and ground on the other. The capsule in the AKG circuit sees 12v on one side and the FET gate voltage on the other, ~6.1v.  It seems it should work as designed... but it's not.
I'm stumped.

 

[Jason Baldock]

Hello There,
I just found this post, and I wanted to ask how to take the C535 apart, as I don't want to break it. My mic works fine in switch positions 2, 3 & 4, but it crackles in the first position (no pad, no HPF). So I'd like to get inside to check things out.
I just need to work out how to get to the circuit board without ruining the mic, haha!
Thanks,
Jason

 

[Khron]

Hello There,
I just found this post, and I wanted to ask how to take the C535 apart, as I don't want to break it. My mic works fine in switch positions 2, 3 & 4, but it crackles in the first position (no pad, no HPF). So I'd like to get inside to check things out.
I just need to work out how to get to the circuit board without ruining the mic, haha!
Thanks,
Jason

Might this help?

https://www.manualslib.com/manual/1109440/Akg-C-535-Eb.html#manual

https://www.manualslib.com/manual/1184132/Akg-C535-Eb.html?page=21

 

[JPO]

It is obviously a "12V inside" mike like many 414a. It works the same on P48 or P24, and slightly lower on P12. I dunno how that capsule can bias; but that's clearly how they made it.

There is 6.1V going into the 10Gig to FET Gate, and 8.4V at FET Source. 2.3V Gate-Source voltage is very reasonable for a JFET. This part works.

You can not direct-connect an audio signal generator to the FET gate.... it needs to sit up near 6V, but your sig-gen against 10Gig will pull the DC down to "0 volts" (millivolts), put FET Source way low, and Q2 will be starved. Here you need a cap. It must be LOW leakage. It must pass audio into 10Gig and about 30pFd. If 0.01uFd works for 1Meg loads, then 1pFd would pass bass, though loaded by the 30pFd(?) capsule. Use 100pFd ceramic or silver-mica.

The "impossible" fact is Q3 5.6V at base and 6.5V at emitter. Even assuming a passive meter (not a VTVM or a DMM), that's not right. Q3 base should be very close to the 6.1V, say 5.8V, and Q3 emitter should be 0.6V lower, near 5.2V.

Device failure is VERY unlikely unless the mike has been plugged into HIGH level line OUTputs. If it is older, the electrolytics 9in and out of Q3) are suspect. Use an AC-coupled listening amp. The sound should be the same on both sides of each cap.

Maybe this will help.