DIY 48V phantom mic humming

[reubenm]

Check the backplate grounding cap. Measure from pin 1, all along the ground plane of your PCB right up to the capsule. You are looking for a break in the ground plane or a bad solder connection around the capsule area.

have checked ground trace,no break of connectivity

 

[HellfireStudios]

Do you have pics of the mic? It would help to know the mic's physical design when helping to alleviate the grounding issue. For instance, if the mic is a LDC, you would need to check the continuity between the mesh grill and the XLR ground pin.
You could also test you Behringer preamp with a known-working mic.

-James-

 

[reubenm]

Do you have pics of the mic? It would help to know the mic's physical design when helping to alleviate the grounding issue. For instance, if the mic is a LDC, you would need to check the continuity between the mesh grill and the XLR ground pin.
You could also test you Behringer preamp with a known-working mic.

-James-

thanks James. the mic is an LDC,i dont have a camera at this point maybe later.will probably bring a friends condenser and try.

i just uncoupled the coupling capacitor from the amp[in case it was leaking and causing feedback] and there was no hum/rumble whatsoever even at high gain.therefore,i've narrowed everything to the amplifier.any help is much appreciated.

 

[HellfireStudios]

While the amplifier .jpeg indicates a standard TO-92c package for the FET, the pdf supplied shows two unrelated packages. You could have the pins for the FET wired incorrectly. If your FET is of the TO-92c ilk, then that package pinout would need to be verified to confirm it is hooked up right in the mic.

-James-

 

[reubenm]

While the amplifier .jpeg indicates a standard TO-92c package for the FET, the pdf supplied shows two unrelated packages. You could have the pins for the FET wired incorrectly. If your FET is of the TO-92c ilk, then that package pinout would need to be verified to confirm it is hooked up right in the mic.

-James-

sorry about my pictures . the package is TO-72 ;4 pin like siliconix pdf says. i'm positive:the drain,gate and source are wired correctly- as in the schematic.the FET ground is grounded. reuben

 

[reubenm]

i'm running the drain at 35V ;after the resistor Rd. is that bad?

 

[MagnetoSound]

... you would need to check the continuity between the mesh grill and the XLR ground pin.

This could be it. The bit that sticks in my mind is that you said:

... when i grip the ground of the XLR connector [bottom, mic] ,the hum attenuates,and the awesome tone of the mic in its fullness and happy can be heard through my monitors.

This is the big cleu: somehow when you grip the connector you are completing a ground connection somewhere in the mic - do you happen to be grasping the headgrille at the same time as you are doing this?

 

[zebra50]

Certainly grounding, as the guys above have suggested. Bad cable?

 

[reubenm]

This is the big cleu: somehow when you grip the connector you are completing a ground connection somewhere in the mic - do you happen to be grasping the headgrille at the same time as you are doing this?

thanks for replying. there's no headgrille on at the moment, i grasped only the XLR female connector.the hum subsides but is not gone. iwas looking at the Nady scm900 circuit, which goes right! but mine doesnt.most condensers are simple and do not cause hums.do no whats wrong here.
  can my amplifier not be grounded? i thot it wouldn't work if that were the case since Rs be disconnected.i am positive it is grounded. i last night 'isolated' the drain neumann U47 style; a 30k before the 100k, capacitor in between to ground-solved nothing.
  this mic has been buzzing/humming for one year now.
      could the peluso CEK12 be leaking dc?i tried the circuit without the capsule still hums.
  when i touch the exposed jacks on the Behringer Xenyx ,the hum subsides by a 1/3 ???

 

[zebra50]

thanks for replying. there's no headgrille on at the moment,

!!!! 

You need something to screen the high impedance section, otherwise you will just pickup electrostatic HUMM!
Put a metal case around the whole thing and then test. Even if it is just a makeshift body in tinfoil.

 

[reubenm]

ok man, i will do that now.

 

[reubenm]

1] i caged the capsule and mount with a steel headbasket, (hardwired) to mic ground

2] i encapsulated all pcb electronics in stainless steel tube (hardwired) to mic ground

there was a 1/20th reduction but otherwise no reduction in hum..

i might have to test the mixer with another condenser(dynamic mic=fine) to check if it has bad phantom power.

 

[Rossi]

What you're attempting to do (perhaps without knowing) is build a KM84 circuit. But you're leaving out essential parts such as the *two* additional filter stages (56k and 10k + caps to ground) of the FET supply voltage. Have a look at the original schematic:

http://www.groupdiy.com/index.php?action=dlattach;topic=44805.0;attach=8094

Also, your FET voltage runs too high. You should have about 10V at the source, not 35V. Remember what you see is the static voltage. In real use, the source voltage goes up and down as it amplifies the gate voltage (i.e. the signal).

Your FET may be the wrong type. I've used a 4416 only once, and IIRC it wasn't suited for a KM84 type circuit. Get some 2N3819 FETs, they're cheap, and even a non-selected, non-optimal one will at least work.

So, put in an extra resistor, ca. 62k, plus an additional cap, 10u or more, for the FET supply voltage (that's one filter stage). Reduce Rd to 47k, replace FET with 3819, replace Rs with a 10k pot and adjust for Vd = 10V. The Source bypass cap should be 4.7u or more. Increase Rg to at least 60M. The usual value is 1 gigaohms (=1000M).

Keep in mind that high impedances are extremely hum sensitive; your mic must be well shielded. Also: work clean, remove any grease and dirt etc from the board and its components with 100% alcohol.

You need a suitable transformer. Not any old 10:1 will do. Primary inductance must be at least around 100H (i.e. 1H secondary inductance).

Hope that helps. If your run into trouble: have another look at the KM84 circuit - and don't try to be smarter than Neumann.

 

[reubenm]

What you're attempting to do (perhaps without knowing) is build a KM84 circuit. But you're leaving out essential parts such as the *two* additional filter stages (56k and 10k + caps to ground) of the FET supply voltage. Have a look at the original schematic:

http://www.groupdiy.com/index.php?action=dlattach;topic=44805.0;attach=8094

Also, your FET voltage runs too high. You should have about 10V at the source, not 35V. Remember what you see is the static voltage. In real use, the source voltage goes up and down as it amplifies the gate voltage (i.e. the signal).

Your FET may be the wrong type. I've used a 4416 only once, and IIRC it wasn't suited for a KM84 type circuit. Get some 2N3819 FETs, they're cheap, and even a non-selected, non-optimal one will at least work.

So, put in an extra resistor, ca. 62k, plus an additional cap, 10u or more, for the FET supply voltage (that's one filter stage). Reduce Rd to 47k, replace FET with 3819, replace Rs with a 10k pot and adjust for Vd = 10V. The Source bypass cap should be 4.7u or more. Increase Rg to at least 60M. The usual value is 1 gigaohms (=1000M).

Keep in mind that high impedances are extremely hum sensitive; your mic must be well shielded. Also: work clean, remove any grease and dirt etc from the board and its components with 100% alcohol.

You need a suitable transformer. Not any old 10:1 will do. Primary inductance must be at least around 100H (i.e. 1H secondary inductance).

Hope that helps. If your run into trouble: have another look at the KM84 circuit - and don't try to be smarter than Neumann.

Hi Rossi.
    dude i hate to disappoint, every aspect of your post is valuable,just that i cannot tolerate the grounded capsule arrangement of the KM83-85...the tone is not 'there'.ive tried that setup.
    my transformer has a primary inductance of 25H...should be enough..seems there's no lack of punch.the coupling coefficient is 99.7% / 0.997.
    i cannot reduce Rd.. 2N3819 has too high input capacitance.[i might try it though]
But you got me thinking about the fet voltage..it may be too high.so far so good except the hum. i will refer back to your post in coming days for reference.thanks for the reply.
BTW,can i ask your opinion what is causing the hum?

 

[MagnetoSound]

i was basing my circuit on the ELAM 251 circuit and U47.

A FET is not a tube.

 

[MagnetoSound]

OK, there are only two possible ways for hum to be picked up in the mic:

1) Bad ground or screening;

2) Power supply ripple (poor filtering), direct coupling of hum into amp circuit.

my phantom powered diy mic with a 77pF capsule makes a 0dbV loud 100Hz hum at minor mixer input gain ...

Are you able to verify whether the hum is in fact ~100Hz or ~50Hz? This will be the biggest indicator as to whether the hum is power related, or a screening problem.

Since you can shut off the FET (by removing Rs) and the hum goes away, it does seem possible that poor filtering may be causing power supply ripple to be entering the FET amp via the capsule polarising network.

Can you please post a schematic with all values shown? It looks like your capsule grounding cap is only 2200pF. This is in fact a crucial filter cap, and needs to be at least 100nF in value to keep polarisation voltage adequately free of ripple.

when i touch the exposed jacks on the Behringer Xenyx ,the hum subsides by a 1/3

I do think there might be a grounding issue as well. Another important, as yet unasked question: is the Xenyx itself well grounded?

 

[Rossi]

Hi Rossi.
    dude i hate to disappoint, every aspect of your post is valuable,just that i cannot tolerate the grounded capsule arrangement of the KM83-85...the tone is not 'there'.ive tried that setup.
    my transformer has a primary inductance of 25H...should be enough..seems there's no lack of punch.the coupling coefficient is 99.7% / 0.997.
    i cannot reduce Rd to 47k.. 2N3819 has too high input capacitance.[i might try it though]
But you got me thinking about the fet voltage..it may be too high.i was basing my circuit on the ELAM 251 circuit and U47.so far so good except the hum. i will refer back to your post in coming days for reference.thanks for the reply.
BTW,can i ask your opinion what is causing the hum?

Well, its seems you already "know" everything.

My advice is simple: build a KM84 type circuit without any modifications (it IS a really good sounding circuit!). Get the basics right, and only then begin to tinker with it. Making two steps at the same time is how beginners projects usually fail. As I said, don't try to be smarter than Neumann. They knew their stuff. Actually, they still do. I know, you think I'm conservative, but the fact is that I've built a fair number of mics, most of which with my self-designed circuits, and in the process I've come to realize that pretty much anything that you see in an established circuit* is there for a reason. If you can't see that reason, it's usually not the circuit's fault but your own. For instance, depending on the housing/mounting, the capsule often cannot operate ungrounded.

In a sense, your hum comes from trying to be too clever. Be humble. Learn. MagnetoSound gave you some valuable hints.

* "established" does not include Chinese circuits.

 

[reubenm]

2) Power supply ripple (poor filtering), direct coupling of hum into amp circuit.




Are you able to verify whether the hum is in fact ~100Hz or ~50Hz? This will be the biggest indicator as to whether the hum is power related, or a screening problem.

when i touch the exposed jacks on the Behringer Xenyx ,the hum subsides by a 1/3

I do think there might be a grounding issue as well. Another important, as yet unasked question: is the Xenyx itself well grounded?

i'll be looking at filter networks i guess

the hum sounds 100-120Hz ish

the Behringer Xenyx 802 doesnt ever hum or buzz with dynamic mic or electric guitar,so the grounding should be ok.


i might have to restart and build filter networks recently with the goal of taking care of any hum,then see whats what later.

 

[reubenm]

MagnetoSound
Jakob E
James @HellfireStudios
Rossi

, thanks for the replies. have my work cut out for me.cheers,reuben