Horrific G7 Problem!

[CurtZHP]

Just fired up the G7.  Hooked it up to a Mackie mixer and monitored through headphones.  Hit the switch on the power supply.  Disappointed by lack of audio, but not really surprised.  Tried turning the pattern switch, got the usual thumps.  Here's where it gets interesting.....

I look down at the Mackie, and I notice that the Phantom Power LED is lit.  Looked at the Phantom Power switch on the back several times to confirm that it is indeed in the OFF position.  ???  The LED must actually read a DC voltage, and obviously that voltage is coming from the mic!

I unplug the mixer, grab the voltmeter, and check the pins on the mic's output XLR.  Here are my findings:

Cardioid pattern = 57VDC
Omni pattern = 0VDC
Figure 8 pattern = 122VDC! 

That's ugly!

Although I checked all the PSU connections when I finished the PSU last year, I checked them again.  Note: these measurements are without the mic connected.

Pin 1 = 0VDC
Pin 2 = 174VDC
Pin 3 = 0VDC
Pin 4 = 0VDC
Pin 5 = 85.5VDC (cardioid), 0VDC (omni), 176.5VDC (fig-8)
Across Pins 6 and 7 = 6.3VDC

Other information:
Capsule: Peluso CEK12  Transformer: OEP A262A3E (windings wired in series.  I was told this helps with capsules that have a top end lift like the CEK12.)

Help!

 

[Mark Burnley]

Hmm,

Check and double check your wiring on the cable- the output is transformer isolated and should have no measurable DC voltage, either between pins 2+3 or between 0V and pin 2 or pin 3.

Most likely you have miswired the interconnect cable, or there is a short within one of the line or chassis connectors.

Check all volatges at connector end that plugs into mic body (without mic being connected) and also meter out (continuity test) pins 2 and 3 of output XLR to mic with power supply switched off, disconnected from the mains, and left to discharge (drain HT caps with e.g. 1W 47k resistor before probing for continuity)
good luck!

Mark

 

[zebra50]

What Mark said - either the cable(s), or the wires inside the PSU from the 7 pin XLR to the 3 pin.
If you can't spot it then post some pix up here and see if any of the eagle-eyes can see where you've gone awry.

 

[Mark Burnley]

Although I checked all the PSU connections when I finished the PSU last year, I checked them again.  Note: these measurements are without the mic connected...

Did you measure these at the connector on the PSU box, or at the end of the cable that would plug into the mic?

As Stewart said, pics may help as well. Sometimes it's something glaringly obvious, but because you've been staring at it for hours, you become blind to things- another pair of eyes is useful.

Mark

 

[zebra50]

I would look carefully at the 7 pin XLR connectors, with a bright light and a magnifying glass if necessary - it would be really easy to have a blob of solder shorting across pins 4 and 5.

 

[Mark Burnley]

Yes, and the other easy fault is when you twist the stranded cable and tin it to land in the XLR pin solder bucket, and one strand escapes and touches the adjacent pin. Good to use Helsyn rubber/silicone sleeving when doing connectors- especially when HV involved!

Mark

 

[CurtZHP]

Gentlemen:

First let me say how much I appreciate your rapid responses to my dilemma.  Nice to know we're all looking out for each other.

I think I may have found the problem, and it wasn't in the PSU or the cable.

Since I wasn't seeing any DC on the XLR output jack when the mic was not connected, I started to suspect the mic.  I also found it intriguing that the voltages at the output seemed to roughly coincide with the pattern voltages.  Although they weren't the same, there was enough similarity to raise suspicion.  (i.e. 0VDC on both when "omni" was selected.)

Upon a very careful examination of the mic PCB, I discovered a very tiny (we're talking human hair sized) trace bridge running from one of the transformer pads (the one for the + output) to the trace leading to Pin 5, the pattern voltage contact.  Just to make sure, I measured it with an ohmmeter and found a short circuit.  A few seconds with a razor blade, and I now measure an open circuit between them.  I can certainly see how introducing upwards of 160VDC to Pin 3 on the mic's XLR connector could provide the results I was getting.

What say you?

 

[MagnetoSound]

Well, putting 160VDC across your secondary is not a great thing to do ...

Is the channel on your Mackie OK?

Just how flukey have you been? 

 

[Gustav]

Upon a very careful examination of the mic PCB, I discovered a very tiny (we're talking human hair sized) trace bridge running from one of the transformer pads (the one for the + output) to the trace leading to Pin 5, the pattern voltage contact.  Just to make sure, I measured it with an ohmmeter and found a short circuit. 

Did you buy this from me or make one yourself? Im really not happy to hear this if its the boards Ive made, since my PCBs are tested for shorts and opens on a probe-tester.

Gustav

 

[CurtZHP]

Rest easy, Gustav.  This was a home brew.  What surprises me is that I checked this thing repeatedly from the time I etched it to the final build.  But then, if you saw the bridge, you'd know why it was easy to miss.  I only spotted it when I backlit the PCB, and even then I wasn't sure of what I was seeing.  That's why I measured it with the meter.

MagnetoSound, I haven't rechecked the channel on the Mackie yet.  I suppose I should, seeing as it belongs to my employer!  ("Honest, Boss, I have no idea what that burning smell is.")

Flukey?  Using a Fluke ahead of time might have avoided such carnage. 

I won't get a chance to try again until tomorrow morning.  I think I will still double check the PSU and the cable, just to be extra sure.

 

[Mark Burnley]

Glad you found the fault,

Hair-bridges are a nightmare- especially between IC pin pads!!

:'(

Mark

 

[CurtZHP]

I just hope I didn't melt the guts out of the transformer!  :-[

 

[CurtZHP]

So I come in the morning, check everything.  No DC on the XLR output.  Good.  Plug everything in, hit the switch and put on the headphones.  Nothing. 

Here's what's interesting.  When I tap on the output transformer, I can hear it in my headphones, but speaking into the mic yields nothing.

???

 

[CurtZHP]

More observations:

I discovered that, while I was getting the pattern voltages into the mic, they were not making it to the back of the capsule.  They were making it as far as the 33M resistor, but no further.  Very close inspection of that resistor showed what appeared to be a hairline crack in its body.  When I gently grabbed it with the needle-nose pliers, it split cleanly in half!  Well, that explains a lot!  So I need to replace that, obviously.

Another interesting tale involving the 100K resistor right before the tube plate.  Upstream of the tube I read about 125VDC.  Good (I think).  On the other side of it, I read nothing.  I removed it from the circuit and measured its resistance just to make sure it was intact.  Putting it back in (which should have certainly taken care of any cold solder joints if there were any), I get the same result.  No voltage to the plate.

No wonder this thing doesn't work!

Is the problem with the 100K resistor, or with the tube? 

I'll replace the 100k anyway.  Resistors are cheap.  Tubes, not so much.

 

[zebra50]

Hi!

Don't expect to be able to read the voltages in the Hi-Z region (i.e. on the capsule side of the 30 meg and 1 gig resistors) accurately with standard multi meter. The impedence is too high (or, put another way, there are insufficient electrons passing through around to get a decent reading).

 

[CurtZHP]

Hi!

Don't expect to be able to read the voltages in the Hi-Z region (i.e. on the capsule side of the 30 meg and 1 gig resistors) accurately with standard multi meter. The impedence is too high (or, put another way, there are insufficient electrons passing through around to get a decent reading).

Well, there wasn't much point in measuring the capsule side of the 33Meg, since it was essentially an open circuit.  I wasn't measuring anything near the 1G resistors.  But there's definitely no plate voltage in this thing, which is why I was suspecting the 100K resistor.

The plans call for the plate voltage to be around 160VDC.  I was reading around 120VDC right before the 100K resistor.  Is this within tolerance?  What other tolerances should I be aware of (i.e. the 80VDC between the two 470K resistors, or the pattern voltages?)

 

[zebra50]

The plans call for the plate voltage to be around 160VDC.  I was reading around 120VDC right before the 100K resistor.  Is this within tolerance? 

Sounds about right, maybe a little low. It's some time since I built mine, but here were my measured voltages..

http://www.xaudia.com/omnip/Other/G7mics/Comments.html

 

[CurtZHP]

Hopefully, once I replace the offending resistors, I'll actually see some plate voltage at the tube and get something out of the capsule.

Zebra, on your diagram, you have red "x"s on the connections to the capsule.  I assume this is to denote that no meaningful voltage measurements can be made at those points.

 

[CurtZHP]

Well, I replaced the bad resistors.  Still no dice. 

What baffles me is the fact that I read no voltage at the tube plate.  None.  The plate supply voltage gets as far as the 10K resistor and that's it.  When I measure at the point where the 10K connects to the 100K, I see at least 120VDC.  When I measure the other side of the 100K (which goes straight to the tube plate) I get nothing.  Where did the 120VDC go???  Looking at the schematic and the PCB, it appears that the 100K resistor is also connected to G2 on the tube, and to the 2u2 capacitor between the tube and the output transformer.  But there is ZERO VOLTAGE!!

If the tube was shot, could that be the cause?  I would think I'd still see 120VDC at that pin anyway.

This isn't making any sense!

 

[Gus]

120VDC at one side of the 100K plate resistor and 0VDC at the other measured referenced to ground?  It reads like the plate is connected to ground.