Gefell conversion confusion

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

peter purpose

Well-known member
Joined
Jun 3, 2004
Messages
2,082
Location
London
Gentlemen,
I have finally got around to converting my Gefell MV692s from 12 to 48V, but have run into a problem.

The conversion .......

gus692.jpg


...... works flawlessly when out of circuit, but when I put it in the mic I'm getting 0.5v..????

Here is where it goes (removing the highlighted stuff), with 48v taken from the centre tap.


gef1.jpg



Can anyone see any glaringly obvious mistakes?

Cheers
peter
 
> Can anyone see any glaringly obvious mistakes?

How did you arrive at R1=100K?

I would expect to need at least 2mA at the power feed. Dropping 48V to 12V, that's 36V/2mA= 18K max. There is 3K4 in the Phantom, so 15K is probably the maximum for R1+R2. However P12 can actually draw more current than P48, so it just may not be possible. Assuming we might want 10mA, 36V/10mA= 3K6, minus the 3K4 in P48, R1+R2 is 200 ohms max!

36V/106K= you only have 0.34mA available now. I bet the HV converter is stalled.

At Your Own Risk: Try R1= 22K. See if the voltage rises over a volt. If so, go fairly fearlessly through 15K toward 3K (you need to reduce R2 also) until you find a good >11V at the mike (or something smokes!). If that "works", decrease the resistor even more to increase current 25%-50% (watching that the voltage does not rise past 13V), so the Zener has work to do and you won't be at the mercy of Phantom tolerances. (This still won't work with sub-48V "P48" unless you trim for the specific Phantom source.)

A more universal alternate is to use a JFET with Idss>10Ma, add a source resistor and short the gate to that, makes a constant current limiter. Trim it to 2mA and see if it feeds the mike, sneak up until the mike is happy, and sneak up another 10%-20% to cover drift and slop.
 
Actually, what I see:

Put a 11.2V-11.7V Zener across C9. This clamps the power rail. Note that the P12 is dropped in D1/2 so the internal working voltage is more like eleven and a half volts. This is slightly critical because the bias voltage is multiplied-up from the rail.

Then yank D1/2 and put in a 10K to 1K resistor or the aforementioned FET strapped for 2mA to 10mA.

If there is space, I might bump C9 to 10uFd or 100uFd to reject crappy P48 juice.

If I were curious, I'd tack a 10 ohm resistor in series with the Zener so I could watch the current in it. Up to 11V it will be zero. As you decrease the resistor at the D1/2 location, or increase the JFET current, the Zener will start conducting. At ~1mA in the Zener (10mV in the 10 ohms) I'd compute the total current and subtract the Zener current to get the mike current. I'd like to end up with the Zener passing more than 2mA (so it is well up on its knee) but total current somewhat under the 10mA max of modern P48. Note that older P48 is 2mA max, and many "Phantom" jacks are not at all a solid 48V +/-4V through 3K4.... wall-wart mike-amps often sag to 30V real fast, so maybe just enough to feed the mike plus 2mA for the Zener.
 
That kind of looks like the circuit in two of my 692s. I purchased them used converted for 48V. What is missing is a small surface mount PNP?

IIRC

The 100K goes to the base from the 48V
the collector connects to 48V
The emitter goes to C1

I guess the added .6V diode drop is for the B to E drop of the pnp

I converted another 692 by removing d1/2 c12? and installing a 12K
I also removed D1/1. I had the microhone open and connected to good 48V phantom supply to pick the resistor. This only works well when the phantom is 48V.

The other pass transistor zener reg can handle a wide range of input voltages
 
Thank you very much indeed gentlemen.

Gus......
It looks familiar, as it was the drawing you sent me.
I simulated with the extra PNP, but that didn't work, so I changed it for an NPN and all seemed hunky dory. Added the extra component to the circuit for real and nothing...???

PRR.....
I went with your second suggestion. Changed C9 to 10uF and strapped a zener accross it. 10k for the other components and wound up with a nice 11.7V.
Re-built the mic and it works... only now I have some serious hiss. Might try swapping the fet for a 2SK170 and see what happens.

Cheers
peter
 
Yes when I looked at my microphones I missed the added smd transistor. For an pnp switch the emitter and collector 48 to the emitter and out from the collector.
 
Just speculating here, but I would think D1 and D2 would have to go from base to ground, and the emmiter and base should not be shorted together.

Peace,
Al.
 
> I just simulated that and got 48V the other side.

What value load (microphone) are you using?

If the answer is "duh?", try both 2K and 6K. Sure, no-load will give 48V, or Zener voltage, but the whole point is to power a load.

The plan posted with 100K into a base-emitter-shorted BJT is essentially equivalent to a resistor of 100K/β plus a diode, where β is the current gain of the transistor, typically 50 to 500. This is not a great idea. The diode does no good and the transistor is just confusing things.

There is a suggestion to implement a current-limiter with BJT, diodes and resistors. It is possible, but a lot of parts, and the "great" (bad) thing about a BJT is that it will pass infinite current (or pass an Ampere for long enough to blow-up everything downstream). A modest size JFET can limit current in fewer parts, and it is a lot harder to randomly mis-connect it for infinite current.

> now I have some serious hiss.

Before messing with external power, try it on a battery. A stack of eight AA cells should run it for hours. Put 100uFd across the battery to swamp battery self-noise. That should be dead-quiet; if not, I would suspect bad capacitors long before I would replace the FET (which was probably fine when new and probably has not been blown if it is passing signal at all).
 
Thanks Paul,

I'm going to stick with your zener across C9 scenario, as it works.

I'll replace some caps and fiddle with the value of the 1k to 10k resistor to fine tune it and report back.

Many thanks

peter
 
Let me try this again c1 and to the right leave as drawn

NPN

48V to the collector and one side of the 100k

other side of the 100k to the base

The base to the top of the zener and diode the other side of the diode string to ground

The emitter to C1(no connection between the emitter and base)

The 100k gives the diodes current to turn on and the junction is connected to the base of the NPN

Think of the circuit as a NPN emitter follower the emitter "tracks" the base with about a .6V offset. The emitter load is R2 and the microphone circuit

That is the circuit in two of the 692s I bought.

If you know you will only use a good 48V phantom supply the resistor 48V mod works well.
 
> 48V to the collector and one side of the 100k ... other side of the 100k to the base ... The base to the top of the zener and diode the other side of the diode string to ground ... The emitter to C1(no connection between the emitter and base) ...

Gefell-drop.gif
?
 
PostPosted: Fri Dec 16, 2005 2:44 am Post subject:
I thought the MV692 was a valve microphone / amplifier?

Why does it have a 7-pin connector?

No, 690, 691 and 692 are transistor-mics. The Valve-version is the MV 582.

And for the 7-pin-connector: Look at the schematics above. Only 4 pins are used.

Chris
 
I have a 48v mv692 and schematics. It looks like you have schematics though. I could take some pics and check the schematic against yours if you want.
 
Pics are always welcome Charlie..... I've built most of my gear by copying pictures.... but as you can see, I've learned nothing..!!

Cheers
peter
 
Back
Top