AKG P414 - C414E preamp + U87 style DC converter strange voltages - any ideas?

[ln76d]

Hey Guys!

Simple project (or maybe not so simple?)!

I put inside AKG P100 original preamp circuit board from C414E (COMB).
P100 use almost the same DC converter as U87Ai.
Simplified schematic in attachment.

First strange thing (maybe it's normal?) is that am getting only 33V from the preamp board to feed DC converter.
What cause that voltage drop? These two transistors suck current that much? Faulty transformer?
At the output of converter am getting only 44V. 
Any ideas?

 

[Khron]

You could try reducing the value of the R20 and/or R21 in the schematic, down to 33-30-27-22k - that should bump up the output voltage

I know i had to tweak those values in my pair of BM600's (which out of the factory, had a U87A-like dc-dc converter).

Regarding the input voltage, 15v difference (48-33) divided by 3.4k (6.8k//6.8k) gives about 4.4mA. Not terribly out of the ordinary, i guess.

I've picked up some schematics of Apex 480 (jfet+pnp) from somewhere online some time ago, with notes of about 23v going into the dc-dc converter.

 

[ln76d]

Thanks Khron!

Unfortunately R20 is already reduced - i updated schematic.
Connection is different also.
Could you share this schematic?
It's quite strange...
I have that DC converter from apex

Full C414 circuit should suck ca. 3mA.
It would mean that original DC converter doesn't draw the current at all or still there's some fault on preamp PCB.
Right now there's let say that 3mA.

Transformer is ok - checked!

So maybe it suck that much

Funny is that i even can't simulate DC converter circuit - it always show at the output 0,5V lower value than the input voltage.

 

[Khron]

You've reduced R20, but R21 is 2.1Meg - that's gonna "choke" the actual input voltage to the dc-dc converter. If you move the components around a bit, R20-21 and C22-24 are just a two-stage RC low-pass.

In addition to that, there might be a chance you'll need to add some voltage to that zener there - i don't have those mics here at home, but i think that's where i added two more 5.6v zeners in series with the existing one (no real idea what voltage that was at), but i know i ended up with about 70-75v at the output (for the Advanced Audio AK12 capsules i used in them).

The dc-dc converter in my BM600's didn't have a transformer "per se"; the magnetics comprised only of two axial 220uH inductors

See the last paragraph in the first post in my thread about these: http://groupdiy.com/index.php?topic=61084.msg773828#msg773828

 

[Khron]

Come to think of it... i kinda wonder just how (drastically?) different the sound of this impedance converter is, from the more widely-used common-drain/PNP configuration. If at all...

 

[ln76d]

I will try to juggle with resistors values and also i will buy few different Zener diodes - starting from 60V.
I don't know what zener sits here almost all the parts are SMD.
Inductors are also 2x220uH and from what i remember Neumann use double inductor with same values.

DC converters have no sound

 

[Khron]

I said IMPEDANCE CONVERTER ie. the jfet & bjt through which the audio signal passes, on its way to the transformer

 

[ln76d]

Sorry!  Today  mind is completely in the different dimension

Difference between what?

Original perception vs C414E?

 

[Khron]

More like in general, between the C414E and the widely-used jfet-pnp combo

 

[ln76d]

Ok i  get it right now

There's always some differnces, even those really small

C414 transformer is really good I compared it to the P100 and yes, both have similar ratio and really small, but sound different!

Currently am using in this microphone chinese "CK12" - "Made in USA" by Audix
With 44V microphone sound sweet except loud and annoying noise.
It's test only right now, but the noise is gone when you touch some components with multimeter probe

 

[Khron]

Some loose grounding somewhere, perhaps?

 

[ln76d]

I was thinking the same, but this is not typical hum and i checked all ground connections. Grounding is ok.
It sounds much more like some freaky oscillation in DC converter.

 

[Khron]

I've played around with some dc-dc converters in ltspice, and i've noticed that in certain circumstances, some can go into some "hiccup" or skipping mode, with bursts happening at audio-frequency intervals. It might be interesting to whip out a scope and see what's going on

 

[ricardo]

To measure the correct output voltage of this type of DC converter, you need an Electrometer ... something with MUCH higher input Z than the usual DVM or scope.

These circuits are not meant to deliver much current.  Even the resistor chain for changing pattern can be a problem.  If you take too much current, they can go into a "squeeging" mode which is very noisy.  I think this is what Khron has seen.

If you are using a converter from a working 414, you are probably getting the correct voltage out.

Also many circuit simulators have problems simulating this type of DC convertor.  You have to kick start.  Try putting a transient spike into T10 gate.

IIRC, the LTspice simulator has a page with special caveats on simulating oscillators.  How are you modelling Dr1, the transformer in the oscillator?

 

[ln76d]

Thanks both of you Guys for help!

With 10M ohm input always is option to measure before or even after 1M resistor. It can drop few volts only like 4-5V.
Voltage from C414 preamp is correct i checked two other preamps.
For simulation am using two 220uH inductors like in the circuit i have. This should be also correct for the original Neumann circuit.
Am using multisim and i didn't found any way to proper work.
How that converter should work?
Is it boosting any voltage to the 60V or only multiplying the voltage?
If multiplying it will have sense 46V - 60V, then 33V - 44V. It will be 1,3-1,4 multiplier.
From the other hand - i found some information that someone measured original perception circuit and get 44V...
Maybe the zener is too small?

 

[Khron]

Well, as i mentioned, i DID have to add a couple zeners in series with the existing diode, to bump up the output voltage

 

[ln76d]

Yes i know
Like always i really appreciate your help.
I guess that it has some similarities to the schoeps converter, where most depends on one zener diode.
It doesn't change the fact that i still don't understand how this converter should work and my question was related to that
I like to know how the things work not only blindly changing parts.
Also, please remember, that this converter is made with smd mostly so i can't freely changing parts for tests because the boards and soldering points have their limitations in durability

 

[ln76d]

Anyone knows how this converter should work?
Is it multiplier or it boost any (with some margin of course) voltage to the specified?

 

[Khron]

I really don't know what problems you guys had with simulating this circuit - it worked on the first attempt, for me.

See the screenshot

White trace = the node where L2 connects to D1/C4/C5
Blue trace = the node between D2 and R6 (the "load")

Yes, i know, there's no filter cap on the output - i was lazy ;D

Quick edit: Adding an output cap just makes the output rise slower. With 1nF, the blue trace gets to about 38v after 10ms. With 100nF, it only reaches about 1.5v after 10ms

And note, that's a plain old 1N4148, not a zener over there

 

[ln76d]

Try with multisim
I tried to change everything in the circuit and always is the same.
I can't use ltspice unfortunately.
You didn't use zener
What if you will use zener 62V?