BM800 Pimped Alice + DC-DC, increase output level

[Jakub W]

Hey everyone!

I drafted the pimped Alice board, based on the article from here https://www.instructables.com/Modify-a-cheap-LDC-Condenser-microphone/.

Then I saw the design in the PDF was extended in this PDF: https://content.instructables.com/F6F/8GN2/IJHFDO44/F6F8GN2IJHFDO44.pdf

In the firs circuit I used a stock capsule from Rode NTK and the output is loud, but really bright (even though I used a de-emphasis filter).

The 1u capacitors parallel to the 100n capacitors are for base extension, which helped especially in the 2'nd circuit.

In the second one I used a K67 style capsule from an Avantone CK-6 I had from a previous mod. and here the tonality is really good.

In both cases a 6n8 cap was used parallel to the drain resistor and both were put together on a perf board.

I like the second attempt better, but the sensitivity is rather low, so the noise floor is high, when I crank it up.

I assume, it may be to the DC-DC converter not providing enough voltage.

I have seen, that both the GenChinaMod and Pimped Alice v 56 don't use a Zener in the drain circuit. Would that help?

I also added a 1.5k resistor between the converter and the audio board. Maybe I should ditch that too?

Looking forward to any insights.

 

[ccaudle]

That schematic shows pin 2 of the XLR connected to shell, which you definitely do not want.

it may be to the DC-DC converter not providing enough voltage.

What voltage are you getting? It looks like the oscillator has a 9V zener, and a voltage doubler on the output. It is really just 18V or a little less? I'm surprised the Rode capsule was so much more sensitive with that capsule bias, Rode usually has between 60V and 70V bias I think.
That DC-DC converter circuit looks a little strange, but might just be the way it is drawn. Have you verified it is operating properly?

 

[Jakub W]

It looks like the oscillator has a 9V zener, and a voltage doubler on the output. It is really just 18V or a little less?

So I either need to remove it or put in a 30V Zener?

I'm surprised the Rode capsule was so much more sensitive with that capsule bias, Rode usually has between 60V and 70V bias I think.

I actually used a different DC-DC circuit there, copied directly from an Audioimprov article, just without the -VBIAS part:

That DC-DC converter circuit looks a little strange, but might just be the way it is drawn. Have you verified it is operating properly?

It would be much quieter without the converter. Also - Is the drain part of the circuit appropriate? As I wrote:

I have seen, that both the GenChinaMod and Pimped Alice v 56 don't use a Zener in the drain circuit.

 

[rogs]

I've used a variation of the same Hartley oscillator configuration (which comes from Schoeps CMC5 schematic origianlly, I believe?) in the latest version of my OPIC project .
Page 5 of this document: https://www.jp137.com/lts/OPIC.TX.pdf shows the DC volatge multiplier schematic.
Fitted with a 12v zener, it allows for a DC voltage within the range range of 55v to 90v to be selected ........

 

[Jakub W]

I've used a variation of the same Hartley oscillator configuration (which comes from Schoeps CMC5 schematic origianlly, I believe?) in the latest version of my OPIC project .
Page 5 of this document: https://www.jp137.com/lts/OPIC.TX.pdf shows the DC volatge multiplier schematic.
Fitted with a 12v zener, it allows for a DC voltage within the range range of 55v to 90v to be selected ........

As far as I can see, it’s connected directly to the xlr con, mine is connected to the 100u cap to the drain circuit. And as I’ve seen in the GenChinaMod, they have no zener there.
https://medias.audiofanzine.com/files/genchinamod-472635.pdf

 

[Jakub W]

What voltage are you getting?

In both circuits I get about 60v after the .1u cap (C5 in my schematic), but after the 1.5M (R1) resistor it drops to about 25V, but this may be a limitation of the multimeter probes.

 

[Jakub W]

I've used a variation of the same Hartley oscillator configuration (which comes from Schoeps CMC5 schematic origianlly, I believe?) in the latest version of my OPIC project .
Page 5 of this document: https://www.jp137.com/lts/OPIC.TX.pdf shows the DC volatge multiplier schematic.
Fitted with a 12v zener, it allows for a DC voltage within the range range of 55v to 90v to be selected ........

OK, given the fact, that I get 60V before R1, I have one question VR1 is a trimpot in the schematic, as I assume, which helps you get up to 90V, which maybe would be a solution in my case... Hmmmm...

 

[ccaudle]

So I either need to remove it or put in a 30V Zener?

Apparently I was way off in my guess of what is going on through the inductors. I will have to think about the voltages around the LC tank circuit, but rogs showed you can get up to 95V with just 12V input, so the 9V zener should be no problem.

I get about 60v after the .1u cap (C5 in my schematic), but after the 1.5M (R1) resistor it drops to about 25V

Do you mean it is 60V at C6, but drops to 25V at C5? That is probably because the input resistance of your voltmeter loads down the output.

Is the drain part of the circuit appropriate?

Do you mean the collector of Q1? You have connected the zener to the gate, and I think typically a zener would be used to set the collector voltage for that circuit.
Or did you mean the drain of Q2?

helps you get up to 90V, which maybe would be a solution in my case...

Many capsules will collapse the diaphragm to the backplate at 90V, the traditional designs are made for around 60V capsule bias.
If you have 60V bias it is not the cause of the low output, something else is different.

Parasitic capacitance and resistance in parallel with the capsule will attenuate the signal, so perhaps you need to clean really the attachment point from the capsule wires.

 

[rogs]

OK, given the fact, that I get 60V before R1, I have one question VR1 is a trimpot in the schematic, as I assume, which helps you get up to 90V, which maybe would be a solution in my case... Hmmmm...

The 12 v Zener regulates the supply derived from the phantom power. With VR1 set fully CW (type 3362P pot) you would expect to measure a sine wave of around 95V (p- p) at the junction of L1 and C4. The D1/D2 voltage doubler turns that into c.95v DC, which you need to measure at the D1 cathode for a more accurate reading from a DMM with an input impdeance of 10M.
With VR wiper rotated fully in the other direction, you will measure c.55v at D1 cathode. The actual DC supply to the oscillator will have been dropped to around 7.5v, measured at Q1 collector.