Groupbuy capsules on BM700/800 microphones. Opinions and suggestions appreciated

[ln76d]

Hi Jon, thanks for the answers.
About the high pass filter cap, I just thought the same thing. I just wanted to be sure the 4u7 was the "stock"  value for that circuit. I didn't want either more or less bass, but I'm sure I wouldn't even notice the difference, if not in particular circumstances, as you said.

If 4.7uF, try MKT capacitor - Siemens/Epcos MKT should fit circuit without any problem. Any radial, small sized MKT.
If higher, try Electrolytic 10uF - 22uF Elna Silmic II.
10uF is reasonable value.

About the biasing method, I'm not sure I understood it completely.
What I understood....I have to test the complete circuit (with transformer and 75pF cap instead of the capsule), connect the circuit through a pre and turn the phantom power on, put the black probe of a multimeter to ground hole and the red probe to the drain hole near the fet, turn the variable resistor screw till I get 11,5V, is everything right until this point? Could I just leave it like this? I mean, I know it isn't a perfect bias, but at least is in the right working range, isn't it?

You need to connect capacitor (don't have to be 500V, 50V-100V is sufficient as also 50pF-100pF range) between electrical signal source (from generator - virtual or equipment) and JFET Gate. Circuit need to be properly grounded and completed.
Don't need to connect of polarisation voltage node.
Setting voltge around 11.5V should be sufficient.

What I didn't undestood...where do I feed the audio signal into the circuit? And how to adjust it by ear? Am I assuming to understand an exact spot where it can be considered biased? Also, as I want to make a "stereo set", can I replicate the bias by ear for the two mics?

How to feed signal, see above.
Setting bias by the ear - apply 1kHz into the input, during adjusting potentiometer in the circuit you will hear change of the level, so your 1kHz will be louder or quieter. During whole operation you need to find  the point where amplitude is highest - it's easy - signal will be increasing during adjusting and after the highest level it will drop very fast. Unfortunately you will have to repeat few times whole operation to catch that point

Another method - same as above but use any VST spectrum analyzer to see where is the higher level during pot adjustement or make complete microphone with capsule installed, and use speaker as signal source for the capsule.
Minus of the speaker as a signal source is that you will have to mount the PCB from the other side than the front of the capsule, to get possibility of potentiometer adjustement.

One last question, about the transformer, I took the tiny neutrik nte10/3 suggested from In76d. I attached the pic for schematic. Are the dots plus, minus, or is the same as long as I refer to them equally?
Still interested if anyone can suggest me if it's worth removing the inner mesh layer from the headbasket.
Thanks

For 10:1 ratio (JB schematic for reference):
RED to C3
BLACK to ground
WHITE to R10
YELLOW to R9

I'm not sure is it  properly working as 7:1 (6.x:1), i think i measured it this way but don't remember well.
For now i have no option to test that.
You can try to use switch for test!
Wire switch (SPDT type ON-ON) between  RED and BLUE and the output of the switch to the C3.
Now you can test two possibilties 10:1 and ca. 7:1.
With 7:1 microphone output level will be higher and if works properly then frequency response will be similar to 10:1.
You can hear what output level is better for your needs.
This operation is independent from biasing etc.

For headbasket - usual one layer sounds better from my experience.
If you setup bias and complete whole microphone circuit with capsule installed, unscrew headbasket.
Do voice test - grab microphone (with whole body metal parts attached and option to take off headbasket) with one hand for the metal base, and with another hand keep headbasket (rather don't cover front of the capsule ). Compare between headbasket installed and without it. Without headbasket you can have problem with hum, so find the place in your listening room where there's lowest problem with hum (it's strongly dependent from your listening gear, amps preamps, psu, wall AC voltage etc.)
You can also try to cover a little the backside of the diaphragm (don't touch capsule) with a hand (second hand always keep metal pipe of the mic body). If you have variable HPF in your preamp, you can use it.
Difference which you should notice is mostly in the upper midrange and high frequency area.
Weak headbasket will usual overbright or make more nasal sounding microphone.
If you find that difference then after removing internal mesh should be better response, more neutral.

 

[Rasta3]

First of all, thanks for the replies.
About the high pass filter cap, I installed the 4.7u. If I'll ever notice lack of bass, I may try the 22u, otherwise I'll be ok with the 4.7u.  I bought the mouser's BOM, so I already own both caps.
I wired the trafo for 7:1, I measured the resistance between the secondary wires (used as primary in this circuit) and should work. If I have problems, I'll wire for 10:1.
For the headbasket, I already read someone else saying better results removing the inner fine mesh, so probably I'll just take them out, without even testing it. I don't think I may have any problems with that, don't I?
For the FET bias...I understood it's better to do the oscilloscope method, because the other methods are not equally precise.
So I'm trying to find someone with a Oscope near me, otherwise I'll just buy one of that tiny DSO138, easily available online and try to follow the procedure myself.
So to the point, how do I connect the capsule test capacitor? Am I right saying I connect the 75pF in series from the signal (from generator or DAW) to the FET gate? While the ground of the signal to the ground of the pcb?
Should I disconnect the FET gate from the turret with the 1Gohm resistor and 4pF cap, for this measure?
For the Oscope, the probes go to drain and ground holes (on poctop pcb)?
I think I got the symmetrical clip thing, just turn the trimmer until I get the waveform clipping simoultaneously on both sides.
Thank you

 

[Rasta3]

Also, can I bias the FET with the trafo connected?
And am I right connecting the capsule test cap the way I described or should just replace the capsule with that cap?
Sorry, but still a bit confused. 

 

[ln76d]

First of all, thanks for the replies.
About the high pass filter cap, I installed the 4.7u. If I'll ever notice lack of bass, I may try the 22u, otherwise I'll be ok with the 4.7u.  I bought the mouser's BOM, so I already own both caps.
I wired the trafo for 7:1, I measured the resistance between the secondary wires (used as primary in this circuit) and should work. If I have problems, I'll wire for 10:1.
For the headbasket, I already read someone else saying better results removing the inner fine mesh, so probably I'll just take them out, without even testing it. I don't think I may have any problems with that, don't I?
For the FET bias...I understood it's better to do the oscilloscope method, because the other methods are not equally precise.
So I'm trying to find someone with a Oscope near me, otherwise I'll just buy one of that tiny DSO138, easily available online and try to follow the procedure myself.
So to the point, how do I connect the capsule test capacitor? Am I right saying I connect the 75pF in series from the signal (from generator or DAW) to the FET gate? While the ground of the signal to the ground of the pcb?
Should I disconnect the FET gate from the turret with the 1Gohm resistor and 4pF cap, for this measure?
For the Oscope, the probes go to drain and ground holes (on poctop pcb)?
I think I got the symmetrical clip thing, just turn the trimmer until I get the waveform clipping simoultaneously on both sides.
Thank you

With headbasket in chinese microphones rather rarely you will find one which have more than one layer of mesh and sounds better or just even equal to the headbasket without it.
If you are making cardioid only microphone, then try to remove only layer from the front of headbasket.
If internal mesh is one piece, then you can try to cut it on a half and replace it from the back side.

Yes, you are correct about connection cap and source of the signal. In series.
Despite the capsule and polarisation voltage node, everything inside microphone should be installed.
Better way to connect oscope is ground to ground and take signal after output capacitor (before transformer).
Then you are avoiding DC voltage on the input of the scope.

 

[Khron]

A possibly easier way to bias it would be with an audio interface and your computer

Either with your DAW of choice (they all have, or should have, some signal generator and some form of spectrum analyzer), or with something like Visual Analyser.

http://www.sillanumsoft.org/download.htm

A virtual oscilloscope, signal generator... Aw hell, there's an overview page there http://www.sillanumsoft.org/prod01.htm

Better way to connect oscope is ground to ground and take signal after output capacitor (before transformer).
Then you are avoiding DC voltage on the input of the scope.

Most (if not all) half-decent scopes DO have an AC-coupling option  : Although probably not those crap pocket-usb things...

 

[ln76d]

A possibly easier way to bias it would be with an audio interface and your computer

Either with your DAW of choice (they all have, or should have, some signal generator and some form of spectrum analyzer), or with something like Visual Analyser.

http://www.sillanumsoft.org/download.htm

A virtual oscilloscope, signal generator... Aw hell, there's an overview page there http://www.sillanumsoft.org/prod01.htm

Any daw and two vst plugins would be sufficient, anyway in this case the higher sampling frequency the better visible result. 96kHz looks pretty ok. Instead scope signal can be taken  after microphone preamp.

Most (if not all) half-decent scopes DO have an AC-coupling option  : Although probably not those crap pocket-usb things...

You know, i know, but we both don't know what would be using any person which read that

 

[Rasta3]

Hi there, I already checked for AC coupling option on those tiny DSOs and it should be there
Anyway I probably found a friend with a Oscope, it seemed almost impossible 
Otherwise I'll try the DAW vst method.
One last thing I didn't understand....what's the polarization voltage node? What actually should I disconnect from the circuit?
Sorry for the dumb questions, but I'm almost done, you won't hear from me anymore
Thanks again

 

[ln76d]

One last thing I didn't understand....what's the polarization voltage node? What actually should I disconnect from the circuit?

It's the part of the circuit which distribute polarisation voltage (ca. 46V) to the capsule and exactly backplate of the capsule.
It's just wire which goes from R2 1Gohm resistor to the capsule. Other wire from the capsule goes from diaphragm to the fet gate.
More precisely - not disconnect, but don't connect R2 to the capacitor for biasing.

Sorry for the dumb questions, but I'm almost done, you won't hear from me anymore
Thanks again

Better ask than make mistakes
At least tell us how it sounds, how do you like it etc.

 

[Rasta3]

Better ask than make mistakes
At least tell us how it sounds, how do you like it etc.

That's exactly my philosophy 
I will, I will, don't worry.
Thanks again

 

[Rasta3]

No way to find someone with a scope, at least in 100 kms from me
So I decided to go with the software scope. I already had Visual Analyzer installed and after confirmation about it from Khron, I decided to give it a try.
At first I had some troubles, I tried to pass through the mixer with the signal, with no result, very very low output from direct out on the channel, so I went with the signal directly connected to the pcb and used the wave volume from the windows mixer to adjust volume.
Another problem I had, it was the coupling cap (75pF) before the FET introduced a 50Hz hum into the circuit, and I had bad moments tryng to trigger the 1KHz signal, with the 50Hz signal summed.
After removing the cap everything seemed to work fine, I took some time to adjust volumes and such but after a while I had a clear sine wave at almost 1KHz, fair enough
The last step was trying to catch the point were both peaks started to clip simoultaneously. I can't say that has been easy, it wasn't. After a bit of frustration I tried to catch the sweet spot, helping me out with the lowest THD and I ended up with one mic measuring 8.4V and the other 8.3V.
I measured even the resistance from the trimmer and it's about 7.4k, I didn't desolder the trimmers from the PCB cause I'm not planning to replace it with fixed resistor.
I attach some pics from the scope, the first one is from the sine wave without clipping, the second with the moment both peaks seamed to clip and the third one with both peaks clipped.
Can I assume I've done everything right? Am I ready to mount to capsule, from wath you see?
Thanks

 

[Rasta3]

Sine wave without clip

 

[Rasta3]

Sine wave with small clip

 

[Rasta3]

Sine wave fully clipped

 

[ln76d]

Weird that you had hum i newer had problem with that during biasing.
Measure your phantom voltage with microphone connected - you can be surprised sometimes
If you have problem with biasing try to change 1kHz level - with to low and to high sometimes is hard to setup properly.

 

[Rasta3]

Yes, it's possible my mixer doesn't have the best phantom power out there   However, do you think the measurement could be fine or they are way too far from the 11.5V spot? Also is the test cap necessary for the bias or can I assume it just worked out, even without that cap? And last, can I mount the capsules and try the mics out or should I redo the bias? Thanks a lot for your support, In76d.

 

[ln76d]

Sorry, i missed your post
If you have lower phantom voltage than +48V then your optimal voltage for proper biasing point also would be lower.
Trust your measurements with virtual scope
Cap isn't necessary.
Definately install capsules and check microphones. Check definately is there similar sound level with both.

 

[Rasta3]

Thank you very much In76d, I definately will try the mics out, very soon (I hope).
Very busy days for now, but you'll here my thoughts as soon as I try them.
Thanks again.

 

[Rasta3]

Hi there, I finally got some news.
I tried the mics out, they sound awesome.
The 2 Fet847 circuits with RK12 capsules sound very smooth and creamy, even with the voices.
While the stock circuits modded with the K47 capsules sound a bit more crispy and high freq, but for the right sources I think it'll help a lot in coming out of the mix.
That said I just have a couple of problems,  the first is very subtle, I noticed a little difference in the CK12s mics, it seems like one of them has less high frequencies than the other. As I said it is very subtle, not accentuated at all, so I tried to open up the mic, turn on the phantom power and measured (with the capsule mounted and without the signal as in the biasing method) the voltage as I did while biasing the FET.
What I noticed is that the first mic (the one with slightly more high freq) starts about 8,2V and after a couple of seconds drops at about 7,8/7,9V.
While the other starts at almost the same 8,2V but continues dropping down the voltage, it's not just for a couple of seconds, I waited some minutes and went all the way down to 6,8V.
What can cause that? Could it be a problem? Did you ever noticed something like that?
The other problem in the other mics (the K47 ones) is my fault, I think I used some on-on-on switches for the polarity  ( or maybe just melted them while soldering) because they always make contact. The right contacts on the sides and both contacts in the middle, so I get nice sounding mics in omni and fig 8 polarity and very bad and low volume sounding while in cardioid polarity. i'll check and swap the switches out.
Anyway the Fet847 have lot less volume than the stock circuits, but, as I understand, that should be fine.
Thanks for any help.

 

[Rasta3]

This are the othe ones.