OPEN SOURCE DIY Mic Project - ORS 87 - Stripped Down u87

[Khron]

@Khron - Could you be kind to point out the schematics with values corresponding to references in your PCB?

Or just a list with references <-> values

Ah maybe you are working with the build guide where they will be noted?

Oops, it seems i neglected to place the values on the schematic on the PCBway page... The build guide will still take a while (batch of ~100 boards, for this and 7 other projects, will only be get shipped out next week, and might take 2-3 weeks to arrive).

But I'll update the schematic to show the values too, later today My bad

 

[AdamFrisen]

Anybody knows if there is any local distributor of Takstar CTS-2 in EU or maybe order directly from them? No email reply so far.

797 Audio has shop in Alibaba, are CY002 equally good as Takstar CTS-2?

 

[micolas]

Are CY002 equally good as Takstart CTS-2?

@kingkorg has previously measured, documented very well, the characteristics of the two capsule models, Takstar and 797, even the comparison with Neumann. The shape of the frequency response is correct for both, it falls within the allowed tolerance.
Only the 797 is slightly brighter, so obviously the capacitor in the de-emphasis circuit has a value just a little higher.

 

[Khron]

@Khron - Could you be kind to point out the schematics with values corresponding to references in your PCB?

Or just a list with references <-> values

But I'll update the schematic to show the values too, later today My bad

Schematic updated

 

[micolas]

Schematic updated

C3 and C7 both have 1uF?

 

[Khron]

No, C3's value still needs real-world testing, will probably end up in the 100n or 47n region (that determines the low-cut frequency, roughly).

You're welcome to experiment with that, though

 

[micolas]

No, C3's value still needs real-world testing, will probably end up in the 100n or 47n region (that determines the low-cut frequency, roughly).

You're welcome to experiment with that, though

What frequency will you hi-pass?
I always found around 80Hz more useful than the higher frequency.

 

[Khron]

What frequency will you hi-pass?
I always found around 80Hz more useful than the higher frequency.

I'll have to see what i get with standard-ish (E3 or E6 series) values. I'm not too concerned with any exact / predetermined numbers, and since it's a 1st-order filter, it's not that critical anyway

PS: Maybe i'm doing something wrong in LTspice, but... does the (standard, part of the feedback network) low-cut filter in the U87(A) also drop the overall level of the signal?

 

[micolas]

I'll have to see what i get with standard-ish (E3 or E6 series) values. I'm not too concerned with any exact / predetermined numbers, and since it's a 1st-order filter, it's not that critical anyway

PS: Maybe i'm doing something wrong in LTspice, but... does the (standard, part of the feedback network) low-cut filter in the U87(A) also drop the overall level of the signal?

With C8=1uF C12=100nF C7+C3=1,.. uF, the extension at low frequencies is wider than in the case of U87 'flat' (with hi-pass off).
* Someone in the group said that using the hi-pass filter in U87 does not affect the overall level, only the lows...(but i think this is perceived as a global decrease in volume, the bass takes up a large part of the loudness)
* Many years ago on a forum someone inside Neumann stated that they achieved the specific midrange tone by reducing the bass more. (...?)

 

[Wordsushi]

So , I am new and read this thread and it made me interested to build it.so i got the Hula cow pcb from pcbways and built it. it worked, but partially, it has low sensitivity and narrow frequency response, i followed the hula cow BOM with the exception I used T-13 transformer, Takstar cts-2 capsule,C-5 33nf and CC-6 156pf, C-7 1uf,cc-9 1uf,fet biased at 8.2 volts and 7.3 volts though same output,pot used is 20kOhm, wiring checked for shortage etc, it seem fine, i have connected capsule front side to signal and bacck plat(blue wire) to pol and the voltage between signal nad pol reads -31v Dc, body from aliexpress.I am sharing the result of the built and seek your guidance to fix the problem.last but not the least I would like to thanks all those who are involved in this open source and are very generous to share their knowledge and making an effort and sparing time to contribute.you are all my heroes.Thank you very much indeed.
test audio is attached.looking forward
View attachment 131887View attachment 131888View attachment 131889View attachment 131890View attachment 131891View attachment 131892View attachment 131893View attachment 131894View attachment 131895View attachment 131896View attachment 131897View attachment 131898View attachment 131899View attachment 131900View attachment 131901View attachment 131902View attachment 131903View attachment 131904

Congrats on your build and even more so, congrats on the additionally difficult task of troubleshooting successfully! Enjoy your new mic and welcome to the club.

 

[khalidgul]

Congrats on your build and even more so, congrats on the additionally difficult task of troubleshooting successfully! Enjoy your new mic and welcome to the club.

Thank you very much sir.

 

[untamedfrontier]

@Khron I was looking at your schematic, and notice r4 & r5 are 10k, as opposed to the magical vintage schematic being 2k2 for those spots. Would it still be as critical to have those 10k resistors closely matched, as is talked about for the 2k2's, or is it trivial?

 

[micolas]

No, C3's value still needs real-world testing, will probably end up in the 100n or 47n region (that determines the low-cut frequency, roughly).

You're welcome to experiment with that, though

Now I remembered that I had actually previously tested this Lo-Cut option in the position you preferred in the KSA 87+ design
There was a time when I was testing different transformers with different core sizes, in the u87i only cardioid circuit.
So with 2N3819 CEN, Idss=7.6mA, D1z=33v, transformer NTE10/3, C8=470nF C12=56nF :

C7 between 1.5...2.2uF slightly increased the bass (by a maximum of 10%, subjective of course)
With C7= 3.3...4.7uF the bass did not increase, (probably being limited by the core size of the transformer)
Conversely, when I reduced C7 from 1uF to 820nF, 680nF, 560nF470nF, I noticed substantial reductions in bass (subjective)
What I'm trying to say is that this capacitor C7 behaves according to a different formula in front of the transformer than a capacitor located between the output of the jFET and the base of the BJT (emitter follower). I think that due to the different impedance and the interaction of C7 with the transformer (with which it forms a filter)
That would mean that a combination
C3=68nF/C7=1uF would not be optimal. As you said, it must be experimented practically.
I hope I managed to express myself in English and it makes sense.

 

[micolas]

View attachment 132068
microphone gain on ssl 2 audio interface, aprox 60% dialled in

Now I was able to listen quietly, with my ears rested, to the microphone test.
It sounds very good, credible, natural, with a nice tonal balance, it has spatiality,
I like it.
Congratulations!

 

[Khron]

@Khron I was looking at your schematic, and notice r4 & r5 are 10k, as opposed to the magical vintage schematic being 2k2 for those spots. Would it still be as critical to have those 10k resistors closely matched, as is talked about for the 2k2's, or is it trivial?

You probably want those as closely-matched as you can get them, indeed. The thought behind increasing their value was, on one hand, BOM optimization (10k's are used in other spots too), on the other hand, since the current requirements are so small, no point in having such low values there, and on the other-other hand, those are in parallel with the input impedance of the preamp, which is low to begin with anyway (1-3k or so), so... why not give the circuit the chance to have a bit more headroom before clipping, by needing to work into a less-low impedance overall?

Haven't measured or even simulated anything in that direction, but common sense validates my hunch.

 

[micolas]

You probably want those as closely-matched as you can get them, indeed. The thought behind increasing their value was, on one hand, BOM optimization (10k's are used in other spots too), on the other hand, since the current requirements are so small, no point in having such low values there, and on the other-other hand, those are in parallel with the input impedance of the preamp, which is low to begin with anyway (1-3k or so), so... why not give the circuit the chance to have a bit more headroom before clipping, by needing to work into a less-low impedance overall.

Haven't measured or even simulated anything in that direction, but common sense validates my hunch.

The separate use of R4, R5 with a value greater than 2.2k for powering the audio circuit and R10, R11 for extracting the polarization voltage of the capsule, in addition to reducing the shunting of the transformer (influencing dynamics and bandwidth), also allows increasing the polarization of the capsule by 0.4...0.6V.
It's not much, but it makes up for other losses.
I think it's an elegant approach.

 

[khalidgul]

@Khron @Wordsushi @micolas @OneRoomStudio @RuudNL Bossses please listen to the test sample for my new build and let me know your honest opinion @micolas gave her feedback but it seemed more of a support for my venture in diy, but i am serious and want to learn things in detail, therefore need your critical feedback.So I used ,T13 for output transformer, takstar microphone capsule, FET used is 2sk163.tesst sample is attached.I fed the mic signal through SSL 2, with gain ddialed in aroud 60 %.
View attachment test ORS87.mp3

 

[AdamFrisen]

@Khron @Wordsushi @micolas @OneRoomStudio @RuudNL Bossses please listen to the test sample for my new build and let me know your honest opinion @micolas gave her feedback but it seemed more of a support for my venture in diy, but i am serious and want to learn things in detail, therefore need your critical feedback.So I used ,T13 for output transformer, takstar microphone capsule, FET used is 2sk163.tesst sample is attached.I fed the mic signal through SSL 2, with gain ddialed in aroud 60 %.
View attachment 132102

Why MP3 and not FLAC or WAV? Even if it is 320 kbps.

 

[Wordsushi]

@Khron @Wordsushi @micolas @OneRoomStudio @RuudNL Bossses please listen to the test sample for my new build and let me know your honest opinion @micolas gave her feedback but it seemed more of a support for my venture in diy, but i am serious and want to learn things in detail, therefore need your critical feedback.So I used ,T13 for output transformer, takstar microphone capsule, FET used is 2sk163.tesst sample is attached.I fed the mic signal through SSL 2, with gain ddialed in aroud 60 %.
View attachment 132102

Sounds pretty darned good to me. Sounds like it should. Build seems pretty quiet. Your voice sounds good. I think you can mark this one as a DIY success!

 

[khalidgul]

Sounds pretty darned good to me. Sounds like it should. Build seems pretty quiet. Your voice sounds good. I think you can mark this one as a DIY success!

Thank you.where can I learn more about biasing the fet for mic.I have only a multi meter.Do I need to buy scope ?