MXL 990 to KM84-ish circuit with RK47

[Khron]

- Why is C7 where it is, and not closer to R5? No need to shoehorn even a single trace between its pins, nevermind two
- Move R5 closer towards the middle of the board, or even flip it 180deg, and move C7 to the edge of the board, now that there's room there
- Why is the trace connecting R3 and R4 not a straight line?
- Speaking of which, swap their positions between them
- If you use a shorter footprint for D1, you'll gain back some board space that's otherwise wasted (maybe even move one or more electrolytics there, so as not to have so many traces snaking over to the other side of the board)
- Flip the trimmer around (rotate by 180deg)
- And since you connected its two ground pins together, there's no longer need for the separate trace connecting them

 

[bdamiano]

Very nice. I've attached some pics of my boards in a 990 and a BM800 flavor (I also have ST55 and TLM102 footprints). The board are from Osh Park. I don't use a pot for R3 but instead have some test points where I clip in a precision pot to set bias to minimum THD. Then, select a resistor that matches the pot (very closely). In the photos, the FET and 1G are not yet soldered to the turret. I save that until we commit to wanting the optional pattern switch or not. If so, I add the second turret and switch. I've been using a Solen "fast cap" for the 1uF. They say "France" on them so they must sound good! Also using the CineMag 5722w but I have made some cool builds with CK12 style capsules with the Tab Funkenwerk T12 too.

 

[SparkleBear]

OK! So I think i've made a " @Khron -proof" revision here ;P LOL...

Put a 50v output cap and optimized all placements and routings as much as possible.

Lemme know what you think ...

 

[Khron]

- The track from C5's positive leg to R4 still cuts a bit close to the "top" leg of R7.
JFET source to trimpot track unnecessarily close to the drain pad - any particular reason?
- And why is the track coming off the drain on the bottom of the board, shoehorned in-between the two grounded pins of the trimpot?
- Insisting on an axial footprint for the output cap? 1u/50v film caps should be easily found in radial 7x5mm or so footprint, the box-shaped kind.
- And i guess you don't want to not-need one of the 1G resistors and the input capacitor...

Other than that, much improved indeed

Especially since this is no high-speed-digital stuff, or even high-power stuff, you ARE allowed to, say, T-junction off a track, not just connect traces at a component pin. Thinking of that drain connection through the trimpot pins...

(Tiny punctuation edit for clarity)

 

[SparkleBear]

And i guess you don't want to not-need one of the 1G resistors and the input capacitor...

And since this is not an SDC, and thus one side of the capsule is not mandatorily connected to ground (through the case), you could go with the "ChinaMicMod" wiring (from micbuilders / Zapnspark), and have the capsule connected between the bias voltage and the JFET gate. That then rids you of one of the 1g resistors, as well as the C1 DC-blocking cap. But that then requires swapping the two output wires going to the XLR as well, to maintain the same polarity.

RE: ChinaMicMod. Does this modification to the schematic look like what you were explaining?
Does it matter if I polarize the diaphragm or the backplate? I am planning on polarizing the backplate and taking the signal from the diaphragm.




I don't mind knocking out a $6 resistor from the design ;P And takes the DC blocking cap out of the signal path...

Let me know if that looks right to you.

 

[Khron]

Better again Although you might still want to come down vertically first, off the positive pad of C5, and then "kink out" to the side. As is, C5's pads look unnecessarily close to being shorted together.

Might also wanna move C6 up a bit, to leave a bit more room for the mounting screw head / spacer there.

And yes, that's the capsule wiring i meant. Although you'll want to triple-check the final polarity of the signal. It's tricky for me to get my head around it, but there's a good chance the polarity will be reversed, with the capsule bias wiring change. But that would only require swapping the two wires going to the output XLR.

- Insisting on an axial footprint for the output cap? 1u/50v film caps should be easily found in radial 7x5mm or so footprint, the box-shaped kind.

 

[SparkleBear]

Better again Although you might still want to come down vertically first, off the positive pad of C5, and then "kink out" to the side. As is, C5's pads look unnecessarily close to being shorted together.

Might also wanna move C6 up a bit, to leave a bit more room for the mounting screw head / spacer there.

done....

And yes, that's the capsule wiring i meant. Although you'll want to triple-check the final polarity of the signal. It's tricky for me to get my head around it, but there's a good chance the polarity will be reversed, with the capsule bias wiring change. But that would only require swapping the two wires going to the output XLR.

I swapped it on the schematic and layout... but ill test it too.

swapped in radial cap... And changed out the zip tie holes for pads to solder solid core wire to secure transformer.

 

[Matador]

The 2N3819 pin out is DGS, but your layout appears to be GDS. Which FET are you using?

 

[SparkleBear]

The 2N3819 pin out is DGS, but your layout appears to be GDS. Which FET are you using?

Oh, when I first designed board I was using J113 from onsemi. But thats a good catch Matador. I am actually using the 2n3819 now so I will swap that around. THANKS SO MUCH!



There we go SGD!

 

[Khron]

If you're going with the rewired capsule bias, you'll only need one standoff, for the gate / R1 / C2 junction.

 

[SparkleBear]

If you're going with the rewired capsule bias, you'll only need one standoff, for the gate / R1 / C2 junction.

I don't understand -- backplate will get bias at c2/r1 junction and diaphragm will go to gate. Don't those need to be separate and lifted?

 

[Khron]

Ah, my earlier approval was a bit of a rushed brainfart

The other 1G resistor gets removed (JFET gate still needs to get pulled to ground), and the biased side of the capsule gets tied directly to the bias voltage, no 1G needed there. C2 can go straight between the drain and gate.

https://audioimprov.com/AudioImprov/Mics/Entries/2015/12/20_Modding_a_BM-800_Mic.html
Like in the second schematic there, but ignoring the switch and rear diaphragm biasing. There's no 1G gate resistor there because the JFET used has a pair of internal antiparallel diodes that virtually act as that.

 

[SparkleBear]

Ah, my earlier approval was a bit of a rushed brainfart

The other 1G resistor gets removed (JFET gate still needs to get pulled to ground), and the biased side of the capsule gets tied directly to the bias voltage, no 1G needed there. C2 can go straight between the drain and gate.

Like in the second schematic there, but ignoring the switch and rear diaphragm biasing. There's no 1G gate resistor there because the JFET used has a pair of internal antiparallel diodes that virtually act as that.

So there is no need for any 1g resistors?? Or R12 in the schema below is taking the same function as the 1G polarization bias resistor on the way to the capsule.

And JFET gate gets pulled to ground through the 1G that I still have present? Or did you mean the drain gets pulled to ground?



I thought I had a handle on it but now I'm a bit confused.. lol

 

[Khron]

Or R12 in the schema below is taking the same function as the 1G polarization bias resistor on the way to the capsule.

Yes. But note there's still a 100nF capacitor to ground, so the capsule connection is, from an AC point of view, "at ground potential". So no need to isolate it through a very-high-value resistor anymore.

And JFET gate gets pulled to ground through the 1G that I still have present?

Yes, if you connect that correctly (from the JFET gate to ground).

 

[solowiej]

I'm new here, but excited to hear how this project is coming along if it's still in the works.