512 Audio Skylight

[Amled87]

So, I picked up two of these for 70$ off eBay, not expecting much. But, after looking over them they have a K47 style capsule, use decent caps (polystyrene, some Rubycon electrolytics and film box caps). The polystyrene caps are 1000pF and 100p, the two larger electrolytics in the center are 100uf/25v.

I looked over the boards, but the actual design inspiration doesn't "jump" out to me. The Jfet is a 2sk30A, there is a c9012 directly in front of it. The back board has (3) 2n2222's, a trimmer and etc. Do these look inspired by anything to anyone? They're very smooth compared to Chinese k67 equipped mics!

I'm assuming schoeps style, but I'm not very familiar with schoeps style schematics and you usually see 4 transistors in those, and this has 5. Or possibly a Jfet/Emitter follower and the 2n2222's are part of dc to dc board? Not familiar here, ang insight?

 

[Amled87]

@Khron Have you had a chance to look at this yet?

 

[Khron]

@Khron Have you had a chance to look at this yet?

Yes and no - that flash glare partially blinded me...

Any chance of some glare-less shots?

10000% positive all three transistors on the oscillator board are 2N2222's?

Ideally, a look at the bottom of the boards would be nice too, but i guess that would be a hassle, what with then being soldered together...

 

[Amled87]

Yes and no - that flash glare partially blinded me...

Any chance of some glare-less shots?

10000% positive all three transistors on the oscillator board are 2N2222's?

Ideally, a look at the bottom of the boards would be nice too, but i guess that would be a hassle, what with then being soldered together...

@Khron

I was pleasantly suprised to find the boards aren't soldered together, there is a connector between them so I was able to get some shots of the backs. The 3 transistors on the "oscillator" board as you called it are labeled PN2222A, PN2222A and P2222A. The two on the main board is a k30a jfet and a s9012 pnp back to back. I also got some images of the capsule, mount and assembly. Also, the mic does have a transformer on the output.

 

[Khron]

That transistor at the edge of the oscillator board seems to be diode-connected (resulting in a Zener-like diode - emitter & base connected together, collector goes to R16).

Q4 & Q5 emitters are connected together on the bottom side; nothing's really visible on the top side. Any chance they're connected in parallel?

Or actually...

https://www.electronics-tutorials.ws/waveforms/astable.html

And the output of that goes to that little step-up transformer? And the trimmer sets the input supply voltage to the whole thing?

 

[Amled87]

That transistor at the edge of the oscillator board seems to be diode-connected (resulting in a Zener-like diode - emitter & base connected together, collector goes to R16).

Q4 & Q5 emitters are connected together on the bottom side; nothing's really visible on the top side. Any chance they're connected in parallel?

Or actually...

https://www.electronics-tutorials.ws/waveforms/astable.html

And the output of that goes to that little step-up transformer? And the trimmer sets the input supply voltage to the whole thing?

The link you provided looks similar to what's going on there. Seems the outputs of Q4 and Q5 transistors go to 1 and 3 of the little step-up transformer. I haven't checked the trimmer (but that seems a logical conclusion, I'll have to get some readings in a few to see what it is adjusting to be sure).

The main board is confusing to me as well. It appears the two 2k2 resistors attached to the center tap (phantom power) has the voltage connected to one side instead of at the junction of the two 2k2 resistors as I've seen in most condenser mics. It appears the amplifier is also a source follower (the coupling cap between the jfet and pnp is a 4.7uF and attached to the source of the jfet). These are from following traces and etc, but the only capacitors attached to the source and or drain is the 4.7uF attached to the source. The 56k (R5 identified by green, violet, black, red and brown) also has a trace on the top of the board which goes to the 100uF about midway down the board on the same side. Also, R3 (5.6k) connects to the 4.7uF cap and then to the base of Q2 (the s9012). The film cap on the main board (6n8) connects to the 4.7uF cap also (traces are connected) but for the life of me can't figure out where it goes (Ground maybe?) R7 appears to be a 10Meg also?

Whole thing is confusing to me to be honest. Definitely a lot more going on then I'm used to!

 

[Khron]

Whole thing is confusing to me to be honest. Definitely a lot more going on then I'm used to!

That's why, when I'm reverse engineering something, i recreate the schematic AND the layout / routing in parallel, for easier double-checking of connections.

 

[Amled87]

That's why, when I'm reverse engineering something, i recreate the schematic AND the layout / routing in parallel, for easier double-checking of connections.

That's what I'm doing my friend (I've took notes from some of things you've done here). It's just being a little more difficult because it's not the familiar jfet to transformer that i'm used to. There's a lot more going on here and it's taking me a lot longer as some of the things don't make sense, like for example the p48 being applied to one end of the the 2k2 phantom resistors instead of the middle junction of the two. Any time I've seen 2k2 resistors and p48, the voltage (center tap) was always applied to the junction of the 2k2 resistors, with one sife supplying the voltage and the other side grounded. That doesn't appear to the case here.

 

[Khron]

the p48 being applied to one end of the the 2k2 phantom resistors instead of the middle junction of the two.

That virtually NEVER happens.

You ARE aware that the phantom power voltage gets applied to BOTH pins 2 AND 3 in the XLR connector, relative to ground (pin 1), right?

If anything, the mic circuitry gets supplied FROM the middle junction of the (most often) 2.2k resistors.

Any time I've seen 2k2 resistors and p48, the voltage (center tap) was always applied to the junction of the 2k2 resistors, with one sife supplying the voltage and the other side grounded.

I'd love to see some schematic examples of that...

As a bonus, draw those up in LTspice and see if/how they work

 

[Amled87]

That virtually NEVER happens.

You ARE aware that the phantom power voltage gets applied to BOTH pins 2 AND 3 in the XLR connector, relative to ground (pin 1), right?

If anything, the mic circuitry gets supplied FROM the middle junction of the (most often) 2.2k resistors.



I'd love to see some schematic examples of that...

As a bonus, draw those up in LTspice and see if/how they work

To be quite honest, I overlooked alot here and I'm slowly getting it worked out. Pins 2 & 3 from the xlr do connect to the 2k2's how you mentioned (I was looking at it wrong) which then has R12 (1k on the board) feed the rest of the power supply. The only real issue I'm having currently is figuring where the R1 1G resistor (its the one connected to the capsules red wire and the 1nF cap) connects to the voltage supply. I am currently at work, and don't have LTspice available, but I'll draw up a hand drawn schematic with the correct part designations shortly.

Edit :
I'll be running this through LTspice when I'm home in a few, but this where I've got. I'll be adding the 2sk30A JFET source follower stage followed by the pnp transistor shortly as well as soon as I have them worked out. Does what I have seem correct @Khron ? I know the values are correct as I've checked them both by color code and meter.

 

[Khron]

Does what I have seem correct @Khron ?

Any reason those diodes are pointing the way they are? Hint: the stripe on one end of the diodes marks the cathode - the line at the tip of the triangle in the symbol (the anode is the base of the triangle).

 

[Amled87]

Any reason those diodes are pointing the way they are? Hint: the stripe on one end of the diodes marks the cathode - the line at the tip of the triangle in the symbol (the anode is the base of the triangle).

I had them drawn backwards, however, as I'm attempting to get a sim working I've removed them from the schematic as they're not important in the sim. However, getting this thing correct is definitely a little challenging. R3 is the source resistor of Q1 (it is connected to the jfet as I have continuity there) however, I get no continuity from it or the 47uF capacitor that's attached to it and ground?

Edit : Found continuity by touching the 1G that goes to ground.

 

[Khron]

I've removed them from the schematic as they're not important in the sim.

... Unless they're Zener diodes and are there to cap the voltage at those points (which they most likely are)...

 

[Amled87]

... Unless they're Zener diodes and are there to cap the voltage at those points (which they most likely are)...

These are the correct places for each component. If I remove the PNP section, I have roughly 22db of gain on the Jfet section, when connected to the PNP section my output is roughly 10db (that is measured from the input side of the 60pf capsule cap to the out of the c12 47uF output cap?). I've got something wrong here as 10db of gain seems way off?

I'm also going to work on incorporating th power supply the way it's supposed to be.

 

[Amled87]

... Unless they're Zener diodes and are there to cap the voltage at those points (which they most likely are)...

So... This thing is basically another Nova/Sterling Audio/GT with a K47 style capsule? Minus the different component choices obviously. It differs from the Nova being a Jfet/pnp instead of a jfet/npn like Nova though.

 

[Khron]

These are the correct places for each component. If I remove the PNP section, I have roughly 22db of gain on the Jfet section, when connected to the PNP section my output is roughly 10db (that is measured from the input side of the 60pf capsule cap to the out of the c12 47uF output cap?). I've got something wrong here as 10db of gain seems way off?

I'm also going to work on incorporating th power supply the way it's supposed to be.

PNP's upside down (consider the power supply polarity, and look at the emitter arrow).

Was it you that i told here, or someone else, to add a 100k resistor after the output capacitor, to represent the preamp input impedance "seen through" a transformer?

R1 1G does nothing there in the simulation, just FYI.

Right-click on the V1 voltage source, and on the right side of the window that pops up, in the "AC amplitude" box, write 1. Then you can probe whatever point you want in the signal path, and you'll get the "absolute dB" (so-to-speak) level of the signal at the respective point.

And you'll want at the very least a 4.5k resistor in series with the 48V supply (since that's fed through a pair of 6.8k resistors in parallel - ie. 3.4k - AND in series with that, a pair of 2.2k resistors in parallel - ie. another 1.1k)

So... This thing is basically another Nova/Sterling Audio/GT with a K47 style capsule?

Why wouldn't it be?

a jfet/npn like Nova

How do you figure that?

https://groupdiy.com/threads/maudio-nova-circuit-mod.78827/

 

[Amled87]

PNP's upside down (consider the power supply polarity, and look at the emitter arrow).

I corrected the PNP and now have 22dB at the output. (I have a 1c AC amplitude entered for the V1 input, but for whatever reason it doesn't show up on the screen, but it does work).

Was it you that i told here, or someone else, to add a 100k resistor after the output capacitor, to represent the preamp input impedance "seen through" a transformer?

It was me. I added that as well. Thank you!

R1 1G does nothing there in the simulation, just FYI.

If R1 does nothing, what would be the correct fitment? The 1N cap before both R1/R2 so that it forms a voltage Divider without the cap in the middle?

Right-click on the V1 voltage source, and on the right side of the window that pops up, in the "AC amplitude" box, write 1. Then you can probe whatever point you want in the signal path, and you'll get the "absolute dB" (so-to-speak) level of the signal at the respective point.

Mentioned this previous on reply.

And you'll want at the very least a 4.5k resistor in series with the 48V supply (since that's fed through a pair of 6.8k resistors in parallel - ie. 3.4k - AND in series with that, a pair of 2.2k resistors in parallel - ie. another 1.1k)

Added 4.5k simulate the 6k8/2k2 resistors. Thank you!

Why wouldn't it be?



How do you figure that?

https://groupdiy.com/threads/maudio-nova-circuit-mod.78827/

Well, for some reason I was thinking the Nova had a npn, but you are correct it is a pnp as well. I wasn't sure what this was because the Nova also doesn't have the 2nd board (it's mostly all surface mount now, so maybe it does have the oscillator and etc too?) I wasn't aware it was a Nova style mic until I just finished drawing everything out lol. That was the point, to learn how to reverse engineer/draw out of a schematic by tracing and etc.

 

[Khron]

If R1 does nothing, what would be the correct fitment? The 1N cap before bith R1/R2 so that it forms a voltage Divider without the cap in the middle?

You can just delete it.

Well, for some reason I was thinking the Nova had a npn, but you are correct it is a pnp as well. I wasn't sure what this was because the Nova also doesn't have the 2nd board (it's mostly all surface mount now, so maybe it does have the oscillator and etc too?)

Cheap companies are cheap - why spend $$$ on R&D, when there are ready-made solutions already?

The Nova does indeed seem to not have any capsule bias voltage booster.

 

[Amled87]

You can just delete it.



Cheap companies are cheap - why spend $$$ on R&D, when there are ready-made solutions already?

The Nova does indeed seem to not have any capsule bias voltage booster.

I'm not sure it actually does anything here to be quite honest. I have roughly 25v at the junction of the 2k2 resistors. The capsule connection on oscillator board when measured is at 19v.

 

[Khron]

The capsule connection on oscillator board when measured is at 19v.

... Because that's got at least a 1Meg (if not larger) series resistor for filtering that voltage, and your voltmeter has a finite input resistance / impedance (1-10Meg), thus forming a voltage divider.

And that's ignoring the very limited current output capability of said oscillator board...

Do you get the same 19V on both sides of R20?