MicBuilders@groups.io | Takstar CM-63 Schematics and teardown

#35478

To continue my bewildering obsession of finding a good and cheap SDC mic, I have completed a mostly completed schematic of the Takstar CM63 microphone.
This one is a bit interesting, as it has very large vents behind the capsule. There are 4 levels of vents, each 2.5mm wide. Compare this to the common MXL 663/991 which has 4 levels of vents at 1.25mm wide. That's twice as wide - the bigger the better - right? But this makes the chamber before the cone quite large. The tip of the cone is inset 10mm and the backplate of the capsule is inset 3mm with 2mm of thread, meaning it's about 11mm from the back of the capsule to the tip of the cone. On the MXL991 I have, it's 7mm deep on the body, 2.2mm thread, with the back plate at 2.2mm deep, so it has 7mm from the backplate of the capsule to the tip of the cone.

The quality of the threads are touted as some of the best of the cheap SDC mics. It's your standard 22mm capsule with 0.5mm thread pitch and 1mm of center pin stickout.
The capsule has a peculiar plate over the top which likely affects HF response and causes some HF resonances.
The capsule is 12mm long with 20.65 thread ID and the retaining ring using the same thread.
The OD of the mic body is 22.1mm and ID is 20mm.
The mic with capsule is 145mm long and 135mm without capsule - making it among the longer 22mm mics (most average 132mm including the Takstar CM90 and MXL991 at 134.5)
The body capsule thread OD is 20.85mm and 2mm deep.

The PCB is 1.2mm thick

17mm wide, but it looks like you can fit up to 19mm without modification.

with 15mm wide step down at XLR connector 2mm long for the XLR body.

94mm long

Offset 2.4mm from center of the body/cone.
The XLR connector has an integrated body contact pin on the ground.

The circuit is modeled after the MXL663 with a few modifications.
It has an unpopulated spot for one of the 12v zeners, and inductor added in series to the positive side of the before the boost circuit zener.
Ferrite beads (I think) on pin 2 and 3, AND a common mode choke (3.2mm by 1.65mm and about 12 windings in parallel)
It also has a ferrite bead from XLR ground to the PCB ground - isolating the PCB from the body and capsule ground.
The high pass and PAD are simply capacitors - adding a capacitor in parallel with the capsule for the PAD, or in series with the amplifier transistors bases (both of them)
it uses a Toshiba 2SK208-Y Jfet which has an impressive 0.5dB noise, 1.2-3.0mA Idss, and minimum 1.2mS forward transfer admittance.

The capsule input capacitor is a 1206 (imperial), leading me to believe it's likely C0G. The high impedance portion of the PCB is coated with an epoxy conformal coating, likely for moisture protection.
The PCB is very clean and clear of flux with exception of the high pass/pad switches which look to be hand soldered.
I'm sorry I didn't get any capacitor values. I don't have the right probe for it.

Overall this is exactly what you'd want to see in a low cost mic, and there isn't a whole lot of improvements to be made here. A skilled person could replace the caps with polystyrene, and replace the electrolytic caps as there's plenty of room. There's also a spot you can add additional 0805 capacitor parallel to the boost zener.

For quick reference I opened up my Takstar CM60 to see the comparison. The CM60 is different in the following ways:
CM60 lacks the common mode choke and ferrite beads but adds caps from pin2/3 to ground
CM60 has holes drilled under the capsule bias resistor and output capacitor.
CM60 is lacking the trimmer pot but has a location to put one.
Through hole caps on the low pass circuit.
No inductor on the boost positive rail.
Given the choice, I'd say get the CM63. The capsule internals are identical, but the CM63 capsule housing is 3mm shorter, putting the diaphragm that much closer to the front.
It would be an interesting experiment to put an identical circuit in each and find the difference in response with the different venting and different grill pattern.
I will say that if you're looking for a 22mm mic body and don't need the PAD and High pass, get the CM60 over any MXL SDC. It's only a few mm shorter and the body is much higher quality.

I think I'm getting better at this. I was going to trace the PCB but it would have been more work.
Next I might do an old AKG Perception 150 I got. It has one of the longest 22mm SDC bodies at 160mm and wide vents. I took a quick peek at the PCB and it's almost identical layout, cone, and components as the Takstar CM63, leading me to believe it was manufactured by the same people.

-Stephanie

Takstar_CM63_v01.PNG

#35479

Le 01/07/2023 à 14:23, Stephanie a écrit :

skilled person could replace the caps with polystyrene, and replace the electrolytic caps as there's plenty of room.

I seriously doubt replacing this cap with a styrene would change anything to teh sound.

The voltage across it is extremely low and CoG are very linear.

BTW, Sw2 is the High pass filter.

For quick reference I opened up my Takstar CM60 to see the comparison. The CM60 is different in the following ways:
CM60 lacks the common mode choke and ferrite beads but adds caps from pin2/3 to ground
CM60 has holes drilled under the capsule bias resistor and output capacitor.
CM60 is lacking the trimmer pot but has a location to put one.
Through hole caps on the low pass circuit.
No inductor on the boost positive rail.
Given the choice, I'd say get the CM63. The capsule internals are identical, but the CM63 capsule housing is 3mm shorter, putting the diaphragm that much closer to the front.
It would be an interesting experiment to put an identical circuit in each and find the difference in response with the different venting and different grill pattern.
I will say that if you're looking for a 22mm mic body and don't need the PAD and High pass, get the CM60 over any MXL SDC. It's only a few mm shorter and the body is much higher quality.

I think I'm getting better at this. I was going to trace the PCB but it would have been more work.
Next I might do an old AKG Perception 150 I got. It has one of the longest 22mm SDC bodies at 160mm and wide vents. I took a quick peek at the PCB and it's almost identical layout, cone, and components as the Takstar CM63, leading me to believe it was manufactured by the same people.

-Stephanie

#35484

Thanks for the spot check. I also mislabelled some inductors and diodes. Fixed them in this schematic. I decided to add the unpopulated cap and zener as well.

I agree there wouldn't be much improvement in swapping the capacitors if they're C0G. I'm not saying you should, just that you could. There's not much I can do to mod this circuit. But I could easily add more ceramics across the electrolytics. A 0805 would nicely fit between the capacitor leads, a 0.1-1uF might help with noise just a little bit. It'd cost pennies and I already have the components. I'm also thinking how the CM60 has holes drilled under the bias resistor and output cap. I could easily do that on a drill press, but would require removing the conformal coating and finding a suitable replacement. I'm not sure how much of a difference that would make, even if I routed out a slot.
I could also fix L3 - typically L2 and L3 are close to each other to couple, I'm pretty sure the slant in it was not on purpose as it doesn't exist on the CM60.

I'm thinking about making a dummy capsule for circuit noise measurements. I'd just remove the membrane from the MXL capsule and solder in a film capacitor of equal capacitance. Then I wouldn't need a quiet place to test.

-Stephanie

Takstar_CM63_v02.PNG

#35485

I have very high opinions of Takstar - here https://youtu.be/5dlWaDQZP4g?t=921 is the CM60 under test.

Best wishes

David P

Show quoted text

On Sat, Jul 1, 2023 at 2:23 PM Stephanie <Qwertys@...> wrote:

To continue my bewildering obsession of finding a good and cheap SDC mic, I have completed a mostly completed schematic of the Takstar CM63 microphone.
This one is a bit interesting, as it has very large vents behind the capsule. There are 4 levels of vents, each 2.5mm wide. Compare this to the common MXL 663/991 which has 4 levels of vents at 1.25mm wide. That's twice as wide - the bigger the better - right? But this makes the chamber before the cone quite large. The tip of the cone is inset 10mm and the backplate of the capsule is inset 3mm with 2mm of thread, meaning it's about 11mm from the back of the capsule to the tip of the cone. On the MXL991 I have, it's 7mm deep on the body, 2.2mm thread, with the back plate at 2.2mm deep, so it has 7mm from the backplate of the capsule to the tip of the cone.

The quality of the threads are touted as some of the best of the cheap SDC mics. It's your standard 22mm capsule with 0.5mm thread pitch and 1mm of center pin stickout.
The capsule has a peculiar plate over the top which likely affects HF response and causes some HF resonances.
The capsule is 12mm long with 20.65 thread ID and the retaining ring using the same thread.
The OD of the mic body is 22.1mm and ID is 20mm.
The mic with capsule is 145mm long and 135mm without capsule - making it among the longer 22mm mics (most average 132mm including the Takstar CM90 and MXL991 at 134.5)
The body capsule thread OD is 20.85mm and 2mm deep.

The PCB is 1.2mm thick
17mm wide, but it looks like you can fit up to 19mm without modification.

with 15mm wide step down at XLR connector 2mm long for the XLR body.

94mm long

Offset 2.4mm from center of the body/cone.
The XLR connector has an integrated body contact pin on the ground.

The circuit is modeled after the MXL663 with a few modifications.
It has an unpopulated spot for one of the 12v zeners, and inductor added in series to the positive side of the before the boost circuit zener.
Ferrite beads (I think) on pin 2 and 3, AND a common mode choke (3.2mm by 1.65mm and about 12 windings in parallel)
It also has a ferrite bead from XLR ground to the PCB ground - isolating the PCB from the body and capsule ground.
The high pass and PAD are simply capacitors - adding a capacitor in parallel with the capsule for the PAD, or in series with the amplifier transistors bases (both of them)
it uses a Toshiba 2SK208-Y Jfet which has an impressive 0.5dB noise, 1.2-3.0mA Idss, and minimum 1.2mS forward transfer admittance.

The capsule input capacitor is a 1206 (imperial), leading me to believe it's likely C0G. The high impedance portion of the PCB is coated with an epoxy conformal coating, likely for moisture protection.
The PCB is very clean and clear of flux with exception of the high pass/pad switches which look to be hand soldered.
I'm sorry I didn't get any capacitor values. I don't have the right probe for it.

Overall this is exactly what you'd want to see in a low cost mic, and there isn't a whole lot of improvements to be made here. A skilled person could replace the caps with polystyrene, and replace the electrolytic caps as there's plenty of room. There's also a spot you can add additional 0805 capacitor parallel to the boost zener.

For quick reference I opened up my Takstar CM60 to see the comparison. The CM60 is different in the following ways:
CM60 lacks the common mode choke and ferrite beads but adds caps from pin2/3 to ground
CM60 has holes drilled under the capsule bias resistor and output capacitor.
CM60 is lacking the trimmer pot but has a location to put one.
Through hole caps on the low pass circuit.
No inductor on the boost positive rail.
Given the choice, I'd say get the CM63. The capsule internals are identical, but the CM63 capsule housing is 3mm shorter, putting the diaphragm that much closer to the front.
It would be an interesting experiment to put an identical circuit in each and find the difference in response with the different venting and different grill pattern.
I will say that if you're looking for a 22mm mic body and don't need the PAD and High pass, get the CM60 over any MXL SDC. It's only a few mm shorter and the body is much higher quality.

I think I'm getting better at this. I was going to trace the PCB but it would have been more work.
Next I might do an old AKG Perception 150 I got. It has one of the longest 22mm SDC bodies at 160mm and wide vents. I took a quick peek at the PCB and it's almost identical layout, cone, and components as the Takstar CM63, leading me to believe it was manufactured by the same people.

-Stephanie

--

- - - - - - - - - - - - - - - - - - - - - - - -
David Pinnegar, B.Sc., A.R.C.S.
- - - - - - - - - - - - - - - - - - - - - - - -
+44 1342 850594

#35545

Thanks for the teardown! Do you have a photo of the cone arrangement? You can make it out in product images but a clear shot would be great if it's possible. I'm building a SDC at the moment and am wondering about adding a cone reflector or not. Cheers :-)

#35554

Thank you for this in depth exposé,

I do share your interest in good SDC's. Although not in a bewildered manner or obsession, sorry.
I just love good SDC's. I found out that good quality and cheap don't always go together unfortunately.
Some of the Schoeps CMC's, AKG C's, Neumann KM, Josephson and Audio-Technica are not cheap.
But I have managed to work on the generic MXL's APEX etc. with good results.
These days, I just rip the PCB's and build my own versions...I'm done with "upgrading" caps etc.

Looking at the schematic of the Takstar CM63 it looks somehow familiar. Other mics use similar circuits.
Where the difference probably lies, the body (vents as you mentioned) capsule, parts quality etc
Many people here hold Takstar in high esteem, with good reasons.

Yes! Could you please share a good clean shot of the cone and head mechanism please?

Stay tuned for more.
M