FEELER - Solid-state microphone upgrade boards (OpAmp, JFET, U87A, FET847)

[Khron]

Motivated by some of the projects i've seen (and even contributed to, in some form or another) around here and in other places, i at long last decided to strive to finalize some ideas and circuit designs i had been dabbling with, on-and-off, for the last couple of years.

While some similar ones (or not) are already available from various sources, they suffer from one or several pet peeves of mine - 15 million different component values, unnecessarily pricey parts, questionable PCB routing, thoroughly inelegant assembly & wiring, and so forth.

While some already commercially available boards do indeed exist, they seem to be intended either for one (and ONLY one) mic body / chassis (thereby complicating mechanical mounting, and/or requiring shoddy fastening solutions etc), or the list of compatible ones is not mentioned at all (despite the existence of even more than one pair of mounting holes). As an added personal challenge, one of the starting points was compatibility with as many (affordable) mic body variants as i could get my hands on, to confirm mechanical measurements.

Another one of the self-imposed requirements was to provide as many useful features as feasible, in order to make up for some of the shortcomings that some of the more affordable capsules seem to suffer from. In a couple of these cases, going full-SMD was necessary in order to manage to fit all the needed components that allow all the tone-shaping possibilities i wanted. The SMD ones i'm seriously considering offering fully-populated (for the most part, anyway - unless otherwise requested).

If anyone has any questions about my design choices, i'll gladly strive to answer and justify them - there's always a (good?) chance you might not like / agree with my responses, though As at least some of you are probably aware, I'm mostly driven by practicality and engineering common-sense. "Vintage accuracy"-fetishists, this is not for you This is for people who would rather prefer to get things working with minimal fussing about, without ten-mile-long BOMs and "esoteric" parts sourced from 20 separate suppliers. Sure, if you enjoy experimenting, be my guest and have at it But if all you want is to put these together, solder the capsule / XLR / transformer (where applicable) wires, here ya go!

"OpAmp Alice" - cardioid-only
- single-PCB design, all SMD
- OPA2991-based (~1.2mA typical current draw in total)
- 36V supply - all the dynamic range you can handle, and then some!
- no capsule coupling capacitor
- impedance-balanced output
- suitable for electret or externally-biased capsules
- passive capsule bias filtering
- output-disconnect circuit to prevent spurious power-up/down noises
- option for attenuator pad (lowering capsule voltage)
- option for low-cut / high-pass (~120Hz, 2nd-order)
- pad & low-cut can either be vertical-mount switches (mechanical modifications to body tube needed), or internal 0.1" pin-headers & jumper-links
- option for RK47 5kHz notch & slight boost at 10k-and-over
- option for K67/K87 8-10kHz notch
- streamlined BOM: only four(!) different resistor values & three(!) different capacitor values (electrolytics & C0G/NP0 ceramics)
- fits Alctron MC001, Apex 435, BM700, BM800, Focusrite CM25 mk3 (possibly earlier ones as well), M-Audio Nova, MXL 440, MXL 2001, MXL 2006, sE 2200A, t.bone SC400, t.Bone SC450

A little "sneak preview" - ignore the slightly messy silkscreen markings, i haven't yet spent the time to (also) make that side of things "pretty". But i'll get to that too...



"OpAmp Alice" - multipattern
- suitable for externally-biased capsules only
- all of the above-mentioned features PLUS...
- interconnected three-board construction
- sturdy design, trivial wiring & assembly (capsule & XLR wires only)
- CMOS-based capsule bias charge-pump
- cardioid / omni / fig-8
- pattern selection can also be a mechanical switch, or an internal pin header & jumper-link



U87A with 21st-century design sensibilities - unleash the hatred
- same mechanical compatibility as the "OpAmp Alice" designs
- all through-hole components (on the audio boards)
- power supply reconfigured for maximum noise filtering
- (up to) ~32V available for the JFET
- pattern switching revised, one single 1Gohm required overall
- no Teflon / PTFE isolation pins needed
- streamlined BOM: only five different resistor values (versus the original's TEN, only in the audio section), & four different capacitor values (versus the original's EIGHT; electrolytics, film & C0G/NP0 ceramics)
- pad / low-cut / pattern can be switches (mechanical modifications to body tube needed), or internal 0.1" pin-headers & jumper-links
- output wiring option available for single-secondary / center-tapped secondary / dual secondary transformer



FET847 - the engineeringly-sane version
- suitable for electret or externally-biased capsules
- same mechanical compatibility as the "OpAmp Alice" designs
- all through-hole components
- single-layer, optimized routing - top copper layer unnecessary, if one pays even a modicum of attention to component placement / layout
- no Teflon / PTFE isolation pins needed
- design flaws of the original project corrected
- optimized BOM: (more) conventional component footprints
- output wiring option available for single-secondary / center-tapped secondary / dual secondary transformer

FET847 multipattern
- suitable for externally-biased capsules only
- same board set as U87A
- only needs one wire jumper, and not-populating the feedback components

CMOS-based capsule bias oscillator
- same mechanical compatibility as the "OpAmp Alice" designs
- all SMD
- 40106-based
- facility for capsule voltage attenuation / pad
- facility for pattern switch

Currently i'm still "polishing" some final board details (mainly on the mechanical side of things), and i should have a first batch of prototype boards in hand in the very near future, so stay tuned!

Unless otherwise requested, i was thinking it's probably most practical to offer the SMD boards ready-assembled, and the through-hole ones as-they-are (so you can trick them out with your favourite Dale resistors and whatever else gets your rocks off spending money on). It would also likely be nice of me to put together some build guides in some form or another; i would expect them to be more informative than pretty though, so you've been warned

But for the time being, this is intended mainly just to gauge if or whatever interest there may or may not be around here, for things like these. Of course, all of these projects and boards are still pending physical prototyping, so keep that in mind - i will of course update descriptions and details if need be, and I'll gladly listen to whatever thoughts or concerns wiser people than i might have. Just try not to be offended in case i end up shooting them down due to... y'know... sanity, and other such "heathenistic" reasons

Edited 05.06.2024 - added a couple other board screenshots
Edited 06.06.2024 - corrected a couple numbers & typos
Edited 17.06.2024 - added optional low-pass to FET847 cardioid-only & updated board screenshot
Edited 24.06.2024 - added MXL 440 mounting holes, updated board screenshots

 

[kingkorg]

Motivated by some of the projects i've seen (and even contributed to, in some form or another) around here and in other places, i at long last decided to strive to finalize some ideas and circuit designs i had been dabbling with, on-and-off, for the last couple of years.

While some similar ones (or not) are already available from various sources, they suffer from one or several pet peeves of mine - 15 million different component values, unnecessarily pricey parts, questionable PCB routing, thoroughly inelegant assembly & wiring, and so forth.

While some already commercially available boards do indeed exist, they seem to be intended either for one (and ONLY one) mic body / chassis (thereby complicating mechanical mounting, and/or requiring shoddy fastening solutions etc), or the list of compatible ones is not mentioned at all (despite the existence of even more than one pair of mounting holes). As an added personal challenge, one of the starting points was compatibility with as many (affordable) mic body variants as i could get my hands on, to confirm mechanical measurements.

Another one of the self-imposed requirements was to provide as many useful features as feasible, in order to make up for some of the shortcomings that some of the more affordable capsules seem to suffer from. In a couple of these cases, going full-SMD was necessary in order to manage to fit all the needed components that allow all the tone-shaping possibilities i wanted. The SMD ones i'm seriously considering offering fully-populated (for the most part, anyway - unless otherwise requested).

If anyone has any questions about my design choices, i'll gladly strive to answer and justify them - there's always a (good?) chance you might not like / agree with my responses, though As at least some of you are probably aware, I'm mostly driven by practicality and engineering common-sense. "Vintage accuracy"-fetishists, this is not for you This is for people who would rather prefer to get things working with minimal fussing about, without ten-mile-long BOMs and "esoteric" parts sourced from 20 separate suppliers. Sure, if you enjoy experimenting, be my guest and have at it But if all you want is to put these together, solder the capsule / XLR / transformer (where applicable) wires, here ya go!

"OpAmp Alice" - cardioid-only
- single-PCB design, all SMD
- OPA2991-based (~1.2mA typical current draw in total)
- 36V supply - all the dynamic range you can handle, and then some!
- no capsule coupling capacitor
- impedance-balanced output
- suitable for electret or externally-biased capsules
- passive capsule bias filtering
- output-disconnect circuit to prevent spurious power-up/down noises
- option for attenuator pad (lowering capsule voltage)
- option for low-cut / high-pass (~120Hz, 2nd-order)
- pad & low-cut can either be vertical-mount switches (mechanical modifications to body tube needed), or internal 0.1" pin-headers & jumper-links
- option for RK47 5kHz notch & slight boost at 10k-and-over
- option for K67/K87 8-10kHz notch
- streamlined BOM: only four(!) different resistor values & three(!) different capacitor values (electrolytics & C0G/NP0 ceramics)
- fits Alctron MC001, Apex 435, BM700, BM800, Focusrite CM25 mk3 (possibly earlier ones as well), MXL 2001, MXL2006, sE 2200A, t.bone SC400, t.Bone SC450

"OpAmp Alice" - multipattern
- suitable for externally-biased capsules only
- all of the above-mentioned features PLUS...
- interconnected three-board construction
- sturdy design, trivial wiring & assembly (capsule & XLR wires only)
- CMOS-based capsule bias charge-pump
- cardioid / omni / fig-8
- pattern selection can also be a mechanical switch, or an internal pin header & jumper-link

U87A with 21st-century design sensibilities - unleash the hatred
- same mechanical compatibility as the "OpAmp Alice" designs
- all through-hole components
- power supply reconfigured for maximum noise filtering
- (up to) ~35V available for the JFET
- pattern switching revised, one single 1Gohm required overall
- no Teflon / PTFE isolation pins needed
- streamlined BOM: only five different resistor values (versus the original's TEN, only in the audio section), & four different capacitor values (versus the original's EIGHT; electrolytics, film & C0G/NP0 ceramics)
- pad / low-cut / pattern can be switches (mechanical modifications to body tube needed), or internal 0.1" pin-headers & jumper-links
- output wiring option available for single-secondary / center-tapped secondary / dual secondary transformer

FET847 - the engineeringly-sane version
- suitable for electret or externally-biased capsules
- same mechanical compatibility as the "OpAmp Alice" designs
- all through-hole components
- single-layer, optimized routing - top copper layer unnecessary, if one pays even a modicum of attention to component placement / layout
- no Teflon / PTFE isolation pins needed
- design flaws of the original project corrected
- optimized BOM: (more) conventional component footprints

FET847 multipattern
- suitable for externally-biased capsules only
- same board set as U87A
- only needs one wire jumper, and not-populating the feedback components

CMOS-based capsule bias oscillator
- same mechanical compatibility as the "OpAmp Alice" designs
- all SMD
- 40106-based
- facility for capsule voltage attenuation / pad
- facility for pattern switch

Currently i'm still "polishing" some final board details (mainly on the mechanical side of things), and i should have a first batch of prototype boards in hand in the very near future, so stay tuned!

Unless otherwise requested, i was thinking it's probably most practical to offer the SMD boards ready-assembled, and the through-hole ones as-they-are (so you can trick them out with your favourite Dale resistors and whatever else gets your rocks off spending money on). It would also likely be nice of me to put together some build guides in some form or another; i would expect them to be more informative than pretty though, so you've been warned

But for the time being, this is intended mainly just to gauge if or whatever interest there may or may not be around here, for things like these. Of course, all of these projects and boards are still pending physical prototyping, so keep that in mind - i will of course update descriptions and details if need be, and I'll gladly listen to whatever thoughts or concerns wiser people than i might have. Just try not to be offended in case i end up shooting them down due to... y'know... sanity, and other such "heathenistic" reasons

This is beyond great news!!!

 

[elskardio]

- design flaws of the original project corrected

You don’t like the 5 minutes warmup time?

 

[Khron]

You don’t like the 5 minutes warmup time?

Tubes are "bad" enough with the warmup time, and we're in 2024 Humanity, as a whole, can do better, and there's no reason to expect anything less than the bare minimum of (no longer) common sense... Wouldn't you say?

 

[ubxf]

Fet Yeah : )

 

[zaraxisof]

Those are good news Chris

 

[Khron]

Added a board example to the first post, for those curious / intrigued...

 

[Wordsushi]

DUDE! This is fantastic news! I'm very intrigued and I love all the different EQ options for the Alice board and all the other forward-looking upgrades you've drawn up for these circuits. This is very, very cool.

 

[Khron]

DUDE! This is fantastic news! I'm very intrigued and I love all the different EQ options for the Alice board and all the other forward-looking upgrades you've drawn up for these circuits. This is very, very cool.

Thank you kindly for the enthusiasm and interest! These have been either in the back of my mind and/or on the back burner for a good couple of years now, but I'm pretty sure this is my time And I'm very much looking forward to confirming these work as well as intended in practice - no reason they shouldn't, but i've suffered brainfarts before...

 

[jp8]

I understand many of your requirements, although in some cases I would make different choices myself, but I still have a few questions and comments.

Q1: are all of them P48-phantom powered? Or do you also consider external power?

Q2: you did not mention RFI immunity as a requirement (which is one of my hobby horses). Did you give it any thought?

Q3: I've seen that wish for a 36V op-amp supply mentioned before in other threads, but I still fail to see the rationale. Assuming P48 phantom supply with a minimum voltage of 44V, taking into account the 1.2mA op-amp quiescent current and operating in Class B, assuming a 1.5k load from the mic preamp, and no polarization oscillator, then a quick and dirty calculation tells me a sustained 15Vpp is the max the circuit can deliver into the preamp input. And any preamp would be happy with much lower voltages and still not produce audible noise, so why would I want such high voltages anyway? But maybe you or someone else could enlighten me as to why one would want so much headroom? I know about Kingkorg's extreme 20-something Vpp from an SDC with 40V polarization voltage, but then I would rather choose to add a -20dB pad to reduce the peak voltage to a level the average preamp can handle, than to tru and design a circuit that can feed this >20Vpp voltage into the mic preamp with the risk of sending it into clipping. But maybe I'm missing something, and I'd be happy to learn.

I look forward to seeing your designs published here or somewhere else! Enjoy and good luck!

Jan

 

[Khron]

Q1 - Yes, all phantom-powered.
Q2 - While i admit to not exactly having RFI-tested any of my previous designs (a Schoeps board and a JFET+PNP+transformer), i also don't recall encountering any such issues with those, so... "So far, so good" - at least for the time being Or at least i'll cross that bridge when i get to it, hence the prototype phase.
Q3 - The padding is actually not a bad idea; i guess i was (at least temporarily) blinded by not wanting to clip the impedance converter inside the mic, but sure, i think i can squeeze in a couple more resistors there. Good call!

 

[Voyager10]

Would it be cheeky to ask about the M-Audio Nova as a donor body?

I have a couple, both the body and headbasket are heavy and solidly made, the main oddity is the mounting rails are close together and won't accept typical tapered PCBs.

I'll post pictures / dimensions if you like.

 

[Khron]

Would it be cheeky to ask about the M-Audio Nova as a donor body?

I have a couple, both the body and headbasket are heavy and solidly made, the main oddity is the mounting rails are close together and won't accept typical tapered PCBs.

I'll post pictures / dimensions if you like.

Cheeky, not necessarily, i'll just have to see how feasible it is, so no guarantees there

Go ahead and snap a photo of the insides, and i can then annotate / indicate which dimensions i'd need

 

[Voyager10]

Thanks! Here's an overview:


Other pics attached below.

 

[Voyager10]

Sorry to spam your thread, but if it's of interest, here's the MXL 440. Nothing like as weighty as the Nova, but very cheap secondhand.

 

[mrgrooves666]

Following anxiously, would love to try some of these Khron!

 

[Silas]

Motivated by some of the projects i've seen (and even contributed to, in some form or another) around here and in other places, i at long last decided to strive to finalize some ideas and circuit designs i had been dabbling with, on-and-off, for the last couple of years.

While some similar ones (or not) are already available from various sources, they suffer from one or several pet peeves of mine - 15 million different component values, unnecessarily pricey parts, questionable PCB routing, thoroughly inelegant assembly & wiring, and so forth.

While some already commercially available boards do indeed exist, they seem to be intended either for one (and ONLY one) mic body / chassis (thereby complicating mechanical mounting, and/or requiring shoddy fastening solutions etc), or the list of compatible ones is not mentioned at all (despite the existence of even more than one pair of mounting holes). As an added personal challenge, one of the starting points was compatibility with as many (affordable) mic body variants as i could get my hands on, to confirm mechanical measurements.

Another one of the self-imposed requirements was to provide as many useful features as feasible, in order to make up for some of the shortcomings that some of the more affordable capsules seem to suffer from. In a couple of these cases, going full-SMD was necessary in order to manage to fit all the needed components that allow all the tone-shaping possibilities i wanted. The SMD ones i'm seriously considering offering fully-populated (for the most part, anyway - unless otherwise requested).

If anyone has any questions about my design choices, i'll gladly strive to answer and justify them - there's always a (good?) chance you might not like / agree with my responses, though As at least some of you are probably aware, I'm mostly driven by practicality and engineering common-sense. "Vintage accuracy"-fetishists, this is not for you This is for people who would rather prefer to get things working with minimal fussing about, without ten-mile-long BOMs and "esoteric" parts sourced from 20 separate suppliers. Sure, if you enjoy experimenting, be my guest and have at it But if all you want is to put these together, solder the capsule / XLR / transformer (where applicable) wires, here ya go!

"OpAmp Alice" - cardioid-only
- single-PCB design, all SMD
- OPA2991-based (~1.2mA typical current draw in total)
- 36V supply - all the dynamic range you can handle, and then some!
- no capsule coupling capacitor
- impedance-balanced output
- suitable for electret or externally-biased capsules
- passive capsule bias filtering
- output-disconnect circuit to prevent spurious power-up/down noises
- option for attenuator pad (lowering capsule voltage)
- option for low-cut / high-pass (~120Hz, 2nd-order)
- pad & low-cut can either be vertical-mount switches (mechanical modifications to body tube needed), or internal 0.1" pin-headers & jumper-links
- option for RK47 5kHz notch & slight boost at 10k-and-over
- option for K67/K87 8-10kHz notch
- streamlined BOM: only four(!) different resistor values & three(!) different capacitor values (electrolytics & C0G/NP0 ceramics)
- fits Alctron MC001, Apex 435, BM700, BM800, Focusrite CM25 mk3 (possibly earlier ones as well), MXL 2001, MXL2006, sE 2200A, t.bone SC400, t.Bone SC450

A little "sneak preview" - ignore the slightly messy silkscreen markings, i haven't yet spent the time to (also) make that side of things "pretty". But i'll get to that too...

View attachment 129648

"OpAmp Alice" - multipattern
- suitable for externally-biased capsules only
- all of the above-mentioned features PLUS...
- interconnected three-board construction
- sturdy design, trivial wiring & assembly (capsule & XLR wires only)
- CMOS-based capsule bias charge-pump
- cardioid / omni / fig-8
- pattern selection can also be a mechanical switch, or an internal pin header & jumper-link

U87A with 21st-century design sensibilities - unleash the hatred
- same mechanical compatibility as the "OpAmp Alice" designs
- all through-hole components
- power supply reconfigured for maximum noise filtering
- (up to) ~35V available for the JFET
- pattern switching revised, one single 1Gohm required overall
- no Teflon / PTFE isolation pins needed
- streamlined BOM: only five different resistor values (versus the original's TEN, only in the audio section), & four different capacitor values (versus the original's EIGHT; electrolytics, film & C0G/NP0 ceramics)
- pad / low-cut / pattern can be switches (mechanical modifications to body tube needed), or internal 0.1" pin-headers & jumper-links
- output wiring option available for single-secondary / center-tapped secondary / dual secondary transformer

FET847 - the engineeringly-sane version
- suitable for electret or externally-biased capsules
- same mechanical compatibility as the "OpAmp Alice" designs
- all through-hole components
- single-layer, optimized routing - top copper layer unnecessary, if one pays even a modicum of attention to component placement / layout
- no Teflon / PTFE isolation pins needed
- design flaws of the original project corrected
- optimized BOM: (more) conventional component footprints

FET847 multipattern
- suitable for externally-biased capsules only
- same board set as U87A
- only needs one wire jumper, and not-populating the feedback components

CMOS-based capsule bias oscillator
- same mechanical compatibility as the "OpAmp Alice" designs
- all SMD
- 40106-based
- facility for capsule voltage attenuation / pad
- facility for pattern switch

Currently i'm still "polishing" some final board details (mainly on the mechanical side of things), and i should have a first batch of prototype boards in hand in the very near future, so stay tuned!

Unless otherwise requested, i was thinking it's probably most practical to offer the SMD boards ready-assembled, and the through-hole ones as-they-are (so you can trick them out with your favourite Dale resistors and whatever else gets your rocks off spending money on). It would also likely be nice of me to put together some build guides in some form or another; i would expect them to be more informative than pretty though, so you've been warned

But for the time being, this is intended mainly just to gauge if or whatever interest there may or may not be around here, for things like these. Of course, all of these projects and boards are still pending physical prototyping, so keep that in mind - i will of course update descriptions and details if need be, and I'll gladly listen to whatever thoughts or concerns wiser people than i might have. Just try not to be offended in case i end up shooting them down due to... y'know... sanity, and other such "heathenistic" reasons

This is great! Would be most keen to build one or two types of these. Good luck with the prototyping…!

 

[Khron]

Thanks! Here's an overview:
View attachment 129716

Other pics attached below.

So that's about 35mm "axially" between screws, and from another photo, about 26mm between screw centers in the other direction, right? Thaaaaaaaat's pretty "toight" indeed, i really doubt i can squeeze things down that narrowly, on the existing boards Might keep it in mind for some later revisions though



Those two dimensions highlighted in light blue i'd still need, as well as the internal diameter of the (top of the) body tube.

Sorry to spam your thread, but if it's of interest, here's the MXL 440. Nothing like as weighty as the Nova, but very cheap secondhand.

That looks a helluva lot like the Alctron MC001 / Apex 435 / t.Bone SC400, actually. I'm reading about 34mm width-ways between screw centers..? And i could use the same dimensions i asked for the Nova (the bottom of the headbasket to screw center and to the bottom of the board, and body tube ID).

 

[Voyager10]

For the Nova - yes, it's 35mm axially between mounting holes, and 27mm laterally. The available width for components on the backside of the board (between the mounting rails) is also about 27mm.

From your picture: bottom-of-headbasket to first mounting hole is 14mm, and to the top of the transformer housing it's 70mm. Body internal diameter is just under 45mm (44.8mm) at the headbasket end (50mm OD, it's 2.5mm thick!).


For the MXL 440: just one mounting screw for each side. Bottom-of-headbasket to these screws is 22mm, and to the XLR housing it's 73mm. Lateral distance between mounting screws is 35mm, and ID of the tube is 44.8mm (it's the same size all the way down).

Thanks!

 

[Khron]

For the Nova - yes, it's 35mm axially between mounting holes, and 27mm laterally. The available width for components on the backside of the board (between the mounting rails) is also about 27mm.

From your picture: bottom-of-headbasket to first mounting hole is 14mm, and to the top of the transformer housing it's 70mm. Body internal diameter is just under 45mm (44.8mm) at the headbasket end (50mm OD, it's 2.5mm thick!).


For the MXL 440: just one mounting screw for each side. Bottom-of-headbasket to these screws is 22mm, and to the XLR housing it's 73mm. Lateral distance between mounting screws is 35mm, and ID of the tube is 44.8mm (it's the same size all the way down).

Thanks!

Yeah, reasonably sure the Nova would need its own board design(s). I'm still at work for another 6h or so, but when i get home i'll double-check / confirm whether those MXL 440 dimensions check out with the ones i suspect or not