Topic: Battery/PiP powered jfet buffer for TSB-120a --> Korg MR-1

[ghw]

Howdy, folks!  I've been a lurker for a long-ish time.  First post, here goes...

So, given all the praise heaped upon the Church CA-11s and the Core Sound tetramic, it seems to me that the Transsound TSB-120a are probably the best deal going in cardioid microphone capsules right now at <$10 each (other contenders?).  Especially desirable is the removable circuit board, which allows one to use the capsule as is, or to supply an external jfet buffer, etc.  The latter is what I would like to do, seeing as how the stock circuit is drastically noisier (27dbA according to Church) than the capsule itself is capable of (<15 dbA assuming Core Sound is to be believed). 

My MR-1 (presently in transit from B&H damn they got cheap!) takes balanced ins, and all the battery box circuits I've seen have had unbalanced outputs.  Ideally, I would like to use all of my pins.  Thus, I'm wondering about battery or PiP powered buffer circuits that would give me a balanced output.

The obvious jfet choices would seem to be the LSK170 or LSK389 (dual) N-channel JFETs, but there are probably other good options.  For a phantom powered application, the tried-and-true Schoeps circuit (see Alice mic) would do, but this isn't a PP application.  When I find a SPICE app that will run on this iMac I'm using I'd like to futz around with a couple ideas I have floating around, but in the mean time, anyone have any suggestions?

Since I don't have the deck in hand yet...do you folks know if the plugin power is available on only one or both of the signal pins?


Thanks,
George

[Church-Audio]

Howdy, folks!  I've been a lurker for a long-ish time.  First post, here goes...

So, given all the praise heaped upon the Church CA-11s and the Core Sound tetramic, it seems to me that the Transsound TSB-120a are probably the best deal going in cardioid microphone capsules right now at <$10 each (other contenders?).  Especially desirable is the removable circuit board, which allows one to use the capsule as is, or to supply an external jfet buffer, etc.  The latter is what I would like to do, seeing as how the stock circuit is drastically noisier (27dbA according to Church) than the capsule itself is capable of (<15 dbA assuming Core Sound is to be believed). 

My MR-1 (presently in transit from B&H damn they got cheap!) takes balanced ins, and all the battery box circuits I've seen have had unbalanced outputs.  Ideally, I would like to use all of my pins.  Thus, I'm wondering about battery or PiP powered buffer circuits that would give me a balanced output.

The obvious jfet choices would seem to be the LSK170 or LSK389 (dual) N-channel JFETs, but there are probably other good options.  For a phantom powered application, the tried-and-true Schoeps circuit (see Alice mic) would do, but this isn't a PP application.  When I find a SPICE app that will run on this iMac I'm using I'd like to futz around with a couple ideas I have floating around, but in the mean time, anyone have any suggestions?

Since I don't have the deck in hand yet...do you folks know if the plugin power is available on only one or both of the signal pins?


Thanks,
George

The capsule I am using is not a stock 120... Its made for me + I modify the capsule further  I also think it depends on what your application is 27dbA is plenty quiet for recording loud shows.. But not very good for recording quiet sound sources. I have tryed other fets and the performance gain was marginal at best. As far as bufffer circuits there are plenty around I like some of the ic chips made by THAT. They make a good chip for doing what you want to do. What is it you want to record?

[illconditioned]

Howdy, folks!  I've been a lurker for a long-ish time.  First post, here goes...

So, given all the praise heaped upon the Church CA-11s and the Core Sound tetramic, it seems to me that the Transsound TSB-120a are probably the best deal going in cardioid microphone capsules right now at <$10 each (other contenders?).  Especially desirable is the removable circuit board, which allows one to use the capsule as is, or to supply an external jfet buffer, etc.  The latter is what I would like to do, seeing as how the stock circuit is drastically noisier (27dbA according to Church) than the capsule itself is capable of (<15 dbA assuming Core Sound is to be believed). 

My MR-1 (presently in transit from B&H damn they got cheap!) takes balanced ins, and all the battery box circuits I've seen have had unbalanced outputs.  Ideally, I would like to use all of my pins.  Thus, I'm wondering about battery or PiP powered buffer circuits that would give me a balanced output.

The obvious jfet choices would seem to be the LSK170 or LSK389 (dual) N-channel JFETs, but there are probably other good options.  For a phantom powered application, the tried-and-true Schoeps circuit (see Alice mic) would do, but this isn't a PP application.  When I find a SPICE app that will run on this iMac I'm using I'd like to futz around with a couple ideas I have floating around, but in the mean time, anyone have any suggestions?

Since I don't have the deck in hand yet...do you folks know if the plugin power is available on only one or both of the signal pins?


Thanks,
George

Yes, the Transsound capsules are pretty good.  I would visit the Yahoo Group called "micbuilders".  Lots of people have experimented with various capsules there.  And powering circuits as well.

If you're not a hacker, or if you just want one set of mics, you might as well buy from Chris Church, though.  You'll get the capsules matched and mounted properly, with top quality parts (Neutrik connectors, etc).  As Chris wires them, these mics will work on plug-in-power on just about any recorder.

I'm mostly working with Omni mics, but for a compact cardioid these are good mics.

  Richard

[Roger Gustavsson]

The noisefloor of a TSB-120A seems to be at -112dbV. I do not think the mic preamp of the Korg MR-1 betters that by a large margin.

 
Roger

[ghw]

To Chris-

I record loud shows, quiet shows, and not shows, so if I'm going to go to the trouble to build a mic (and I am) I'd like to see the full noise floor potential of a 1/2in. capsule.

Um, as far as THAT products go, they have mic preamp ICs and line drivers and (really cool) transistor arrays, but all of the above are BJTs and thus have inadequate input impedance to buffer the output of a bare electret.  

How did you get transsound (or whomever) to manufacture you custom capsules?  That is pretty incredible.


To Richard-

Chris' two-wire approach is certainly elegant, especially for devices which use a conventional stereo mini-plug (thus unbalanced) plugin power arrangement.  However, the MR-1 uses dual mono (balanced, mini-trs plugs with plugin power and, as I stated, I'd like to use all my pins.  Call me neurotic.  On a more serious note, I operate in some terribly noisy (in every sense) environments, and long cable runs are often extremely useful.

I'm building a coincident mic for a couple of reasons, one of which is mono compatibility in a stereo device.  When I want to work with omnis, I have a pair of DPA4060s I'm pretty happy with.  I do, however, need to build a battery box for 'em since I was previously powering them with a Mic2496, which I'd like to part with.

[ghw]

Roger,
Good point.  Unfortunately, the Sonic Studios tech review, which shows the noise plot of the MR-1 after 36dB gain, doesn't list the width of their FFT bins, so its hard to (roughly) integrate accross the spectrum and derive a total noise figure.  Reading off the chart is misleading, as it would lead one to believe that the EIN of the deck is below the theoretical minimum for 150ohm source (-102dBu noise 20-20k after 36dB gain --> -138dBu EIN).

At any rate, I'm still interested as a balanced output is way useful from an rf immunity standpoint, and I run long cables.

[illconditioned]

Roger,
Good point.  Unfortunately, the Sonic Studios tech review, which shows the noise plot of the MR-1 after 36dB gain, doesn't list the width of their FFT bins, so its hard to (roughly) integrate accross the spectrum and derive a total noise figure.  Reading off the chart is misleading, as it would lead one to believe that the EIN of the deck is below the theoretical minimum for 150ohm source (-102dBu noise 20-20k after 36dB gain --> -138dBu EIN).

At any rate, I'm still interested as a balanced output is way useful from an rf immunity standpoint, and I run long cables.

I suggest a "less is more" approach.  First, I would get a deck with a (known) low noise input.  Something like a Minidisc, HIMD variety, comes to mind.  Then see that the limitation really is, the mics, the preamp, or the cables/drive.  I'm really suspicious of the MR-1, especially those noise peaks shown in the sonicstudios plots.

By careful wiring you may find that plug-in-power works just fine.  I've run a ton of mics (Countryman B3, Sennheiser MKE2, MKE40, AT853, Church Audio's mics) all from plug in power on my Sony PCMD50 (both loud and quiet shows) and Edirol R09 (loud shows and hot mics).  It sure simplifies things if the mics go right into the recorder.  No extra batteries or connections to worry about.  No extra circuits/wiring to mess up.

By the way, one project I'm currently working on is a combined battery box, preamp, that has a fixed gain of +15dB.  I'll put a mini stereo socket (with power switch) on input and a pigtail miniplug output on it.  I envision this as an input stage, whenever I need a true 9v mic power, or when I need a bit of gain, eg., for noisy mic inputs like the Edirol R09.

Now, to test this thing I've learned to use RMAA with a decent soundcard.  It is way to easy to make a mistake and never discover it.

  Richard

[Church-Audio]

To Chris-

How did you get transsound (or whomever) to manufacture you custom capsules?  That is pretty incredible.

Not that incredible actually when you buy allot of capsules and you ask for changes they make them. I dont buy 2-3 at a time I buy large quantities. Thats how I am able to match them so well. And I think its great that you want to build your own mic. But if you think my mics are just stock off the shelf products you would be wrong. Thats all I was saying. And there is no point in going for a mega low noise floor when your recorder cant even touch the noise floor your going for but what do I know Good luck with your build this is a great place to learn about live recording and there are many guys here that know much more then I do about electronics.

[udovdh]

Interesting thread.
Is it really that this TSB-120A capsule can work up to 120 dB at 1.5V over a 3K resistor? 
Or am I misreading some info? (probably)

What (simple) replacements for the FET are tested and found to be really improvements?
I mean, a small batterybox with a few parts is not an issue....

These capsules might be an interesting DIY project, at $10 each?
Or is there a European source for 10 euro/each? 

[Church-Audio]

Interesting thread.
Is it really that this TSB-120A capsule can work up to 120 dB at 1.5V over a 3K resistor? 
Or am I misreading some info? (probably)

What (simple) replacements for the FET are tested and found to be really improvements?
I mean, a small batterybox with a few parts is not an issue....

These capsules might be an interesting DIY project, at $10 each?
Or is there a European source for 10 euro/each? 

You should use a 4.7k at 9 volts the stock cap has a distortion of 8% at 114db @1k with a 2.4k resistor @ 9 volts.

[udovdh]

Interesting thread.
Is it really that this TSB-120A capsule can work up to 120 dB at 1.5V over a 3K resistor? 
Or am I misreading some info? (probably)

What (simple) replacements for the FET are tested and found to be really improvements?
I mean, a small batterybox with a few parts is not an issue....

You should use a 4.7k at 9 volts the stock cap has a distortion of 8% at 114db @1k with a 2.4k resistor @ 9 volts.

Ouch, 8% is tad much!
Thanks for the tip, how much does the distortion improve with 4k7?

[Roger Gustavsson]

You should use a 4.7k at 9 volts the stock cap has a distortion of 8% at 114db @1k with a 2.4k resistor @ 9 volts.

How much will that 4.7k resistor lower the sensitivity of the mic capsules?


Roger

[Church-Audio]

Interesting thread.
Is it really that this TSB-120A capsule can work up to 120 dB at 1.5V over a 3K resistor? 
Or am I misreading some info? (probably)

What (simple) replacements for the FET are tested and found to be really improvements?
I mean, a small batterybox with a few parts is not an issue....

You should use a 4.7k at 9 volts the stock cap has a distortion of 8% at 114db @1k with a 2.4k resistor @ 9 volts.

Ouch, 8% is tad much!
Thanks for the tip, how much does the distortion improve with 4k7?

brings it down to 0.5% at 114 db at 1k

Chis

[Church-Audio]

You should use a 4.7k at 9 volts the stock cap has a distortion of 8% at 114db @1k with a 2.4k resistor @ 9 volts.

How much will that 4.7k resistor lower the sensitivity of the mic capsules?


Roger

It brings down the output by 10db.

[udovdh]

Thanks for the tip, how much does the distortion improve with 4k7?

brings it down to 0.5% at 114 db at 1k

That's a lot better.
How does this mic stack up to the competition?
E.g. what amount of dB's for K=1% (1KHz)?
Is this competitive performance? (I think I lack oversight over the industry)
Compare to e.g. MM, DPA, etc?

[Roger Gustavsson]

The TSB-120 with a 4.7k resistor would lower the output level to about 1.6mV/94dB. Will the noisefloor drop by 10dB as well, to -122dBV? That will call for an external low noise mic preamp for recording low level events.

Roger

[Church-Audio]

The TSB-120 with a 4.7k resistor would lower the output level to about 1.6mV/94dB. Will the noisefloor drop by 10dB as well, to -122dBV? That will call for an external low noise mic preamp for recording low level events.

Roger

The total output of the capsule drops so yes noise floor would go down relative to the drop in output. You always need a good quality low noise preamp when your dealing with low acoustic levels   

[jobseek2001]

Chris, there's a discussion of some sorts going on at micbuilders about the 8% distortion figure and whether a 4K7 would do any good.
Care to explain here or there?

[Church-Audio]

Chris, there's a discussion of some sorts going on at micbuilders about the 8% distortion figure and whether a 4K7 would do any good.
Care to explain here or there?

Yeah its simple... I measure my distortion with a FFT analyser and a super low distortion sound source directly coupled to the capsule then I take my 4.7 k resistor and the distortion goes from 8% to 0.5% in an instant. Its very simple not much to tell my methods are correct in my measurements. This works not only with my capsules but with some other capsules as well. I have measured my source with a DPA 4006 measurement mic that has extremely low distortion my sound source at 114 db at 1k is less then 0.004%.

I know my distortion measurements are based on the mic capsule / acoustic properties and not just on purely fet distortion like other tests are. That is why some mics that have problems with the acoustic properties of the capsule and not electrical properties do not benefit from this mod.
This mod absolutely works no question about it.

Chris Church

[ricardo]

I measure my distortion with a FFT analyser and a super low distortion sound source directly coupled to the capsule ...

Is this a B&K 4231 Calibrator?

[Church-Audio]

I measure my distortion with a FFT analyser and a super low distortion sound source directly coupled to the capsule ...

Is this a B&K 4231 Calibrator?

bv

I dont want to give away all my little secrets.... The name and make of the device shall remain my little secret but its not a regular calibrator if that is what your asking. 

[ricardo]

Chris, I'm puzzled by the 10dB drop in sensitivity when you replace the internal 2k2 with 4k7.

Did you measure this on your "FFT analyser and a super low distortion sound source directly coupled to the capsule .. " setup?

[illconditioned]

I measure my distortion with a FFT analyser and a super low distortion sound source directly coupled to the capsule ...

Is this a B&K 4231 Calibrator?

bv

I dont want to give away all my little secrets.... The name and make of the device shall remain my little secret but its not a regular calibrator if that is what your asking. 

Just my opinion, but I don't like this secrecy.  I've said this before, and I'll say it again, there is lots of room for both vendors and diy types here.  Hiding this information does no good.  People in the know (competitors) will figure out how to do this anyway.  You (Chris) are offering a great product at a great price.  And you're constantly innovating, with new products that we (think we) need.  That is what sells your product.  IMO this is what puts you above others (like Soundpros, Core sound, and sonicstudios, for example).

  Richard

[ricardo]

There's a basic misunderstanding going on between micbuilders and here . . . they are talking about a common drain configuration, Chris is using common source.

You are right.  I was assuming the TSB120 used as in their literature; common drain / source follower.

GHW, you mention a Church type 2 wire connection.  Is there a link to this?

Chris, if it isn't too secret  , is your recommendation, "replace 2k4 with 4k7", to do with an external load resistor for common source PiP operation?

I was under the impression you were advocating replacing the internal (SMD on tiny PCB behind the capsule) 2k2.

[Church-Audio]

There's a basic misunderstanding going on between micbuilders and here . . . they are talking about a common drain configuration, Chris is using common source.

You are right.  I was assuming the TSB120 used as in their literature; common drain / source follower.

GHW, you mention a Church type 2 wire connection.  Is there a link to this?

Chris, if it isn't too secret  , is your recommendation, "replace 2k4 with 4k7", to do with an external load resistor for common source PiP operation?

I was under the impression you were advocating replacing the internal (SMD on tiny PCB behind the capsule) 2k2.

Ok I will tell you what part of the gear that I use to test the mics is a burr brown 1k oscillator its a chip that is encased in epoxy. It has one of the lowest distortions available for a 1k oscillator and I couple that with a Mylar driver that is mounted in a machined housing. But its not something off the shelf. I use a 4.7k resistor in almost all of my mics and it works provided you have approximately 10k or so load impedance and a 9 volt supply.

Chris

[illconditioned]

There's a basic misunderstanding going on between micbuilders and here . . . they are talking about a common drain configuration, Chris is using common source.

You are right.  I was assuming the TSB120 used as in their literature; common drain / source follower.

GHW, you mention a Church type 2 wire connection.  Is there a link to this?

Chris, if it isn't too secret  , is your recommendation, "replace 2k4 with 4k7", to do with an external load resistor for common source PiP operation?

I was under the impression you were advocating replacing the internal (SMD on tiny PCB behind the capsule) 2k2.

Ok I will tell you what part of the gear that I use to test the mics is a burr brown 1k oscillator its a chip that is encased in epoxy. It has one of the lowest distortions available for a 1k oscillator and I couple that with a Mylar driver that is mounted in a machined housing. But its not something off the shelf. I use a 4.7k resistor in almost all of my mics and it works provided you have approximately 10k or so load impedance and a 9 volt supply.

Chris

Thanks for the information.

I will say that you are probably the only builder to systematically tests their gear.  I have only recently started testing (PC with EMU 1212m soundcard, RMAA program), and it is certainly more tricky than I imagined.

  Richard

[ricardo]

Ok I will tell you what part of the gear that I use to test the mics is a burr brown 1k oscillator its a chip that is encased in epoxy. It has one of the lowest distortions available for a 1k oscillator and I couple that with a Mylar driver that is mounted in a machined housing. But its not something off the shelf. I use a 4.7k resistor in almost all of my mics and it works provided you have approximately 10k or so load impedance and a 9 volt supply.

Thanks for this Chris.  Now everything is clear.

I will say that you are probably the only builder to systematically tests their gear.  I have only recently started testing (PC with EMU 1212m soundcard, RMAA program), and it is certainly more tricky than I imagined.

Richard (and anyone else), do you know of free/inexpensive software/gear to measure acoustic frequency response/impulse response of speakers & mikes?
Electrical stuff is easy with RMAA and other stuff.

The nearest I know of is Professor Angelo Farina's Aurora plugins to Audition.  Not exactly free but the Audition license is the main expense.

I use MARS, a programme by CRC, (Canada) for impulse  responses and my own klunky DOS software for frequency response and signal processing to get a quasi anechoic measurement.

For commercial stuff, I like Clio; especially ClioQC for production testing.  Their inexpensive measurement mikes are one of the few whose calibration I trust.

[Church-Audio]

Ok I will tell you what part of the gear that I use to test the mics is a burr brown 1k oscillator its a chip that is encased in epoxy. It has one of the lowest distortions available for a 1k oscillator and I couple that with a Mylar driver that is mounted in a machined housing. But its not something off the shelf. I use a 4.7k resistor in almost all of my mics and it works provided you have approximately 10k or so load impedance and a 9 volt supply.

Thanks for this Chris.  Now everything is clear.

I will say that you are probably the only builder to systematically tests their gear.  I have only recently started testing (PC with EMU 1212m soundcard, RMAA program), and it is certainly more tricky than I imagined.

Richard (and anyone else), do you know of free/inexpensive software/gear to measure acoustic frequency response/impulse response of speakers & mikes?
Electrical stuff is easy with RMAA and other stuff.

The nearest I know of is Professor Angelo Farina's Aurora plugins to Audition.  Not exactly free but the Audition license is the main expense.

I use MARS, a programme by CRC, (Canada) for impulse  responses and my own klunky DOS software for frequency response and signal processing to get a quasi anechoic measurement.

For commercial stuff, I like Clio; especially ClioQC for production testing.  Their inexpensive measurement mikes are one of the few whose calibration I trust.

The best software out there IMO is winaudiomls I have the lab edition with all the plugins.. but you dont need to get that elaborate. I also use the emu 1212m soundcard as my measurement interface. Here is a link to an ebay auction for the basic version of the software I use. I have over $3000 in software from Winaudiomls and I can honestly say they do update and fix any bugs that I find with in 24 hours.. thats pretty impressive..
check this link from time to time he has deals on ebay for his software. http://www.dr-jordan-design.de/Special_offers.htm

http://www.dr-jordan-design.de/complete.htm

[Roger Gustavsson]

There's a basic misunderstanding going on between micbuilders and here . . . they are talking about a common drain configuration, Chris is using common source.

You are right.  I was assuming the TSB120 used as in their literature; common drain / source follower.

GHW, you mention a Church type 2 wire connection.  Is there a link to this?

Chris, if it isn't too secret  , is your recommendation, "replace 2k4 with 4k7", to do with an external load resistor for common source PiP operation?

I was under the impression you were advocating replacing the internal (SMD on tiny PCB behind the capsule) 2k2.

Is this what we are comparing? Inside the red line is what Church is doing and inside the green the standard schematic. The series capacitor and the 10k resistor represents the input of the preamp.



Roger

[Church-Audio]

There's a basic misunderstanding going on between micbuilders and here . . . they are talking about a common drain configuration, Chris is using common source.

You are right.  I was assuming the TSB120 used as in their literature; common drain / source follower.

GHW, you mention a Church type 2 wire connection.  Is there a link to this?

Chris, if it isn't too secret  , is your recommendation, "replace 2k4 with 4k7", to do with an external load resistor for common source PiP operation?

I was under the impression you were advocating replacing the internal (SMD on tiny PCB behind the capsule) 2k2.

Is this what we are comparing? Inside the red line is what Church is doing and inside the green the standard schematic. The series capacitor and the 10k resistor represents the input of the preamp.



Roger

actually thats not what I am doing.

[Roger Gustavsson]

Sorry Chris!

Should have put it as a question... What about these schematics?





Roger

[Church-Audio]

Sorry Chris!

Should have put it as a question... What about these schematics?





Roger

Thats it. Distortion does go from 8% to 0.5% with this mod.

[Roger Gustavsson]

Thank you Chris!

[scottwu]

I measure my distortion with a FFT analyser and a super low distortion sound source directly coupled to the capsule ...

Is this a B&K 4231 Calibrator?

bv

I dont want to give away all my little secrets.... The name and make of the device shall remain my little secret but its not a regular calibrator if that is what your asking. 

Just my opinion, but I don't like this secrecy.  I've said this before, and I'll say it again, there is lots of room for both vendors and diy types here.  Hiding this information does no good.  People in the know (competitors) will figure out how to do this anyway.  You (Chris) are offering a great product at a great price.  And you're constantly innovating, with new products that we (think we) need.  That is what sells your product.  IMO this is what puts you above others (like Soundpros, Core sound, and sonicstudios, for example).

  Richard

Just as a comment you can DIY a small Hi-Q piston chamber and get some very low distortion at very high SPL (without ear damage . Exciting it at resonance gives excellent extra rejection of harmonics. I used a 2.5" Yamaha driver scavenged from a computer speaker and a 3-4" piece of heavy mailing tube. It measured <.1% at 140dB. I also used a medical grade MEMS pressure transducer as a measurement device. It makes a VERY insensitive and fairly low frequency device (15 microV/Pa, <3kHz) but it is calibrated to the equivalent of .1dB, has response to absolute DC, and using it way below its rated range it is inherently very linear. I was able to easily measure several mics all the way to hard clipping. Obviously this is only for omni's.

The second way to measure THD is to use two tones and two speakers and derive the mic's transfer function from the IMD spectrum which can only come from the mic (pretty elegant). This is documented in a couple of AES papers.

As for commercial interests polluting DIY, Chris is definately one of the good guys with the white hats.