Where to find genuine WM-61A capsules - or a better alternative?

[gentlevoice1]

Hi all,

It has been a while since I have visited the forum - good to see that things appear to be well here

My reason for posting right now is that I am in the process of making a measurement setup for headphones and to that end am looking for some genuine WM-61A capsules. Alternatively a suitable electret microphone replacement with similar or better frequency response (an extended & linear HF response is important), similar or better SNR, and similar or better distortion specs (as compared with a Linkwitz modded WM-61A; https://www.linkwitzlab.com/images/graphics/microph1.gif).

Any chance one of you may know where to obtain genuine WM-61As - or a similar/better alternative? If possible I would like to go down to 0.005% THD at ~90 dB SPL ...

Cheers & thanks for considering,

Jesper M

 

[Khron]

What's your budget?

 

[gentlevoice1]

Hi Khron ... & thanks for replying. As to your question I reckon it depends on what is available. But for good electret capsules I would be ok with EUR 50 per capsule - do you have anything particular in mind?

Jesper M

 

[Khron]

I was just thinking that, in the grand scheme of things, the measurement mic i got myself a few years back isn't even all that pricey, at about 200e or so (calibrated, both as graph and as text / frd file). But depends how "stuck" you are on DIY-ing it or not

 

[gentlevoice1]

Hmm ... I need it for headphone measurements - so I need two and they should be small size so they can be fitted inside a model head. For best quality still at a reasonable price I have not seen an alternative to DIY'ing ... but maybe I have missed something?

Cheers, Jesper

 

[Khron]

Is it mandatory to measure both sides of the headphones at once?

 

[rock soderstrom]

Hi Jesper, I got two of them. I bought them over 10 years ago from a serious local source. I do not need them, make me an offer.

 

[gentlevoice1]

Hi both ;-) ...

@Khron: That is a good question ... I reckon there could be something related to structural vibrations shifting to the other side - but you may have a point. In any case my feel is that I'd like two sides - just in case. May I BTW ask which microphone you have in mind - maybe knowing this would ease assessing whether it could be feasible ...

@rock soderstrom: Thanks for the offer! I may be interested - just think I will wait a little bit to see if there may be a suggestion for a newer/better microphone. Please hold them (admittedly smiling here as you have had them for 10 years) and I will get back to you.

Cheers, Jesper

 

[Khron]

Hi both ;-) ...

@Khron: May I BTW ask which microphone you have in mind - maybe knowing this would ease assessing whether it could be feasible ...

Cheers, Jesper

I don't mean to advertise, and i have no affiliation with them; they just happened to have the cheapest calibrated solution i could find at the time, in Europe anyway.

https://www.isemcon.net/ashop/index.php?cPath=2_3&MODsid=1cfab6ecb2a5aa022a5c2bbd957572f8
I got the EMX-7150, and while i haven't used it particularly extensively, i'm really happy with it, and there's that peace-of-mind factor somewhere in there too.

 

[gentlevoice1]

Hi Khron - thanks for the link! ... It looks fine but in my context there would be three "challenges":
- one is that I already have a couple of BK 4133 capsules and if I were to measure e.g. loudspeakers I would use these capsules with appropriate electronics.
- The second would be this capsule's distortion level. I would like to be able to go down to 0.005% THD at ~90 dB SPL (time will show if this is possible) and this capsule's distortion level is somewhat higher.
- Third there is the price. Since I can make the ancillary electronics myself - and likely will prefer to be able to customize how it works - I prefer to pay only for the capsule.

So, all in all, I appreciate that you shared which microphone you were thinking of so that it could be clarified that I am looking for a different microphone "concept". And you made me reconsider my specs for this capsule - thus I will add the distortion level wish to post #1.

In any case thanks for considering and replying

Jesper M

 

[k brown]

Best current substitute for the WM-61 is the Primo EM273; plus it doesn't need the Linkwitz mod, as it's already three-wire. Waay less self-noise than the Panasonic, too.

https://micbooster.com/microphone-capsules/209-primo-em273-microphone-capsule.html

 

[gentlevoice1]

Hi k brown ... Thanks also for your feedback and link! Looks indeed interesting, not least if I can correct the HF response - I understand this should be possible e.g. in REW.

BTW I happened to find this thread discussing how to maybe improve on the EM273:

https://www.audiosciencereview.com/...or-a-measurement-mic.10079/page-3#post-294808
Cheers,

Jesper M

 

[kingkorg]

You need a capsule that is calibrated and/or comes with calibration file. Using just something like random wm61 won't get you there as these can vary a lot. Maybe just buy two (used?) mics like umik and use the capsules, or even the electronics.

On the other hand, head model, ear canal will alter the response of the capsule significantly, to the point you might wonder if you even need a calibrated capsule, or just a stable reference.

Headphones are difficult, so many resonances, goofy responses, honestly i'm not even sure you need anything near flat. Just flat-ish so you can compare headphone models to each other.

Also no point chasing better noise spec for measurements, it doesn't affect the measurement, especially not at 90db. It's either low thd or low noise. You can't have both.

You won't find anything close to 0.005% THD. Simply because THD spec is published for electronic part of the capsule/circuit. As far as i know no capsule manufacturer provides THD values of the capsule itself, i mean the THD caused by the mechanics of the capsule, not electronics. They will always be higher than 0.005%. Also don't get me wrong, no one on earth can hear even 0.1%, if nothing else simply because your ear will distort above that. But i'm sure an audiophile would argue that.

 

[gentlevoice1]

Hi kingkorg,

Thank you also for your feedback. And, yes, it is also my impression that headphone measurements are more "diluted" in terms of precision than other audio measurements but nevertheless I prefer being on the safe side so to speak.

It's either low thd or low noise. You can't have both.

Hmmm ... a bit puzzled about this one ... I would reckon that a large microphone membrane would have both low self-noise and low distortion (everything else being equal) as the movement of the membrane would be less for the same output - and in general large membranes are less self noise (relative to their output level). But you likely know more about this than I do ... but would it mean that the WM-61A (likely) is lower thd than the primo 273?

no one on earth can hear even 0.1%

The "challenge" here IMHO is that measurements and the human hearing do not really correlate. As an example I have made electronic audio designs which essentially measure completely identical, however, their sound is like day & night. Typically, the "challenge" - again IMHO - is that when signals become complex, like in music, high(er) distortion levels typically will also mean high intermodulation distortion and then the distortion levels become significant - at least to my ears. Anyway, I realize this is an on-going debate and I do not really wish to open it here - I hear what you are saying and am fine with that

Have a fine day, Jesper M

 

[kingkorg]

Hmmm ... a bit puzzled about this one ... I would reckon that a large microphone membrane would have both low self-noise and low distortion (everything else being equal) as the movement of the membrane would be less for the same output - and in general large membranes are less self noise (relative to their output level).

It is important to distinguish if we are talking about THD comming from the capsule itself or the impedance converter that follows. As we don't know exact figures of capsule distortion it is best to leave it alone for now. I am pretty sure though it is higher than 0.005 at 90db.

As for the impedance converter THD, which is specified in the sheets, the larger capsule or the more polarization voltage, the more output, the lower noise but more signal that drives the impedance converter and hence the higher THD.

This is off course generalization, you can design a high voltage circuit, maybe with some bootstraping, cascode, tube cathode out, or even opamp at first stage to minimize the effect. But we are talking about capsules with built-in fets, some don't even have high ohm resistors but diodes for gate biasing. Not good THD wise.

 

[gentlevoice1]

Hi again ... & thanks once more for your feedback.

It is important to distinguish if we are talking about THD comming from the capsule itself or the impedance converter that follows. As we don't know exact figures of capsule distortion it is best to leave it alone for now. I am pretty sure though it is higher than 0.005 at 90db.

Just to make sure I understand you correctly: What you are saying is that the electret microphone capsules that have been mentioned in this thread would all have distortion levels that are higher than 0.005% at 90 dB SPL - is that correctly understood? Then, if so, it would mean that if I wanted a lower distortion figure I would need to choose another capsule, right?

As far as I can gather the BK4133 capsules (I have two of these) would be lower than 0.005% THD at 90 dB SPL ... They are 10% distortion at 165 dB SPL and 3% distortion at 160 dB SPL. So - if this can be assumed to be a linear relationship (?) - they would be about 0.003% at 130 dB SPL. A completely different magnitude of distortion relative to the other capsules mentioned here. I am wondering if I am missing something ... ??

Anyway, I digged a bit deeper into possible low distortion circuitries (for measurement) and happened to come across the circuitry inserted in this post (courtesy Scott Wurcer). At 50mVpp, 1 kHz sine the distortion level is appr. - 144 dB (2H) for the circuitry in the middle. Good enough for my purposes.

Also the "regular" phase splitter circuitry is shown at the bottom in the attachment. With adjusted R16 & R17 it has a 2H distortion level of appr. - 114 dB (2H). May be better for recording purposes, I assume, as it can be made without capacitors (Vout+ actually distorts a little less than the balanced Vout+ & vout- combined) ...

Cheers, Jesper M

 

[kingkorg]

As far as I can gather the BK4133 capsules (I have two of these) would be lower than 0.005% THD at 90 dB SPL ... They are 10% distortion at 165 dB SPL and 3% distortion at 160 dB SPL. So - if this can be assumed to be a linear relationship (?) - they would be about 0.003% at 130 dB SPL. A completely different magnitude of distortion relative to the other capsules mentioned here. I am wondering if I am missing something ... ??

That sounds right to me, but what is the noise spec of this capsule since it takes so high SPL? BK measurement capsules are very different to wm61 and other commercial capsules. They have very high polarization voltage and large spacing between diaphragm and backplate, wich helps with high SPL.

Anyway, I digged a bit deeper into possible low distortion circuitries (for measurement) and happened to come across the circuitry inserted in this post (courtesy Scott Wurcer). At 50mVpp, 1 kHz sine the distortion level is appr. - 144 dB (2H) for the circuitry in the middle. Good enough for my purposes.

50Vpp is not much, large diaphragm capsules that put out several volts of signal. You would have to know how much a specific capsule puts out, and this will depend on polarization voltage or how much the electret is charged. I don't know how he made the conclusion 50mVpp is 144db. That doesn't sound right to me.

 

[gentlevoice1]

Good morning ... a quick feedback on your post yesterday ...

but what is the noise spec of this capsule since it takes so high SPL?

According to the BK datasheet the thermal noise is specified to be 20 dB (A-weighted). Thus actually only a SNR of ~70 dBA relative to a 90 dB SPL. Not too impressive IMO ...

As it is I also have a couple of Gefell Microtech MK221 capsules and they have a 15 dBA noise level re. 94 dB SPL, i.e. 79 dBA SNR ... Well, better, but I really would like to avoid the 200 VDC polarisation voltage for headphone measurements ... which also is the reason I was interested in the WM-61A in the first place ...

I don't know how he made the conclusion 50mVpp is 144db.

I hasten to add that the - 144 dB (2H) is from an LTSpice (FFT) simulation of this circuitry that I made yesterday with a 50 mVpp signal level. I chose this signal level because it would be relevant for me in relation to measuring headphone distortion. However, a 1 Vpp signal level gives appr. - 118 dB 2H - which IMHO still is very good considering it is a very simple circuitry with no (global) feedback.

The circuitry at the bottom yields appr. - 78 dB 2H at 0.66 Vpp input level (appr. 1 Vpp output level).

BTW I found an internet site where they had tested the WM-61A at 100 dB SPL with a couple of sennheiser headphones and the distortion level had been 0.1%. They assessed that most of the distortion came from the headphones .. yet the dynamic headroom of the wm-61A had been estimated to be 68 dBs ... just FYI should someone be interested.

Anyway, with the feedback in this thread I guess I will have to consider what to do. I reckon the Primo EM273, which k brown suggested, would be a feasible option and to this end I have sent an email to micbooster asking them if they would/could supply FR curves, or a calibration file, with their capsules. I will post their reply here.

Thanks again to all of you for helping out with this ....

Cheers, Jesper

@rock soderstrom: Given what I have learned here it seems that I cannot use the WM-61A capsule for my purposes. So I won't be needing your capsules - but many thanks for making them available!

 

[gentlevoice1]

Hi ... just a quick follow-up on my contact with micbooster: Their matching is only for levels at 1 kHz sine and they do not provide any further documentation in this context. Cheers, Jesper