Doesn't that in fact prove precisely what @kingkorg was saying? That sensitivity depends DIRECTLY on the voltage you POLARIZE it with? As in, the capsule on its own, without a bias voltage, has NO INHERENT sensitivity?
Soliloqueen's k87(k67) and k47 capsules
- Thread starter kingkorg
- Start date
Help Support GroupDIY Audio Forum:
Doesn't that in fact prove precisely what @kingkorg was saying? That sensitivity depends DIRECTLY on the voltage you POLARIZE it with? As in, the capsule on its own, without a bias voltage, has NO INHERENT sensitivity?
I could see how reference sensitivity at a known voltage could be useful. It's just like anything else that has an input voltage. Technically you could run a motor at any voltage but many of the specifications of the motor are given at a specific reference voltage. I wasn't aware that capsules were specified this way but it's not unreasonable.Doesn't that in fact prove precisely what @kingkorg was saying? That sensitivity depends DIRECTLY on the voltage you POLARIZE it with? As in, the capsule on its own, without a bias voltage, has NO INHERENT sensitivity?
I believe that the reference sensitivity measured under standard established conditions is an important criterion for evaluating the capsule.
Attachments
This chart says that the sensitivity was measured at 47 volts polarization which is unusual
I agree, but there doesn't seem to be such standard. The paper you attached has nothing to do with that specific capsule. Why, i have no idea. It doesn't state if the noise of the amp is taken into account. They do however provide accurate measurements with their samples, but not all the conditions under which they were taken. No mention of polarization voltage either. Not to mention the sensitivity is also dependent on frequency.
If you are into weeds of this kind it is presumed you are able to do these tests yourself. U87 and U67 sensitivity figures are not limited by the capsule. Passive capacitance of a known capsule design should be more than enough to indicate it's noise performance.
Last edited:
![Electronics Soldering Iron Kit, [Upgraded] Soldering Iron 110V 90W LCD Digital Portable Soldering Kit 180-480℃(356-896℉), Welding Tool with ON/OFF Switch, Auto-sleep, Thermostatic Design](https://m.media-amazon.com/images/I/41gRDnlyfJS._SL500_.jpg)
It depends. Over a wide range the output voltage goes linearly with polarisation voltage. A critical polarisation voltage exists. Above this value the capsule shows a strange behaviour can collapse sometimes triggered by a loud acoustic signal. Therefore it is important for a manufacturer to indicate the max. allowable pol. voltage. For electret capsules life is easier..
Sure, max bias voltage would be a nice-to-know parameter, but that still doesn't mean a(n externally polarized) capsule has "a" certain inherent sensitivity.
Thread 'How to fry a microphone capsule' How to fry a microphone capsule
Kingkorg,
Did you do this extreme test at 125V in a Rode NT1 kit (black)?
If so, then has it just become the lowest self noise microphone on the planet?
Kingkorg,
Did you do this extreme test at 125V in a Rode NT1 kit (black)?
If so, then has it just become the lowest self noise microphone on the planet?
thinking of doing charted matched pairs next batch. you guys don't mind if we don't have FR comp for our speaker response yet, right? you understand that you can't tell how the capsule sounds from an uncompensated QC graph, only how close the match is
edit:
the reason we don't save the charts for every capsule is that our DAAS license is for windows 95, so we had to salvage a windows 95 shitbox to do our floor FR tests. it's got an awful CPU and it hangs when you open the save dialogue(not permanently, but for like 15 seconds so it's a pain in the ass). the system is so old that it saves the charts in pcx. we can only afford the time loss of saving charts for matched pairs (even then, only maybe). eventually i will build a pc and install like, windows fundamentals for legacy pcs on it and migrate the installation over, if that's possible. if not i'll have to crawl ebay to build a faster rig out of real period parts but until then we can only do MP charts. ideally we'd migrate to arta or rew, but that's a big procedure change. at least this time i can verify we're actually doing it unlike with the factory. still pissed about that.
edit:
the reason we don't save the charts for every capsule is that our DAAS license is for windows 95, so we had to salvage a windows 95 shitbox to do our floor FR tests. it's got an awful CPU and it hangs when you open the save dialogue(not permanently, but for like 15 seconds so it's a pain in the ass). the system is so old that it saves the charts in pcx. we can only afford the time loss of saving charts for matched pairs (even then, only maybe). eventually i will build a pc and install like, windows fundamentals for legacy pcs on it and migrate the installation over, if that's possible. if not i'll have to crawl ebay to build a faster rig out of real period parts but until then we can only do MP charts. ideally we'd migrate to arta or rew, but that's a big procedure change. at least this time i can verify we're actually doing it unlike with the factory. still pissed about that.
Last edited:
I think it's quibbling over terminology, and different senses of "sensitivity," but a sensitivity to the product of sound level * polarization voltage seems like a well-defined and potentially useful thing to know when thinking about capsules and circuits. If you know how that polarization-voltage-dependent "sensitivity" is different from one capsule to another, you can predict the change in output level if you do a normal capsule swap, while keeping the circuit and its polarization voltage constant, or a circuit swap while keeping the capsule constant.
Matt Nolan
Well-known member
I don't know if you can still do this or find the right files, but I used to run Win95 inside a virtual machine (just to use an old scanner for which there were not any newer drivers and one old piece of photo editing software). Might be worth investigating rather trying to build a legacy machine in the real world. To get files in and out I had to use external drives, which was bit clunky, but it was supposed to be possible to get the VM to see the drives of the host machine, I just couldn't make it work.
Couldn't you just create a win95 system inside a virtual machine on a modern pc? Or would that be wishful thinking on my part...
Ah I see @Matt Nolan had the same idea.
We would still need hardware compatibility and working driver pass through for the hardware parts of the daas system. This would take some experimentation to see what legacy connector add-on cards are available for modern computers, which would work in pass through to the virtual machine invisibly, etc. it could certainly be done and it's something I would consider but it wouldn't be quick to set up unless I was physically there.
took a break from QC to finish packing and shipping every domestic preorder myself, so almost everybody in the US should be getting their flat k47 capsules by saturday. international orders and replacements for the previous batch will follow.
there's one thing i need some feedback on: very long term stability. Can you guys note any tuning drift or capsule burn in over the first 3, 6 and 12 months? and then at 5 years. if they drift, replacement is also free. pretty sure i have this dialed in, but hard to test synthetically
there's one thing i need some feedback on: very long term stability. Can you guys note any tuning drift or capsule burn in over the first 3, 6 and 12 months? and then at 5 years. if they drift, replacement is also free. pretty sure i have this dialed in, but hard to test synthetically
some charts from this batch for transparency:
here it is zoomed in a bit more vertically and without any smoothing:
those nulls at 6 and 9k are from room-related junk in the rear lobe. this room is not fully treated acoustically so there are some caveats like this here. All in all, knowing how these capsules are supposed to look in my setup, these units are performing pretty well! listen tests are good. Some of them are slightly sparky, but that's not really a bad thing. Pretty pleased.
edit:
and (pink) vs kingkorg's measurement of one (orange) and a neumann k47 (blue)
other than the aforementioned nulls, the measurement is basically spot on. That being said, we don't have the volume and procedure benefits of a factory, so unit to unit consistency may be slightly lower than before, especially on the K87. Not super low, but not "assembly line" precise matching like at the factory. There may be slightly brighter and darker k87s. Maybe you guys could trade units with each other to get what you like. I've yet to personally see one more than +-1dB though.
here it is zoomed in a bit more vertically and without any smoothing:
those nulls at 6 and 9k are from room-related junk in the rear lobe. this room is not fully treated acoustically so there are some caveats like this here. All in all, knowing how these capsules are supposed to look in my setup, these units are performing pretty well! listen tests are good. Some of them are slightly sparky, but that's not really a bad thing. Pretty pleased.
edit:
and (pink) vs kingkorg's measurement of one (orange) and a neumann k47 (blue)
other than the aforementioned nulls, the measurement is basically spot on. That being said, we don't have the volume and procedure benefits of a factory, so unit to unit consistency may be slightly lower than before, especially on the K87. Not super low, but not "assembly line" precise matching like at the factory. There may be slightly brighter and darker k87s. Maybe you guys could trade units with each other to get what you like. I've yet to personally see one more than +-1dB though.
Last edited:
JeffKall
Well-known member
Hey, Ari …
I just got notice yesterday that my two flat k47’s have shipped … and I am ecstatic !
I know they will be well worth the wait … and that they will definitely elevate the sound of my two D-EF47’s from Dan …
And whenever the two flat k87’s ship … they will likely do the same for Dan’s two D-67’s !!!
Exhilarating times for all of us awaiting the arrival of our Arienne flat k47 capsules !
Thank you, Ari, for all your hard work, pragmatism and determination to take the time to do them right !!!
Best Regards
Last edited:
Matador
Well-known member
There seems to be an analog here with speakers, which publish "sensitivity" at 1m distance with 1W input power. I would think that even if you fixed the polarization to a known standard value, it would be difficult to calibrate the excitation source from test to test. Like, is there a known "Pascal" standard sound source and/or "test condition", that generates 1 Pascal under a standard condition?
Sound calibrators (1Pa / 94dB SPL @ 1kHz) -> measurement microphone (with a known sensitivity), the output of which can be compared with the microphone-under-test (assuming a known fixed-gain of the circuitry).
Similar threads
