Personally I doubt that something cheaper than ACO will actually meet with the quality of your other test equipment.
Back 45+ years ago, I had another problem. The only distortion measurement test equipment that I had was a Heathkit IMA, a Heathkit audio oscillator, and a Donner (vacuum tube) wave analyzer. Still, I needed a pretty good microphone to measure the distortion of individual drivers for the Wall of Sound we were designing for the Grateful Dead. I was fortunate in finding this B&K mike capsule, but it needed a preamp to make it whole. Still, I'm sure that my actual cost today would be between $500 and $1000 given the inflation over the years. Like any good tool, when you find something that actually works well, be it a hand tool, a loudspeaker, or even a measurement microphone, you tend to keep it indefinitely, because it doesn't really go obsolete like electronics does. Today, I have given away both my IMA and my wave analyzer, and I bought better test equipment over the decades. But I still have my measurement microphone, that I am sure still works OK for most anything, except extremely high frequencies, above 18KHz. The newer mikes aren't really much better.
Back 45+ years ago, I had another problem. The only distortion measurement test equipment that I had was a Heathkit IMA, a Heathkit audio oscillator, and a Donner (vacuum tube) wave analyzer. Still, I needed a pretty good microphone to measure the distortion of individual drivers for the Wall of Sound we were designing for the Grateful Dead. I was fortunate in finding this B&K mike capsule, but it needed a preamp to make it whole. Still, I'm sure that my actual cost today would be between $500 and $1000 given the inflation over the years. Like any good tool, when you find something that actually works well, be it a hand tool, a loudspeaker, or even a measurement microphone, you tend to keep it indefinitely, because it doesn't really go obsolete like electronics does. Today, I have given away both my IMA and my wave analyzer, and I bought better test equipment over the decades. But I still have my measurement microphone, that I am sure still works OK for most anything, except extremely high frequencies, above 18KHz. The newer mikes aren't really much better.
I think we should move this microphone stuff to a different thread.
I would be happy to make some measurements and share them of a handful of microphones. I just checked the mike preamps I have for the condenser microphones. The B&K 2639's have at least -96 dB HD2 and about -110 dB HD3 at the equivalent of 100 dBSPL with a 1" microphone (about 100 mV in). My two vacuum tube B&K's have about -60 dB HD2. I will be selling the vacuum tube preamps so if there are any questions that I should explore please bring them up before they are gone.
Is there any consensus on IM testing of a microphone being useful for predicting HD? Further should those IM sources be HF or lower frequency? I was thinking of using 250Hz and 350 Hz from two separate speakers however I need some validation since this stuff takes a lot of time.
I suspect the bulk of the "measurement microphones" of the Behringer ECM8000 type all come from one Asian vendor so maybe I'll get one of those and I have capsules from Primo and Panasonic (discontinued for several years) so once setup I can run several through the testing.
I would be happy to make some measurements and share them of a handful of microphones. I just checked the mike preamps I have for the condenser microphones. The B&K 2639's have at least -96 dB HD2 and about -110 dB HD3 at the equivalent of 100 dBSPL with a 1" microphone (about 100 mV in). My two vacuum tube B&K's have about -60 dB HD2. I will be selling the vacuum tube preamps so if there are any questions that I should explore please bring them up before they are gone.
Is there any consensus on IM testing of a microphone being useful for predicting HD? Further should those IM sources be HF or lower frequency? I was thinking of using 250Hz and 350 Hz from two separate speakers however I need some validation since this stuff takes a lot of time.
I suspect the bulk of the "measurement microphones" of the Behringer ECM8000 type all come from one Asian vendor so maybe I'll get one of those and I have capsules from Primo and Panasonic (discontinued for several years) so once setup I can run several through the testing.
I'm on this. I own a Gefell MK301 1/4" capsule with 1/2" adapter that I assume is a reasonable baseline for mic distortion. Just need to build a new buffer for it, as the cheep chinese B&K 2639 + supply/gain-stage knock-off I also have is junk (it works basically but is so extremely microphonic to be unusable in practice, also I don't know if they copied the circuit properly, there is a lot of hum and ULF-drift/noise in the output signal).
With this I should be able to compare to the performance of the cheap Transsound capsules (with a variant of the bootstrapped buffer Scott Wurcer shared here).
With this I should be able to compare to the performance of the cheap Transsound capsules (with a variant of the bootstrapped buffer Scott Wurcer shared here).
I found another reference from the same 2012 conference that you should look at, this reads directly onto my original reference (Wyberg) and my results. They quote some calculations using non-linear acoustics to get a more accurate model which looks almost exactly what I measured. Their figure 4 is almost exactly what I see (with different time constants to fit my spark and mic). They have some info and measurements that you might find useful.
https://hal.archives-ouvertes.fr/hal-00810828/document
There is need to know the thd/harmonics of the DUT --- without knowing the thd of the measuring system is a No-Go for me.
So. here is some useful data for you to think about.... The 1/2 inch diameter mic made under the Dayton and other names has harmonics in the 0.1 - 0.25% range at 90db SPL.
0.1% at 500Hz region.
This is marginal, but price is right, so I think we could use them for most things related to dynamic speaker/driver design and mods.
Thx-RNMarsh
[Dayton OmniMic, EMM6, ECM-8000, MiniDSP, etc]
So. here is some useful data for you to think about.... The 1/2 inch diameter mic made under the Dayton and other names has harmonics in the 0.1 - 0.25% range at 90db SPL.
0.1% at 500Hz region.
This is marginal, but price is right, so I think we could use them for most things related to dynamic speaker/driver design and mods.
Thx-RNMarsh
[Dayton OmniMic, EMM6, ECM-8000, MiniDSP, etc]
Last edited:
This is the most important thing, the other mic stuff is probably a distraction here. I certainly understand how tedious it can be taking measurements especially when you have to apply your own judgement on the methodology and possible confounders.
Isn't IM just a different representation of the one and same non-linearity as HD??? So why to worry?
The electrets seem to be fine for this ??
See # 12665 above.
If the dynamic driver mods leave us with <0.1% then we have to go to the more expensive mic's.
But then, if it is <0.1% maybe we dont care below that?
Anyway, one of those electret's is what i will be using for now.
-RNM
Demian -- if you want to measure IM on electret microphones, One of the cheapest is the EMM6 spec'ed at 1% THD at 127db SPL.
See # 12665 above.
If the dynamic driver mods leave us with <0.1% then we have to go to the more expensive mic's.
But then, if it is <0.1% maybe we dont care below that?
Anyway, one of those electret's is what i will be using for now.
-RNM
Demian -- if you want to measure IM on electret microphones, One of the cheapest is the EMM6 spec'ed at 1% THD at 127db SPL.
Last edited:
http://www.daytonaudio.com/media/resources/390-801--dayton-audio-emm-6-quick-reference-guide.pdf
VERY flat FR. Well under 100 USD.
-RNM
Then buy the Dayton and take a chance, Richard. Following the guidelines that I have already stated, this microphone should have low enough distortion at 100-110 SPL for most measurements. The Dayton company most probably has already done the distortion comparison with a reference mike. Just use the highest voltage specified to power it, to keep the electronics linear.
IM in this case is just an easier way to test for distortion in a microphone. Linear sources are costly, so if you use 2 separate sources, then their individual non-linearities that generate their harmonic distortion from a single tone can be ignored (on a graph) and the microphone's distortion estimated. It just gets more confusing to use it for any other reason.
Last edited:
I dont need to buy one, I already own it and several others similar to it. But other DIY's might be interested in buying it to do their own testing and driver mods. .
-RNM
Well, since no one else solved the question of what is the distortion of the microphone or the "problem", I researched it further myself. As was pointed out, the capacitor mic could be low enough in distortion though very expensive.
I found what I have will work at low cost and reported that here for others interested.
Now we can continue with JN's suggestions on how to lower the THD of drivers.
THx-RNMarsh
Last edited:
Drivers have arrived (Monacor SPH-135TC). Quick test shows that the coils are not perfectly matched, certainly they are not bifilar wound. One coil is definitely hotter (higher resulting BL) and I will use this coil for the JN-feedback with a bit of attenuation to get the matching. Exact amount will have to be determined by measuring the back-EMF voltage with external exitation and trim the attenuation for deepest null.
Also of concern is the positional dependence on inductance, as they are probably different.
jn
Speaker time response
I hope this is something that John C might like to see 😉. It is a 2-way speaker response to a 454us wide rectangular impulse. Bottom trace is the el. impulse, top trace is the speaker&room response. Time delay about 5ms is the time needed for sound to travel from the speaker to the microphone.
I hope this is something that John C might like to see 😉. It is a 2-way speaker response to a 454us wide rectangular impulse. Bottom trace is the el. impulse, top trace is the speaker&room response. Time delay about 5ms is the time needed for sound to travel from the speaker to the microphone.
Attachments
- Status
- Not open for further replies.