Hello
I just found this thread... I have PUI 5024 electrets wired to a Zoom H1n. My objective with the 5024 was a low noise high sensitivity electret. It is just fine for that. My question is about the bandwidth. I have used them to make audio measurements. The frequency responses I get have a cut off from about 5kHz or just a bit higher. To try to find where is the problem, I have tested the H1n mic input with an attenuated signel with REW, it is ok. I tested also the H1n with a Dayton Audio iMM6 mic, it is ok also.
So my question : does somebody know a possible HF cut off with the PUI 5024 (the spec says 20-20k)? a possible effect of the low voltage of the H1n mic input (around 2.5V)?
Writing this question let me think I should make a test with a wide range loudspeaker (or a tweeter) and compare the mic I have (umik1, iMM6, PUI 5024 and some other low cost electrets).
Thank you for your help
Did you use a long and cheap cable? With 40 metres of cheap PVC-dielectric cable, it could just be the cut-off frequency of the 2.2 kohm pull-up resistor and the cable capacitance.
5K is low, I have not had any problem like that with any mic capsule. A telephone receiver, yes. 😉
Like Marcel, I suspect that something in your wiring is making a low pass filter. The PUI capsules work fine on 2.5V, they just won't have as much headroom as running a little higher. 2.5V does not affect the frequency response a far as I can tell. Some of the larger PUI capsules don't get all the way to 20 kHz, but they certainly have no problem at 5 kHz.
Like Marcel, I suspect that something in your wiring is making a low pass filter. The PUI capsules work fine on 2.5V, they just won't have as much headroom as running a little higher. 2.5V does not affect the frequency response a far as I can tell. Some of the larger PUI capsules don't get all the way to 20 kHz, but they certainly have no problem at 5 kHz.
+ @MarcelvdG
Thank you for you quick feedback. Unfortunately no long cable to explain a capacitance across the mic terminals, neither an added cap.. To understand that this cutoff shouldn't happen is in itself an information and the PUI remains then a good choice. I will try to inform you after some comparison tests with other mics I have. I have 2 stets of those mic, for one, the capsules are sold at the end of one 3.5mm jack. I have 2 other types of electret assembled in the same way. Easy to exchange.
Reading those threads about mic lead me again to The sound Sleuthers page. I wonder if I have wired the mic exactly as in this page : I don't remember if I connected or not the shield of the cable at the mic side. According to your experience, does it matter?
Sorry, one additional question : I bought one H1n which I find handy but I was badly surprised by the "poor" noise performance when the knob gain is pushed a bit. I prepared a preamp powered on a 9V battery with some transistors but not built for now... does it make sense this preamp idea? Any advice for an other recorder?
Thanks again.
I'm not having noise problems with the PUI HD capsules and the Zoom H4n or a Tascam. With the little Sony recorder I recently bought noise is rather noticeable, but that's the recorder. On some of my builds I've used thin Mogami cable and it never picks up hum, buzz or any other noise. But I have also used just plain old non-shielded earbud wires with no noise problem. Of course they are only about 4 feet long, but using the Zoom H4n input the wire had to be laid across a PSU transformer to pick up any noise. With short cables and a low(ish) input impedance, mic noise had not been a problem for me, especially with the high output of the PUI HD series. With the old Panasonic capsules, maybe just a little.
Your preamp might help if going in line level is quieter on the H1n than mic level.
Your preamp might help if going in line level is quieter on the H1n than mic level.
I looked again and found that I did use the 5024L in early experiments. I remember it rolling off above 10K, but there was some response higher than that. Not much there in the top octave, to be honest. I used only unbalanced wire.
If I understand you correctly, you connected the microphone capsule to an unbalanced input with a balanced cable, connected the two internal wires as intended, but possibly connected the shield only on one side. That may have an impact on hum and susceptibility to RF interference, but not high-frequency roll-off.
Do you have any specifications related to the recorder noise floor?
For what it's worth, some time ago, I helped a field biologist who uses PUI microphones and a cheap sound card for a bird monitoring project. In his case, the noise floor could be lowered by about 2 dB with an extra preamplifier, if I remember well.
Your understanding is correct. Clear for me the consequence. I will check the wiring and probably correct it if wrong.
I just check again but I have found nothing about the noise from the specification. I found this Avisoft site showing noise evaluation of recorders. It seems they use a 150Ohm load instead of the mic. I haven't fully understood for now the conversion in dBu... probably by the clipping voltage. is it something you are familiar with? I am wondering about the tool able to give easily the level of a record in dBA or full range 20..20k
Thanks. I think I found also your preamp schematics.
At Digi-Key, I see three variants of the PUI AOM-5024, all with a sensitivity spec of (-24 +/- 3) dB with respect to 1 V/Pa and with a noise level of -80 dB A-weighted with respect to 1 Pa.
That corresponds to an equivalent acoustic noise level of about 14 dB(A), which is better than my 200 euro AKG C900 stage microphones (17.5 dB(A)). The maximum sound pressure handling is worse, though.
Anyway, -24 dB -80 dB = -104 dB, so the A-weighted microphone output noise is about -104 dB with respect to 1 V RMS.
0 dBu is the voltage that would produce a power of 1 mW into a 600 ohm load, no matter what the actual impedance may be. That is, about 0.7746 V RMS, between -2 and -3 dBV.
Hence, the noise level of the recorder has to be well below -101.5 dBu A-weighted in order not to mess up the noise floor.
Looking at the site you found:
H5: -121 dBu(A)
H4n: -107 dBu(A)
H2: -99 dBu(A)
It appears to be getting worse quickly with a decreasing number after the H, so chances are that an H1 would benefit from an extra preamplifier.
That corresponds to an equivalent acoustic noise level of about 14 dB(A), which is better than my 200 euro AKG C900 stage microphones (17.5 dB(A)). The maximum sound pressure handling is worse, though.
Anyway, -24 dB -80 dB = -104 dB, so the A-weighted microphone output noise is about -104 dB with respect to 1 V RMS.
0 dBu is the voltage that would produce a power of 1 mW into a 600 ohm load, no matter what the actual impedance may be. That is, about 0.7746 V RMS, between -2 and -3 dBV.
Hence, the noise level of the recorder has to be well below -101.5 dBu A-weighted in order not to mess up the noise floor.
Looking at the site you found:
H5: -121 dBu(A)
H4n: -107 dBu(A)
H2: -99 dBu(A)
It appears to be getting worse quickly with a decreasing number after the H, so chances are that an H1 would benefit from an extra preamplifier.