Great, many, many, many thanks! It will take about a week until the LME49721 arrives.
If possible, the lesser parts, the better... I could also go back from 94µF electrolytics to 5.6µF Wimas, if that helps. But in the datasheet I do not find data of the the input impedance of the LME49721 to calculate the corner frequency.
If possible, the lesser parts, the better... I could also go back from 94µF electrolytics to 5.6µF Wimas, if that helps. But in the datasheet I do not find data of the the input impedance of the LME49721 to calculate the corner frequency.
Last edited:
Funny - even Texas Instruments support does not know the input impedancde of the LME49721 .
They propose 100kΩ, so I assume I can put 5.6µF WIMA foils back in, instead of the 94µF Electrolytic caps:
https://e2e.ti.com/support/audio-gr...9866/lme49721-what-s-lme49721-input-impedance
They propose 100kΩ, so I assume I can put 5.6µF WIMA foils back in, instead of the 94µF Electrolytic caps:
https://e2e.ti.com/support/audio-gr...9866/lme49721-what-s-lme49721-input-impedance
Did you measure the response over the Phantom power resistors? Is the voltage stable over the entire range?
Many thanks for the suggestion, but no need to do so. All measurements were made with Phantom Power switched off, using sweeps @mic and @line level. This thread has changed to the topic of putting in a small preamp into the Xvive P1 so that the camera can record the MKH-416 signal @line level which shows better results.
Looking at the input noise current and bias current specifications, it must have some kind of FET input. The open-loop input impedance should then be far above 1 Gohm.
This doesn't help much, because the differential closed-loop input impedance of the circuit of post #58 is only 2 kohm, the sum of the values of the input resistors.
R1 and R7? Don't forget, I am a layman. Do we need them, when I go back to 5.6µF? Layman thinks: With just no resistors or 10ohms, the incoming signal is louder...
The impedance is determined by the capacitors then, except above about 3 kHz, where it is determined by the 10 ohm resistors. Hence, you get 40 dB more gain than you asked for at high audio frequencies, with a roll-off below 3 kHz (assuming that the microphone impedance is negligible and the 47 ohm resistors are shorted - with 47 ohm + 10 ohm, the gain rolls off below 500 Hz and is about 25 dB too high at high audio frequencies).
What do you propose? Keep the 94uF or work with 5.6uF? I assume, both need high resistance, to address leakage with electrolytics or to even out frequency reponse because capacitance can’t go higher with the space available? As far as I remember the MKH-416 has 10ohm output impedance.
Marcel,
many thanks! But there is no space
to put in four 47uF/63V caps seperately. The only way is to stack them tightly as pairs soldered together at the angled legs. Space in the Xvive is very restricted, as the battery consumes 4/3 of the space between the XLR connectors. No space to put R12/13 and diodes between the caps. The network to sink the current must be after them on a seperate, small board.
The opamp with feedback and output resistors must be on a second board. Space is also restricted for 10uF/10v caps.
many thanks! But there is no space
to put in four 47uF/63V caps seperately. The only way is to stack them tightly as pairs soldered together at the angled legs. Space in the Xvive is very restricted, as the battery consumes 4/3 of the space between the XLR connectors. No space to put R12/13 and diodes between the caps. The network to sink the current must be after them on a seperate, small board.
The opamp with feedback and output resistors must be on a second board. Space is also restricted for 10uF/10v caps.
Last edited:
I'm wondering if that 45 dB fixed voltage gain you asked for isn't much too high.
Your microphone has a sensitivity of 25 mV/Pa. Assuming -36 dB means -36 dBV RMS, that is 15.85 mV RMS, so 0.634 Pa, or 90.02 dB SPL relative to 20 uPa. If you ever want to film at a rock concert, or just about any venue with music, the level is bound to exceed 90.02 dB SPL in the loudest passages.
The gain you asked for is 45 dB. Assuming -36 dBV RMS is the correct level, you get 9 dBV RMS at the output. In case of a sine wave, that's 7.9716 V peak-to-peak, more than twice the highest level that fits in a 3.7 V supply. At 3.3 V, the microphone amplifier can at most handle 82 dB SPL when the gain is 45 dB.
Could you check what voltage level you need to drive your camera all the way to full scale via the line input when it is set to maximum gain? I know there is a -10 dBV in the specifications, but I assume that's a nominal level rather than the maximum level at maximum gain.
Could you then also record and post the camera's noise with shorted line input at maximum gain?
This is another option with fewer capacitors. I don't like it much for two reasons: all the gain is in the second stage, which has very little loop gain left at 20 kHz, and it amplifies its own offset almost 180 times, reducing headroom. It is probably good enough, though, and it gets better if the required gain should be less than the 45 dB assumed so far. R5, R7, R1, R2 and R4 may need to be updated when the required gain is not 45 dB.
Last edited:
Dear Marcel,
many thanks! Why is it built-in 2.2µF coupling capacitors from the phantom power supply,
and not 5.4µF? (In fact two stacked 2.2µF plus 1µF, all paralleled, as seen in post #29)
The Panasonic S5 Mic / Line Preamp is really noisy. The -12db setting gives the best S/N ratio,
as it equally rises with every +3dB step.
I can send you an audio file, just with the self - noise, on Saturday.
Like it is now, the transformer, acting also for going from balanced to single ended,
adds about 6dB. So it is about 50dB. With -12dB at the cam, relatively loud city noises,
cars passing by are below 0-dB peak. A Motorcycle is little over. So yes, unuseable for a concert,
But useable for Interviews or cinema verité like action, where one makes sure beforehand,
that the surroundings are not too loud.
When the LME49721 exceeds 82 dB SPL at 3.3V, what will I hear? Crackling or a rather soft distortion? What is your proposal for the gain? All the best, Salar
many thanks! Why is it built-in 2.2µF coupling capacitors from the phantom power supply,
and not 5.4µF? (In fact two stacked 2.2µF plus 1µF, all paralleled, as seen in post #29)
The Panasonic S5 Mic / Line Preamp is really noisy. The -12db setting gives the best S/N ratio,
as it equally rises with every +3dB step.
I can send you an audio file, just with the self - noise, on Saturday.
Like it is now, the transformer, acting also for going from balanced to single ended,
adds about 6dB. So it is about 50dB. With -12dB at the cam, relatively loud city noises,
cars passing by are below 0-dB peak. A Motorcycle is little over. So yes, unuseable for a concert,
But useable for Interviews or cinema verité like action, where one makes sure beforehand,
that the surroundings are not too loud.
When the LME49721 exceeds 82 dB SPL at 3.3V, what will I hear? Crackling or a rather soft distortion? What is your proposal for the gain? All the best, Salar
Last edited:
You would hear hard clipping, so a crackling-like sound.
My proposal for the gain is whatever gain is needed to amplify the microphone's noise to just above the camera's noise floor. Higher gains don't improve the signal-to-noise ratio when you are recording soft sounds much, while they do reduce the maximum signal handling. I need the measurements I asked for to put a number to it.
My proposal for the gain is whatever gain is needed to amplify the microphone's noise to just above the camera's noise floor. Higher gains don't improve the signal-to-noise ratio when you are recording soft sounds much, while they do reduce the maximum signal handling. I need the measurements I asked for to put a number to it.
Last edited:
OK, many thanks for the suggestions!
I´ll record them this night, input shorted, mic and line level.
My place is quiet, so I can also record the MKH-416 as well.
I´ll record them this night, input shorted, mic and line level.
My place is quiet, so I can also record the MKH-416 as well.
Dear Marcel,
here we go, the three files should be self-explanatory.
Recorded mic-level shorted, mic-level with MKH-416 attached, line level shorted.
In incrementing steps, first mute, then -12dB / -9dB / -6dB / -3dB / 0dB / +3dB / +6dB.
All 16bit / 48kHz
Noise level with mic-sensitivity rises dramatically, with line sensitivety,
the rise in noise is barely audible.
So as you proposed, amplification of your circuit does not seem to be
All the best, Salar
https://filetransfer.io/data-package/bDeNTJMh#link
here we go, the three files should be self-explanatory.
Recorded mic-level shorted, mic-level with MKH-416 attached, line level shorted.
In incrementing steps, first mute, then -12dB / -9dB / -6dB / -3dB / 0dB / +3dB / +6dB.
All 16bit / 48kHz
Noise level with mic-sensitivity rises dramatically, with line sensitivety,
the rise in noise is barely audible.
So as you proposed, amplification of your circuit does not seem to be
All the best, Salar
https://filetransfer.io/data-package/bDeNTJMh#link
As you will have to remove the electrolytics and place film capacitors again, I thought mounting just one capacitor per side would be more convenient for you than mounting all three per side. Assuming the circuit is in a metal enclosure that shields it from electric fields - is it indeed in a metal enclosure? - just one capacitor per side should be sufficient for the circuit of post #71, or a variant of it with less gain.
Thanks! I don't know when I have the time to analyse them, maybe tonight, maybe this weekend.
What signal voltage level do you need to drive the line input to the edge of clipping at -12 dB and at +6 dB?
unless the gain has to be reduced very drastically, then it might need a bigger capacitor after all. To be continued...
No clue. My old portable field mixer generates a variable, analog 1khz Testtone. I assume, i cannot measure the voltage like with a Voltmeter, set to DC. But I have a scope.
But this I can only do during the weekend.
The foil caps were already used, 5.4uF. Happy to put them back in. About the noise recordings: As the transformer in the Xvive, that converts the mkh-416 signal to single ended, has not been removed yet, it adds about 6dB to the MKH-416 recording. The other recordings were made with a shorted 3.5mm plug inserted inti the cameras mic input. The Cabinet of the Xvive is metal, probably die cast aluminium
Last edited: