Power an analog accelerometer 3V, from 48V Phantom Power

[FvM]

If i connect the accelerometer output to XLR pin 2, then i measure 7-8V at the output.

Apparently the real circuit doesn't conform with the schematic. In the latter, there's an 1 uF coupling capacitor between accelerometer output X and XLR pin, so I don't see how the voltage should be raised by connecting the accelerometer. The schematic in post #14 has Z-diodes for protection, have they been omitted now?

 

[ELEFANTOMET]

The Zener diodes havent been removed, although they are from the circuit. The 1uF capacitor is also still there. The circuit is build exactly like on the schematic, except for the missing zener, thats in the circuit but not 2nd schematic.

To block the 48Vdc phantom power do i need another value capacitor or should 1uF be fine?

 

[FvM]

1 uF isn't bad, of course the capacitor must have sufficent voltage rating (and correct polarity if using polarized capacitor). I won't use the circuit without protection Z-diodes to avoid possible damage in case of inadvertent hot plugging.

 

[ELEFANTOMET]

So this is how the final schematic looks like. I never figured out what was going on with the decoupling capacitor, measuring 48V on the left side, and 7V on the right side. But the problem got solved some how.

Ive put in a voltage follower for impedance matching, because of the 32kOhm at the outputs of the accelerometer. And now it works great.
Thanks for all your help FvM! Much appreciated




Last question though

Any ideas on how to test several accelerometers, if they perform the same way?

Ive been think of placing them on a board in a circle. And smash with a hammer, right in the center of the accelerometer circle. And then look at the frequency response from the impulse.

A nice thing could also be, to place them on a loudspeak membrane, and then do a frequency sweep. But there is really no good way to mount them on the loudspeaker membrane(glue them to the membrane is no option).

So does anyone have any good ideas how to do a frequency vibration sweep?