0
\$\begingroup\$

Piezoelectric mics and accelerometers are called high impedance aka high Z sensors. And I know that the high impedance sensor output needs to be converted into a low impedance signal first. I think this impedance transformation is needed for carrying the signal with minimum power loss.

But my question is more fundamental here. Let’s take a piezoelectric accelerometer/mic/sensor and model it. As far as I know it is modeled as a voltage source with a series capacitor. Something like:

enter image description here

Now if my model is correct for a piezoelectric sensor, why are these called high impedance sensors? Obviously Xc is frequency dependent which means it is hard to say whether the impedance is high or low here. Where am I wrong?

\$\endgroup\$
14
  • 1
    \$\begingroup\$ Look at the typical capacitance and the typical frequencies. \$\endgroup\$
    – Chris Stratton
    May 9 '19 at 18:49
  • 1
    \$\begingroup\$ What is a crystal dielectric? Answer: an insulator with piezo properties \$\endgroup\$
    – Tony Stewart EE75
    May 9 '19 at 18:57
  • 1
    \$\begingroup\$ the range of frequencies it's sensitive to is not infinite \$\endgroup\$
    – dandavis
    May 9 '19 at 19:06
  • 1
    \$\begingroup\$ But I think if this is buffered since the buffer has almost infinite input impedance I guess then the signal will not attenuate at low frequencies. \$\endgroup\$
    – floppy380
    May 9 '19 at 19:49
  • 1
    \$\begingroup\$ @HelpMee No, because if the stress across the crystal develops too slowly, the voltage generated will be too small and just disappear into the thermal noise of the crystal and buffer. \$\endgroup\$
    – DKNguyen
    May 9 '19 at 20:28
1
\$\begingroup\$

High impedance generally means that as a source, the source impedance is high. And by high, it can also mean that the signal from the sensor needs to be amplified and the amplifier needs to have a high input impedance. If the load impedance on the sensor is low, it will pull the sensor voltage lower and create error in the sensor measurement.

\$\endgroup\$
8
  • \$\begingroup\$ In my case the source impedance is Xc correct? \$\endgroup\$
    – floppy380
    May 9 '19 at 19:24
  • \$\begingroup\$ Is ampllifer’s huge input impedance providing frequency independence ?. \$\endgroup\$
    – floppy380
    May 9 '19 at 19:47
  • \$\begingroup\$ If you compensate it correctly to the sensor's impedance \$\endgroup\$
    – Voltage Spike ♦
    May 9 '19 at 19:53
  • \$\begingroup\$ What do u mean by compensation? \$\endgroup\$
    – floppy380
    May 9 '19 at 19:54
  • \$\begingroup\$ Frequency compensation in the feedback loop, like integrators, filters differentiators \$\endgroup\$
    – Voltage Spike ♦
    May 9 '19 at 19:54
0
\$\begingroup\$

The comments try discreetly quide your thinking to the right direction without putting the truth straightly against your face. The brutal technical fact is that whatever you connect to your sensor output, it needs some current. Generally you can see there's a resistor as the load. Together with your capacitor they are a high pass filter. The slower are the voltage changes the more they attenuate in the hp filter. If you know the capacitance and the frequency range of the acceleration, you can calculate the needed minimum load resistance for a given error due the filtering.

Learn RC high pass filter.

\$\endgroup\$
5
  • \$\begingroup\$ If C is very high and the frequency is very high then Xc becomes almost zero so the sensor becomes a low impedance. Since there is such dependency why are these sensors called high Z. There must be some typical values for C and f for such sensors then. What are they roughly? \$\endgroup\$
    – floppy380
    May 9 '19 at 19:33
  • \$\begingroup\$ Your sensor doesn't work properly at frequencies beyond the range that the manufacturer has specified in the datasheet. The range can be for ex. from 1Hz to 10kHz. If the piezo crystal has capacitance = 1000pF (=imagination) the average reactance is about 30 kOhms. That's not especially little, At 1Hz it's about 160 MOhm. You should consider Hi-Z as a description for the needed load resistance. Many practical sensors surely have the needed Hi-Z input preamp integrated and they output standard instrumentation signals. \$\endgroup\$
    – user287001
    May 9 '19 at 20:18
  • \$\begingroup\$ Something like a small diaphragm capacitor mic element (Without its fet buffer) will be maybe 30pF, and you want to be reasonably flat to below 50Hz.... \$\endgroup\$
    – Dan Mills
    May 9 '19 at 22:48
  • \$\begingroup\$ @DanMills Is the model for a piezoelectric sensor in my question wrong? (Not related to your comment) \$\endgroup\$
    – floppy380
    May 9 '19 at 23:59
  • \$\begingroup\$ @HelpMee "If C is very high and the frequency is very high" -- In practice the two are not both true. The capacitance of a piezoelectric sensor is very low, and the frequency at which it is useful is also low, in the grand scheme of things. You won't be using them at GHz frequencies, and their capacitance isn't on the order of mF. \$\endgroup\$
    – nanofarad
    May 10 '19 at 0:09
0
\$\begingroup\$

I think the term "high" comes from radio broadcasting, and the commercial audio realm, pre-1970s.

In the Morse-code radio era before vacuum tubes, headphones were either of telephone-receiver type (magnetic speaker,) or of piezo crystal type. Across the audio spectrum, the coils-phones have low impedance, well below 1K, while the crystal headphones have far higher impedance.

Years later, up until the 1970s, normal microphones have a few hundred ohms output impedance (since microphones are dynamic mics; coil/magnets-based, before the arrival of cheap electret microphones.) Loudspeakers are "low impedance" as well, Zinp of 8ohms, or perhaps 4ohms, or 72ohms.

After the 1970s, along come piezo devices, with impedance far, far higher than loudspeakers, higher than microphones, higher than anything except those 1920s-era crystal headphones. From that viewpoint, piezos are "high" impedance, with "high" meaning "far higher than loudspeakers and coil-based microphones," measured at audio frequencies.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.