amplifier with very large input resistance?

[xmat]

Hello, to all.

I have a transducer with a very large output resistance, more than my multimeter can measure ( >>1Mohm).

I want to connect it into an absolute value circuit, presumably it (also) needs an amplification..

My question is:

Is there an easy/specific way to make a low noise amplifier of high input impedance to "link" the transducer to the absolute value circuit?
Can anybody suggest a schematic?
(I am looking for a cheaper solution to the one an "instrumentation amplifier" can provide..)

And what are the main problems when the amplifier does not have an appropriate input resistance? (i.e low signal...anything else?!?)

PS:The transducer is of a condenser material, similarly-working as a condenser electret capsule;
yet it does not have a factory-build FET amplifier to it, so it cannot be easily linked to other circuits.

Hope somebody can help,

regards,

xmat.

 

[williB]

well we could guess , but could you give the part number or better yet the spec sheet for the part in question..

 

[eblc1388]

I want to connect it into an absolute value circuit, presumably it (also) needs an amplification..

An opamp in voltage follower mode gives very high input impedance and you can possibly arrange it to give some gain too.

 

[chemelec]

Just Because it Measures a Very High "DC Resistance" with a Meter, Doesn't Always mean it is a High Impedance.
DC Resistance and Impedance are TWO DIFFERENT THINGS.
Although it could be High Impedance.

In Most cases, an Input Inpedance of about 1 Megohm is acceptable for High impedance devices. Even for Crystal or Pizo Mics.

Take care.........Gary

 

[xmat]

thanks for the tips

Hello, guys.

In reply to your answers,

I am trying to connect a "b-band transducer"

(see link www.b-band.com )

into an analog to digital converter.

Connecting that into a normal amplifier gives a very noisy and low level (microvolts) signal.

So I am assuming that the problem is impendance.

This transducer is made of a special material, it works like a "capacitor"...It is similar to a consender mic capsule (without the build in FET amplifier the capsules have..).

Connecting it to a FET amp taken out from a capsule, I noticed some problems in the signal.
The signal's sensitivity was becoming high or low as time passed by,
plus the transducer behaved as a "short circuit" randomly; like a bad contact!But the transducer is fine..

That's all, hope I made myself clearer now, and somebody can suggest a more specific solution.

Thanks in advance,
xmat

 

[Optikon]

Hello, to all.

I have a transducer with a very large output resistance, more than my multimeter can measure ( >>1Mohm).

I want to connect it into an absolute value circuit, presumably it (also) needs an amplification..

My question is:

Is there an easy/specific way to make a low noise amplifier of high input impedance to "link" the transducer to the absolute value circuit?
Can anybody suggest a schematic?
(I am looking for a cheaper solution to the one an "instrumentation amplifier" can provide..)

And what are the main problems when the amplifier does not have an appropriate input resistance? (i.e low signal...anything else?!?)

PS:The transducer is of a condenser material, similarly-working as a condenser electret capsule;
yet it does not have a factory-build FET amplifier to it, so it cannot be easily linked to other circuits.

Hope somebody can help,

regards,

xmat.

Very high input impedance amplifiers have FET type inputs. You can even find some that have bias currents below 1pA. To minimize offset errors, you need something like this.

I didnt see you specify how low of noise you need. One inherent problem with the noise of your system is it will depend on your source impedance which, is high (> 1Mohm). So your noise floor is at best, dominated by the thermal noise of this impedance. But it will be larger due to the amplifiers noise contribution as well.

You might try and focus on how to reduce your source impedance first to make the amplifiers job easier noise-wise. Is there a way you can parallel transducers? If they can somehow be paralleled with multiple units, you can easily reduce that impedance.

Also, while your down in micro-volt land, don't forget to consider thermo-couple junction effects in your wiring scheme. Errors like that can easily blow your transducer signal out of the water..

 

[Roff]

All op amps need a path for input bias (or leakage) current. The random behavior you are experiencing is consistent with a capacitive or piezoelectric transducer with no DC path to GND. You need a resistor from the input to GND (or to a bias voltage, if you have only one supply). You may also need to AC couple the transducer to the input.
If you use a JFET input amplifier, the resistor can be 1 - 10 Megohms unless you expect the amplifier to get hot.

 

[xmat]

Thanks!!

All op amps need a path for input bias (or leakage) current. The random behavior you are experiencing is consistent with a capacitive or piezoelectric transducer with no DC path to GND. You need a resistor from the input to GND (or to a bias voltage, if you have only one supply). You may also need to AC couple the transducer to the input.
If you use a JFET input amplifier, the resistor can be 1 - 10 Megohms unless you expect the amplifier to get hot.

Thanks!! That looks exactly what I needed!!

Really appreciated!Keep up the good work here!

Seeya!

 

[Roff]

Good deal. I hope it works for you.