Maker Pro
Maker Pro

Current source in front of JFET Input Amplifier

W

wolti_At

Jan 1, 1970
0
Hello,

I am currently working on a prototype for the second generation of an
electrode amplifier. It is used to amplify the signals of pH and ISE
electrodes where the former ones have a very high output resistance in
the 100-1000megaohm range. The input amplifier I am using is based on
two matched JFET fets in a source follower configuration. The second
JFET is used to get a zero DC offset. That is the first input is
connected to the gate of Q1. The second JFET has its Gate-Source tied
together and is connected to the source of Q1 which is also the
output. The amplifiers I am using are from linear technologies (LS830,
http://www.linearsystems.com/datasheets/LS830-3.pdf). This output
signal is then band-limited, amplified and digitized using an ADC.

The second task is now that I would like to measure the resistance of
the electrodes. This implies that I have to connect something to the
input signal which I can then measure. I though a current source would
maybe be a good idea. The problems here is that the output current
must be very low (1nA) to get approx 1V for 1000Megaohm and of course
it should not affect the input at all;) I tried to come up with a
solution based on two sense resistors with a factor of 1:10000 where
the first resistor was programmed by the MCU with a constant current.
An opamp then compares the two sense resistors and drives an P-Channel
MOSFET which is connected to the input lines.

A small simulation showed me that there are problems with the leakage
current of the MOSFET and that it reduces the dynamic performance due
to its drain-source capacitance. Does anybody know a good device for
this or has maybe a better idea on how to measure the resistance of an
electrode?

Thanks,
Christian
 
W

wolti_At

Jan 1, 1970
0
Hello,

I am currently working on a prototype for the second generation of an
electrode amplifier. It is used to amplify the signals of pH and ISE
electrodes where the former ones have a very high output resistance in
the 100-1000megaohm range. The input amplifier I am using is based on
two matched JFET fets in a source follower configuration. The second
JFET is used to get a zero DC offset. That is the first input is
connected to the gate of Q1. The second JFET has its Gate-Source tied
together and is connected to the source of Q1 which is also the
output. The amplifiers I am using are from linear technologies (LS830,http://www.linearsystems.com/datasheets/LS830-3.pdf). This output
signal is then band-limited, amplified and digitized using an ADC.

The second task is now that I would like to measure the resistance of
the electrodes. This implies that I have to connect something to the
input signal which I can then measure. I though a current source would
maybe be a good idea. The problems here is that the output current
must be very low (1nA) to get approx 1V for 1000Megaohm and of course
it should not affect the input at all;) I tried to come up with a
solution based on two sense resistors with a factor of 1:10000 where
the first resistor was programmed by the MCU with a constant current.
An opamp then compares the two sense resistors and drives an P-Channel
MOSFET which is connected to the input lines.

A small simulation showed me that there are problems with the leakage
current of the MOSFET and that it reduces the dynamic performance due
to its drain-source capacitance. Does anybody know a good device for
this or has maybe a better idea on how to measure the resistance of an
electrode?

Thanks,
Christian

Hello,

If somebody followed the thread - I found myself quite a good solution
where I am capacitively coupling in an AC voltage which is then
rectified and measured with my standard input amplifier. Using some
math and two different frequencies I am then able to calculate the
load resistance.

Kind regards,
Christian
 
Top