Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

Impedance problem

Best Regards,
Im working on a Wheatstone bridge using a DC source of 18 volts (two 9 V batteries in series), three resistance of 50 MΩ (actually each 50 MΩ resistance is two 25 MΩ resistors in series) and a variable resistance from 1MΩ to 50MΩ. The output of this bridge is plugged to my USB 6009 DAQ using two Voltage follower configurations, so I can match the impedance of the DAQ with the output of the bridge. My problem comes at the moment when I measure the voltage in the output of the bridge. It should measure 0V when no change is in the sensor that I'm using and around 9V when it is at its peak value of response. The measurements readings are around 2 V, no matter the input applied to the sensor. Later I took out the sensor and the left part of the bridge (leaving only the DC Source and four resistors of 25 MΩ each), to start troubleshooting my circuit. What I encounter was that between my ground and the last resistor, there was 18 V, but the voltage between any of the four resistors alone was around 2 V. The Ohm's law was being violated, as the voltage divider principle was failing. What I thought at that moment was that my DC Source wasn't giving me the amount of current that I need for, but I calculated and what I need are 0.54 µA. Then I tested with my laptop's DC power supply, that has a 12 V output with a maximum of 2 A, but I got the same results. The voltage divider principle was failing. Between all the resistors there were 12V, but on each one alone, there were only around 1 V. I changed my multimeter batteries to new ones and got the same result. What is happening ? Does anyone knows what is the cause of this phenomenon? Can anyone suggest me what to do? Thanks for reading this thread, your suggestions would be very appreciated!.

Juan.

0 Kudos
Message 1 of 4
(3,003 Views)

Doing measurements with impedances in the 50MOhm range is not so easy.

You only need a 40nA current to get 2V at 50M 😮 . That is pretty well in the region of OPs bias current, what circuitry do you use?

While instrumental amplifier or OPs exists that have bias currents in the fA range .. actually building such amplifiers is still art.

Analyze the currents in your circuit and keep in mind that even a PCB is a resistor, and quite low if not cleaned 😉  every temperature difference will give you EMF and and...

For testing I like the dead bug approach with mostly air as isolator and intense cleaning / drying .

 

If you tell us the physics of your sensor, then there might be an other way to read it.

 

 

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


0 Kudos
Message 2 of 4
(2,981 Views)

Thanks for your advice Henrik!

 

Well Im using a water vapor sensor, wich I connect in parallel to a 50 Mohms resistor, due to its high impedance. What I found, was that my multimeter has an input impedance of 10 Mohms, so when Im  making voltage measurementes, that 10 Mohms input resistor is in parallel with the measured resistance of 25 Mohms. Due to the ratio between this two resistance, my entire circuit changes, because now I have 3 resistores in series, plus the one that Im measuring, that changes to an equivalent resistance of 7.05 Mohms. This equivalent resistance, that is the result of measuring in parallel for voltage, changes the entire current on the circuit, so I ended up reading a false signal, due to the impedance mismatch. What I did to check this theory was that I took the measured 25 Mohms resistance and connect both of its terminals to two OpAmp as voltage followers (JFET= high input impedance, low output impedance) to couple the impedance between it and the multimeter. After this, my reading were as espected. Around 4.5 V for the 18 V source, and 3.4 for the 12 V source.

0 Kudos
Message 3 of 4
(2,960 Views)

Water vapor sensor... you mean these PCB board with shiny gold traces??

I would be carafull with measuring voltages higher than 2V  even if your impedance is high .. it's not good for longterm stability... (if these sensores ever had them 😉 )

 

My approach would be using a C decoupled rectangle 5V DO to get a AC exitation (about 200Hz to 1kHz) , using an AC amplifier and do the signal processing in software (rectify, Lock In Amp, filter)

 

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


0 Kudos
Message 4 of 4
(2,953 Views)