Driving an analog pressure sensor (+5V) to an ADC (+3V)

Hello guys,

I need your comments and suggestions. I have the next stuff:

- A pressure sensor MPXV7002, output at full scale +5V@0.1mA, the sensor has a bandwidth of 1KHz.

- An ADC (AT91SAM7S256) with Vref +3V (CL = 1pF).

I've just thought the schematic in the picture attached.

R1 = 1.5K
R2 = 2.2K (5V to 3V).

C1 = C2 = 0.1uF

The OAs are the LM358.

I don't know if this is the best way to interface the ADC.

Thanks in advance.

This would work with a few remarks:

1. LM358 is not a rail-to-rail OP Amp so you will need to power it at minimum 6.5V to have a range of 0.5V to 4.5V in the first Op Amp.

2. 0.1uF cap at the second Op Amp output is a no-no. The second Op Amp is prone to oscillate. Insert a small resistor between the Op Amp output and cap to avoid instability. Another way is to move the filter cap C2 in parallel with R2 and recalculate its value.

3. Are you going to use the full range of 0.5V to 4.5V ? In other words, is your application designed for -2 to +2 kPa? If not, if for example you need just 0 to +2 kPa, then you will waste half of your ADC dynamic range. This, in effect, means lower ADC resolution.

It may turn out that a voltage divider with a 0.1uF cap at the A/D input is sufficient, and the OpAmp buffers are not needed. Cap alone may be able to provide sufficiently low output impedance. Easy to prototype and test. Your current relatively low 2.2k resistor also helps to keep the output impedance low.

K.I.S.S. (keep it simple)

AdrianN said:

This would work with a few remarks:

1. LM358 is not a rail-to-rail OP Amp so you will need to power it at minimum 6.5V to have a range of 0.5V to 4.5V in the first Op Amp.

2. 0.1uF cap at the second Op Amp output is a no-no. The second Op Amp is prone to oscillate. Insert a small resistor between the Op Amp output and cap to avoid instability. Another way is to move the filter cap C2 in parallel with R2 and recalculate its value.

3. Are you going to use the full range of 0.5V to 4.5V ? In other words, is your application designed for -2 to +2 kPa? If not, if for example you need just 0 to +2 kPa, then you will waste half of your ADC dynamic range. This, in effect, means lower ADC resolution.

Click to expand...

Thank you very much, it's been an usefull response.

1 & 3. Yeah I know it, ,maybe i haven't explained it correctly. 5V is the max value of the sensor, but I won't expect that value in my application (I didn't say that earlier, sorry). The application is a pitot to measure the speed of a Radio Control boat (max speed about 10Km/h), so I won't reach 5V at the output sensor. With max 4V at the output, you're right, I have to check out the datasheet of the OP, anyway, any other suggestion selecting another OP?

This sensor will measure Vy speed (left /right), so -2kPa make sense xD.

2. I will test both options, thanks again.

kender said:

It may turn out that a voltage divider with a 0.1uF cap at the A/D input is sufficient, and the OpAmp buffers are not needed. Cap alone may be able to provide sufficiently low output impedance. Easy to prototype and test. Your current relatively low 2.2k resistor also helps to keep the output impedance low.

K.I.S.S. (keep it simple)

Click to expand...

I will test that too, it was my first option but i started to mess all with the impedances bla bla xD

I'll keep you tuned on Monday with the results.

The simple divisor voltage works xD

With R2 = 2.2K and Cout = .1uF , the BW is about 700Hz (more than enough). The Cout provides enough low impedance for driving the ADC.


Thank you very much for your comments guys.