Amplifying Light Sensor Output

Thread Starter

blimey99

Joined Mar 31, 2017
4
Hi all,

I have an Apogee SQ-500 quantum sensor that I hook up to my multimeter in order to get a reading from (it measures photosynthetically active light in units of micromoles per meter squared per second).

My problem is that my multimeter only reads down to a millivolt and one millivolt is equal to 100 micromoles per meter squared per second with this sensor. As a result, I can only get a reading of 100umols (1mV) 200umols (2mV), 300umols (3mV) etc., but I would like to get more detailed readings (e.g - 100, 110, 120). To do this, I'll need to amplify a 10 microvolt signal up to a 1 milivolt signal, so each millivolt becomes equivalent to 1 micromole per meter squared per second.

From what I understand, I'll need some sort of amplifier that can gain this signal up x100, but I don't know enough about this to find the right component. Can you please help a noob out?
 

#12

Joined Nov 30, 2010
18,224
This is a really bleeping difficult differential mode instrumentation amplifier you want. The input current needs to be about 1 pico-amp to input 10 uv through 1 meg at 10% accuracy, and I'm sure you aren't looking for 10% accuracy.

You have two choices that I can think of. Buy the recommended equipment in the PDF or wait until somebody smarter than I am comes along to figure a better way. (Yes, there are a few analog designers better than me on this site.) Maybe this chip will help. Maybe I'm acting like a dolt right now, but I don't see how to do this without designing a circuit you probably couldn't build.
 

Attachments

Dodgydave

Joined Jun 22, 2012
11,285
Came across this circuit on another Apogee site...it has a gain of x100, U2 is used to make a virtual ground for U1, using a 3 to5V supply.


sch.png
 
Last edited:

#12

Joined Nov 30, 2010
18,224
I can see a few reasons why I did not think I could make that work.

1) Common mode rejection of noise pickup in a twisted pair input. There isn't any.
2) Voltage limit on the chip of 5.5 volts. In full sunlight, this amplifier will need to output 25 volts DC. The ultimate sensor output limit is 40 volts.
3) Initial offset voltage guarantee of 10 uV at room temperature to 25 uV across temperature changes. After a gain of 100, that is 1 mV to 2.5 mV error on the low end of the measurement range and the Thread Starter already has that much resolution.

This circuits' gain is 100:)...and you can probably filter out most of the noise by slowing it down to about 10 Hz:rolleyes:...but the amplifiers' full range is less than 1/4th of the daylight the sensor was designed to measure and the +/- 1 mV problem is still there.:(
 

Thread Starter

blimey99

Joined Mar 31, 2017
4
Thank you both for your input! I had initially seen this little guy and figured it's got to be doable somehow. For a novice like me though, I may be better off just biting the bullet and shelling out for that piece - it's just so damn pricey here in Canada; by the time it's on my door it'll be over $60!
 

#12

Joined Nov 30, 2010
18,224
by the time it's on my door it'll be over $60!
I did one yesterday where the retail price was $150, and all it did was flip a relay and turn on a light.
Compared to him, you're getting a bargain.:D

I do have some curiosity as to why you would want a high limit of 1/4th of sunlight and a low limit so low that I can't calculate it without looking up the math. Looking for ghosts in a cave?
 

Thread Starter

blimey99

Joined Mar 31, 2017
4
I did one yesterday where the retail price was $150, and all it did was flip a relay and turn on a light.
Compared to him, you're getting a bargain.:D

I do have some curiosity as to why you would want a high limit of 1/4th of sunlight and a low limit so low that I can't calculate it without looking up the math. Looking for ghosts in a cave?
Can you please explain how you derive those limits? My ultimate goal is to be able to accurately compare the output of different COB LEDs. In my current state, I have no accuracy - if I compare 2 chips and they both read 8mV (800umol/m^2/s), I don't get anything out of it. If I were able to compare 2 chips and get readings of 890umol/m^2/s on one and 860umol/m^2/s on the other, I'd be able to draw some meaningful conclusions.

Apogee does sell a version of this sensor that comes with a reader that provides the accuracy I'm looking for - I just didn't want to spend the extra money for it since I figured I'd be able to find a way to get the same accuracy with just the sensor and a little elbow grease.
 

#12

Joined Nov 30, 2010
18,224
Can you please explain how you derive those limits?
Let's see. According to the datasheet, the reading for sunlight is 25 mV. If you amplify that by 100, you would have to get the amplifier to output... 2.5 volts?
Yikes. It looks like I slipped a digit. I previously calculated 25V for sunlight and 40V for maximum output.
Oh well. That eliminates one problem with the schematic in post #2.:D
 

theatrus

Joined Apr 4, 2017
1
Schematic in post #2 and the product link are ~nearly the same schematic, if that helps you save any cash (I'm the guilty one for the original schematic on ReefCentral and the board). Also, you can bring R4 to about 1k (from 100k) to get the full range out of the sensor from a 3V supply like the coin cell - blueAcro board does this. The op-amp doesn't require much of an offset. Add a low-pass on the input and you're good to go.

The original application was measuring PAR in an aquarium setting, where the useful ranges are in the 100-800 range. The design of the sensor head even introduces a 10% error when immersed in water (gain should be 110 for this application ;)), which was only discovered after the initial release by the Apogee; a few % out from the amplifier over temp is perfectly fine in those applications, though you can dress it up at bit to improve the performance (i.e. common mode noise) if you need more more analytical.

Hope that helps!
 
Last edited:

Thread Starter

blimey99

Joined Mar 31, 2017
4
Schematic in post #2 and the product link are ~nearly the same schematic, if that helps you save any cash (I'm the guilty one for the original schematic on ReefCentral and the board). Also, you can bring R4 to about 1k (from 100k) to get the full range out of the sensor from a 3V supply like the coin cell - blueAcro board does this. The op-amp doesn't require much of an offset. Add a low-pass on the input and you're good to go.

The original application was measuring PAR in an aquarium setting, where the useful ranges are in the 100-800 range. The design of the sensor head even introduces a 10% error when immersed in water (gain should be 110 for this application ;)), which was only discovered after the initial release by the Apogee; a few % out from the amplifier over temp is perfectly fine in those applications, though you can dress it up at bit to improve the performance (i.e. common mode noise) if you need more more analytical.

Hope that helps!
Thank you! I was thinking it looked very similar and hoped it was the same.
 
Top