Comprehending wiring data sheet for ME3 C2H4 gas sensor

Thread Starter

Mohammed Omer_1549158839

Joined Feb 2, 2019
13

I am humbly requesting an English translation of this image. The only thing I figured out is that the box on the top left is the ME3 sensor I am trying to connect to an Arduino. It has three pins, as indicated by the reference sheet. But aside from that, nothing on here tells me anything about how to connect this to an Arduino Uno. All I need is for you to identify the pins of the sensor. Anything extra would be greatly appreciated but that's all. Thanks greatly in advance. The link of the pdf image of the sensor I am using is: pdf image link
 

Papabravo

Joined Feb 24, 2006
21,159
You might be thinking that the circuitry is inside the sensor. I don't think that is the case. The AD 8572 is a operational amplifier that requires ±3 VDC power supplies. So we have an external battery which feeds a linear regulator with a +3VDC output. That feeds the MAX660 inverter which produces the -3 VDC output. I have no way to know what the operational amplifiers are doing. My guess is that "C" is an input to the sensor, while R & W are outputs. Best suggestion is to build the circuit in the diagram and look at the signal referred to as SIGN1 to see what it does under various conditions. Both of the operational amplifiers have capacitors in the feedback loop so they are performing an "integration" function.
 

Thread Starter

Mohammed Omer_1549158839

Joined Feb 2, 2019
13
I think C stands for counter electrode. R and W stand for reference electrode and working/auxiliary electrode, respectively. I just want to connect this to my Arduino Uno, so the schematic is unnecessary for now. I believe SIGN1 can be a power source too or an external source. If you can please give me instructions on what transistor I would use to connect this to my Arduino that would be great. Many thanks in advance.
 

wayneh

Joined Sep 9, 2010
17,496
I just want to connect this to my Arduino Uno, so the schematic is unnecessary for now.
Uh, no. You're going to have to do some kind of signal conditioning of the sensor's current output, probably to convert it to a voltage, in order for your Arduino to make sense of it. I do not understand that schematic but I imagine it's delivering a voltage output that is proportional to the current from the sensor, which in turn is a measure of ethylene concentration.
 

Papabravo

Joined Feb 24, 2006
21,159
As I said in my previous post, that circuit is essential to connecting the sensor to anything. Either get more information from the manufacturer or find a sensor from a manufacturer that will support you. That is perhaps the crappiest datasheet I've ever seen.
 

Raymond Genovese

Joined Mar 5, 2016
1,653

wayneh

Joined Sep 9, 2010
17,496
What passes as a datasheet is sometimes shocking https://p.globalsources.com/IMAGES/PDT/SPEC/026/K1059654026.pdf

That circuit is described pretty well in this thread https://forum.arduino.cc/index.php?topic=579782.0

Here is my rant though - it is foolish in most cases to use a sensor without any level of calibration. True, there are situations where presence or absence is the goal, but honestly, do you want to mess with a sensor without the ability to calibrate?
Well I think the calibration is easily done in software as long as you get a decent signal out of the thing.
 
Here is my rant though - it is foolish in most cases to use a sensor without any level of calibration. True, there are situations where presence or absence is the goal, but honestly, do you want to mess with a sensor without the ability to calibrate?
I installed and maintained Hydrogen sensors. calibration was not done at the sensor. It was done at the alarm unit.
pre-mixed calibration gas was introduced to a tube that flooded the sensor. Generally, you had to do a two-point calibration. We checked the results yearly.

With any sensor, you generally need to know the gas your detecting. When you can't calibrate, it's time to replace the sensor.

Our other detectors were tape based. The machine sampled the atmosphere through a chemical tape.The color change was measured to a concentration, but we had no idea what the gas was that caused the alarm.

So, the sensor signal conditioning is nominal values. Fine tuning occurs at the alarm panel. I think there was 4-wire interface,

We had ceiling and duct sensors. Same deal, a port (separate plumbing) floods the sensor with calibration gas mixes. There was a 24 h warm-up period.
 

wayneh

Joined Sep 9, 2010
17,496
I did a little more searching and found that this supplier will ship the sensor and circuit in an integrated unit, and supply a calibration record as part of that. Seems like the way to go.

I saw references to the datasheet saying the output is 0.04µA per ppm. If it's there, I didn't see it! But anyway the circuit converts current to voltage and integrates it, delivering 0-3.3V for 0-100ppm.
 
Well I think the calibration is easily done in software as long as you get a decent signal out of the thing.
I disagree. In the absence of a standard, no software is going to calibrate a sensor like the on the OP has. If you have to, you can rely on calibration data issued by the manufacturer. Of course, again, no such calibration data is available to the OP for that sensor.

Edited to add. Yes, I saw a note (in the thread I linked to) about getting calibration data from the manufacturer - as you noted in a subsequent post. I would, however, like to see a response in that regard - i.e., with that sensor. IOW saying it (the company) and doing it for a single unit is different.

Not always, but in the case of cheap and poorly documented sensors, frequently. I recently bought a CO sensor and the calibration data is coded on a label on the sensor. It cost 10X what a cheap CO sensor costs.
 

wayneh

Joined Sep 9, 2010
17,496
In the absence of a standard, no software is going to calibrate a sensor like the on the OP has.
In the absence of standards, no sensor from any manufacturer can be fully calibrated. You have to rely on the claim of 40nA per ppm and hope for the best after you adjust slope and intercept in software. Supplied calibration information might be at the lot level, or at the level of individual sensors but as you well know, you pay for that. Probably cheaper than trying to do it yourself, though.
 

Thread Starter

Mohammed Omer_1549158839

Joined Feb 2, 2019
13
The manufacturer told us that the sensor is calibrated from before. However, I don't trust them and since it is a Chinese New Year, they are not replying to my requests. I don't think the calibration is a problem since we have done that. I connected this to Serial monitor and it shows values that stay constant around 250(ppm?) for a while then jumps up to 550. Its a strange sensor, and right now I can't do anything.
 
In the absence of standards, no sensor from any manufacturer can be fully calibrated. You have to rely on the claim of 40nA per ppm and hope for the best after you adjust slope and intercept in software. Supplied calibration information might be at the lot level, or at the level of individual sensors but as you well know, you pay for that. Probably cheaper than trying to do it yourself, though.
Oh ok, so, when you said:
Well I think the calibration is easily done in software as long as you get a decent signal out of the thing.
It was, from my point of view, plainly inaccurate. Now, it turns out that we both understand what a standard is - good deal.
 
The manufacturer told us that the sensor is calibrated from before. However, I don't trust them and since it is a Chinese New Year, they are not replying to my requests. I don't think the calibration is a problem since we have done that. I connected this to Serial monitor and it shows values that stay constant around 250(ppm?) for a while then jumps up to 550. Its a strange sensor, and right now I can't do anything.
When you say, "I don't think the calibration is a problem since we have done that" - can you please explain?

The sensor data sheet states that it has a detectable (useful) responseover 0-100 ppm range and a maximum of 200 ppm. How are you getting values from the sensor and at what concentration of C2H4???

If you have only one reference standard for C2H4 - that is C2H4 in a known ppm concentration and you put the sensor in that environment, you can check the accuracy of the sensor at that concentration, but you can't calibrate the sensor (which would be required to quantity other ppm concentrations) - do you agree with that?

You need, at least, two reference standards. You simply can't assume that the error (e.g., a percentage off) is the same at all concentrations.
 

Thread Starter

Mohammed Omer_1549158839

Joined Feb 2, 2019
13
When I said, "I don't think the calibration is a problem since we have done that"
I'm sorry if I confused people but I was trying to say that the manufacturer attached calibration info and claimed they already did. However, to be sure we calibrated the sensor on the Arduino code.
Let me attach a readout of what the sensor is giving us.

289
289
288
289
289
290
293
293
292
293
293
292
292
This is what I see from the Serial monitor.
I agree with you on the reference standard, I have tried exposing the sensor to pure hydrogen since that was a value and the readout dropped to 90.
Thank you for the help very much.
Also, I think the values are getting multiplied somewhere so that is why they are providing such a high reading
 
When I said, "I don't think the calibration is a problem since we have done that"
I'm sorry if I confused people but I was trying to say that the manufacturer attached calibration info and claimed they already did. However, to be sure we calibrated the sensor on the Arduino code.
Let me attach a readout of what the sensor is giving us.

289
289
288
289
289
290
293
293
292
293
293
292
292
This is what I see from the Serial monitor.
I agree with you on the reference standard, I have tried exposing the sensor to pure hydrogen since that was a value and the readout dropped to 90.
Thank you for the help very much.
You're very welcome. The data that you have posted makes no sense to me - sorry.
 

wayneh

Joined Sep 9, 2010
17,496
I think it might have been around 160 ppm.
Well that made me chuckle. My first boss was a tyrant and as bad as that sounds, it was often made worse by the fact that he was often right in his tyranny. The first time I turned in a report containing the words "about" or "around", it came back to me with big red slashes. Does "about" mean ±5%, or ±50%, or ±100%?, he demanded to know. The lesson was soon learned: Words like "about" and "around" have no place in technical writing. You may know what they imply today as you write the word but no one else knows what they mean and in a few years when you read your own writing, you won't remember either.

So far we still have no reason to suspect the sensor is doing anything other than exactly what it's supposed to.
 
Top