Unit of measure gp2y1010au0f

[pasgabriele]

Hi everyone,
i'm new in the forum and this is my first topic!
I control my gp2y1010au0f with ESP Easy correctly. I have configured the sensor in ESP Easy Device tab.
The return values of sensor is about 2400. But I don't understand what is the unit of measure used by ESP Easy. It isn't ug/m^3! Can you tell me what is the unit of measure?
Thanks you so much!

[pasgabriele]

I have analyzed the source code of ESP Easy for dust sensor (_P018_Dust.ino). I think that the output value is the sum of 25 lecture of raw value of analog pin:

Code: Select all

for (x = 0; x < 25; x++)
{
digitalWrite(Plugin_GP2Y10_LED_Pin, LOW);
delayMicroseconds(280);
value = value + analogRead(A0);
delayMicroseconds(40);
digitalWrite(Plugin_GP2Y10_LED_Pin, HIGH);
delayMicroseconds(9680);
}

So I have added to code the following lines to convert analog raw value to um/m^3 (standard measure):

Code: Select all

value = value/25;
//0 - 5.0V mapped to 0 - 1023 integer values (if you use 3.3V, map 0 - 3.3V to 0 - 1023)
value = value*(5.0/1024);
//linear eqaution taken from http://www.howmuchsnow.com/arduino/airquality/ to convert read voltage to ug/m^3
//Chris Nafis (c) 2012
value = (0.17*value-0.1)*1000;

Do you think it's a correct use?

[Shardan]

I'm not sure about that.

As far a s i know the ESP's analog input can take a maximum of 1V -> 1024 steps.
So i don't see how the map calculation for 5 (or 3.3) volts to 1024 steps should work.

Look at
http://www.instructables.com/id/ESP8266 ... g-Sensors/

[pasgabriele]

Thanks for reply Shardan,
now I'm using a Wemos D1 and on the Wemos website (https://www.wemos.cc/product/d1.html) the manufacturer says that A0 pin has 3.3V max input voltage. Then can I use my map calculation?

[Shardan]

Well, you can put 3.3V on the ADC-Input, it won't blow the chip.
If it can resolve 3.3V or if everything above 1V is maxed out to 1023 is on another note.

Maybe just try the simple circuit mentioned in the link i gave to check out what happens.

[papperone]

Well, you can put 3.3V on the ADC-Input, it won't blow the chip.
If it can resolve 3.3V or if everything above 1V is maxed out to 1023 is on another note.

Maybe just try the simple circuit mentioned in the link i gave to check out what happens.

What he meant is that with Wemos D1 (or compatible) there's already a volyage divider on the small board which makes the AD0/ADC input capable to accept 0-3.3V generating 0-1023 value (10 bit)...

[Shardan]

I just checked with my WeMos sample and i agree: the WeMos D1 has a divider buit-in.
3.3V works with that input.

I mostly use ESP-chips directly i wasn't aware of that.

With an internal voltage divider that works.
But as a note to all, for other boards like the nodeMCU or when using an ESP-Chip an external voltage divider (4.7K/2.2K for 3.3v, 8.2K/2.2K for 5V) is necessary for that.

[Shardan]

Hello again,

i've changed my dust sensor adding a voltage divider as described.

I didn't change the source code but used a value formula in the device definition instead.

For using 5V with the sensor that gives a formula: (%Value% *170)/5120 -100).
Using 3.3 V with the sensor gives (%Value% *170)/7757.57 -100).

Calculating the value for 3.3V by hand along the given source code change or the formula both give same negative values even i'm smoker so the value should be relatively high in my room.

The direct value reading gives 4486.
So the source code gives:

4486 / 25 = 179.44

179.44 * (3.3V / 1024) = 0.578273V

(0.17 * 0.578273 - 0.1) * 1000 = -901 !

I doubt if the factor 0.17 is correct. Following the datasheet the GP2Y10 has a typical output of 0.5V for 100µg/m³ dust concentration.
0.578V should read around 116µg/m³ minus the zero offset.

[vojtishek]

Hello Shardan,

I was successfull to connect this sensor. I'm using Wemos D1 flashed on ESPEasy.
However the value is 221 and after application of formula -95.

What does this value specifically means?
Is this 95µg/m³?

Thank you for help.

Vojta

[vojtishek]

Just came into another formula, which in my eyes showing better value, but still confused as not possible to correlate against something.

calcVoltage = voMeasured * (3.3v / 1024.0);
dustDensity = (0.17 * calcVoltage - 0.1)*1000

So in my case when voltage for sensor is 3.3v and measure 221

calcVoltage = 221* (3.3/ 1024.0);
dustDensity = (0.17 * 0,71 - 0.1)*1000
-----------------------------------------------
26,7mg/m3

This can be close to reality as I used sensor in clear environment.

[Shardan]

Hello Vojta,

I've tested several GP2Y10 meanwhile. Sadly I have some bad news:
There is no "universal" formula for this sensor.

I tested a dozen of GP2Y10.
They are varying in zero point and slope more then 30%.
Says a formula working for me will most likely not work for your sensor.

So the formula has to be calibrated for every sensor individually.
Even worse the sensor values "move" over time due to dust adhered inside and aging of the sensor LED
so calibration has to be repeated now and then.

To be honest i don't know any way to get an exact calibration without using a very expensive commercial dust meter for reference.
To get a rough calibration at least i tried to go another way. The application sheet for GP2Y10 says:

・The output voltage Vo of this sensor is the sum of output voltage at no dust Voc and output proportional to dust density ΔV. Output proportional to dust density ΔV is shown as follows. ΔV = Vo - Voc (Vo : monitor value)

・Output voltage at no dust Voc is caused by the stray light ocuurring in this sensor. This sensor makes Voc voltage even at dust density 0mg/m3. If dust attached within this sensor increases, Voc becomes bigger. On the other hand, if dust attached within this sensor decreases, Voc becomes smaller.

・To store Voc in the memory of application is necessary to calculate ΔV from monitor value Vo. If monitor value Vo lower than the memorized Voc appears, this monitor value Vo should be stored in the memory of application as a new Voc.

・If monitor value Vo maintains a bigger value than the memorized Voc for a certain period of time, this monitor value Vo should be stored in the memory of application as a new Voc.

So i suggest the following steps:

- remove all formula values fom the task
- place the sensor near an open window at a rainy day (rain washes dust off the air!). Avoid smoking
- Let the sensor settle down for about an hour.
- Refresh page and note the shown value in the task every minute, get at least 5..10 values.

According to the application note I'd take the lowest value from the list so i never run into negative values. For example lets say the value is 1500.
The first step for formula is to get rid of this Zero Offset: Set Formula to "%value% - 1500".
(Replace the "1500" with your value from list! Remember, this is an example value!))

Now we can take care of the sensors slope value.
These parameters are given:
The sensor sketch sums up 25 results from the sensor.
The sensor gives an output voltage of 0.5V per 100µg/m³ typicaly.
The ADC from ESP8266 gives 1024 "ticks" for the 0 ... 3.3 Volt range (See circuit description above in this thread!!)

Let's do some calculation:

One "tick" from the ADC eqales to 3.3V / 1024 = 0,0032 V ( in the shown above circuit!!)

Now let's see what our value gives, As we have 25 measurements summarized we have to deal with that too.

Formula: (%value%-1500) * 0.0032 Changes "ticks" into voltage

Formula: ((%value%-1500) * 0.0032) / 25 Reduce the 25 summarized values to one average value.

Now we have a voltage. Datasheet says "Typical 0.5V per 100µg/m³", so let's "translate" the voltage into µg/m³

Formula: ((%value%-1500) * 0.0032 / 25) / 0.5 Gives the value in "100µg/m³" units, but we want 1µg/m³ units....

Formula: (((%value%-1500) * 0.0032 / 25) / 0,5) * 100

This formula can be somewhat reduced to:


(%value%-1500) * 0.0032 * 100
------------------------------------- = (%value%-1500) * 0,0032 * 8 = (%value%-1500) * 0.0256
25 * 0.5


So as a result: Formula: (%value%-1500) * 0.0256
Again, please keep in mind that "1500" is an example value and will not work for your sensor!

Reminder:
This is based on the datasheet's "typ" value for slope, 0,5V per 100µg/m³.
The min/max values are 0.35 ...0.65 V per 100µg/m³.
So this formula will give only an approximated value.
An exact calibration needs a professional dust meter for reference

I will test this these days and update the post accordingly.

Regards
Shardan

[vojtishek]

Hello,

that sounds logical. I will definitely test that and let you know.
Thanks for that.

We have a meteo station in my city which showing hourly results of dust PM10 and PM2.5. I know where is located, so maybe go there and compare the results.
This can be interesting.

[Shardan]

Hello all,

sadly my programming knowledge is about 20 years old and worthless nowadays.

There are some things that should be implemented with the "dust" module for the GP2Y10

Oversampling
The sketch starts an oversampling with adding 25 measurements from the sensor.
Any reason why the result is not divided by 25 to give average directly?

Zero point calibration "down".
Due to the application sheet:
The zero value (offset) should be stored in the sketch.
If a value lower as the stored zero value is measured this should be the new zero value (offset).

This shouldn't be too difficult to implement. Set a "ZeroV" variable to a high value at start and
check if to set it down with every measurement cycle.

Zero point calibration "up".
Over time the Offset moves due to dust adhered inside the sensor and aging of the sensor's LED.
If the sensor does not come down to the stored zero value, the lowest measured value over
a certain time should be the new zero value.

This might be more complex to implement.
Easiest way might be to reset the stored zero value to a high value and let it do the "down" way again.
Might be triggered manually? ("Recalibrate")
Disadvantage: As long as the new zero point is calibrated the measurement gives bad values.

Maybe an idea to calibrate zero offset automatically, without that disatvantage:
Use a variable "zeroV" for current measurement as described above, set it to a default value at start.
Set a second variable, e.g. "zeroCal" to a high value at start and calibrate it "down" as described above.
This may run with the usual measurement cycle of the sensor too.
Every now and then (Once a week?) copy the zeroCal value into
the zeroV current value variable, reset zeroCal to a high value and start the cycle again.

Another point that counts for this:
Store the zero value in the RAM or in the flash?
Storing in RAM says it is lost every time you reset / power the device and has to be recalibrated.
Storing in flash solves loosing the zero value at the cost of write cycles to the flash.

Maybe do both, RAM for current working value and flash for long time storing to reduce flash writes?
Maybe if we get a new "zeroCal" written to "zeroV" once a week, save it to flash and use it as the default value for next start?

I'm not sure about this.


Just my ideas

Regards
Shardan

[pasgabriele]

Congratulation Shardan, great work!!!

[Shardan]

Hello,

that sounds logical. I will definitely test that and let you know.
Thanks for that.

We have a meteo station in my city which showing hourly results of dust PM10 and PM2.5. I know where is located, so maybe go there and compare the results.
This can be interesting.

Definitely interesting.

In the formula (%value%-1500) * 0.0256 the "1500" stands for the zero offset.

The 0,0256 gives the slope, so how much a shown non-zero value changes at a given change of dust density.
This should correct the range of "0.35V ...0.65V per 100µg/m³" given from the sensor's datasheet.

This is important:
Always set the zero offset value first, as described above!
If this value isn't correct it's completely senseless to fiddle with the slope value.

If you have set a zero value for your sensor you may take the sensor to your local station.
Check your sensor's value against the station's value and change the "0,0256" value accordingly.


@pasgabriele

Thanks for the compliment
I hope i didn't make a stupid mistake somewhere ....


Regards
Shardan

[MonkeyTown]

Just wanted to say thanks to Shardan.

I have been playing around with this sensor for a few days. The output seems so fidgety and unreliable that I almost thought I have a dud sensor. Looking it up, it seems that not many people are having much luck with it.
I almost thought about buying a new one, after trying different formulas, using Arduino, NodeMcu and still getting the same inconsistent readings.

Thanks again for all the work you put into this. You might think your programming skills are a bit rusty, but your programmer's mind is still spot on! Cheers!

[vader]

@Shardan

I'm now also owner of that dust "toy". You wrote:

Formula: ((%value%-1500) * 0.0032 / 25) / 0.5 Gives the value in "100µg/m³" units, but we want 1µg/m³ units....
Formula: (((%value%-1500) * 0.0032 / 25) / 0,5) * 100

Is it not * 1000 to change the unit mg --> µg?

Regards

[Shardan]

@Shardan

I'm now also owner of that dust "toy". You wrote:

Formula: ((%value%-1500) * 0.0032 / 25) / 0.5 Gives the value in "100µg/m³" units, but we want 1µg/m³ units....
Formula: (((%value%-1500) * 0.0032 / 25) / 0,5) * 100

Is it not * 1000 to change the unit mg --> µg?

Regards

As said, you get a value in 100µg units, not 1000µg, to get 1µg you need the factor 100, not 1000.

There is a lot of things you have to take into calculation.
The sensor does not give a direct value. It gives a value of about 0,4mg/m³ = 3V, more or less.
The analog to digital converter can take 1V max, so you have to use a voltage divider.
The more the converter does not give a voltage value but "ticks", every tick stands for a voltage step.

And so on. All these factors have to be calculated into that formula so the formulas get quite crude.

Anyways some things about the GP2Y10 should be mentioned.
It was never designed to give clear readings - it does not differ between PM1.0, PM2.5 and PM10.
I've tried to calibrate it with a commercial dust meter, at last i gave in.
This sensor was designed for switching air cleaners and such, just by comparing a given value with the sensor value.
It was made for this purpose and it does the job well. For home control driving an air cleaner it works.
As it does not differ particle sizes so if the particle size differs you get different values with the same concentration of dust .
So as a conclusion a real calibration is more or less impossible. What you can get is a "realtive Dust" value, showing
the over all air quality.

If you need real dust concentration metering i recommend using PMSx003 sensors, the plugin is
availlable now with ESPEasy 2.0.0 testing or development versions.

Regards
Shardan

[vader]

Are your values also jumping from some µg (<10) to over 100µg and so on...? Look a bit useless those values!?

You wrote "As said, you get a value in 100µg units". Must it then not be divided by 100, instead of multiplied, to get 1µg units?

[Shardan]

if you get 0,1 as a reading with a 100µg unit this is 10µg as it says mathematically 0,1 * 100µg.
So you have to multiply by 100......
If you want to translate Kg into g you won't divide by 1000...

[vader]

if you get 0,1 as a reading with a 100µg unit this is 10µg as it says mathematically 0,1 * 100µg.
So you have to multiply by 100......
If you want to translate Kg into g you won't divide by 1000...

Sure? Check this out from a web page for pupils:

Welche Gewichtseinheiten gibt es?
Für Gewichte verwendet man folgende Einheiten:
1 Tonne (t). Eine Tonne sind 10^3 Kilogramm.
1 Kilogramm (kg). Ein Kilogramm sind 1000 Gramm.
1 Gramm (g). Ein Gramm sind 1000 Milligramm oder 10^-3 Kilogramm.
1 Milligramm (mg). Ein Milligramm entspricht 1000 Mikrogramm oder 10^-6 Kilogramm.
1 Mikrogramm (myg). Ein Mikrogramm entspricht 1000 Nanogramm oder 10^-9 Kilogramm.
1 Nanogramm (ng). Ein Nanogramm entspricht 10^-12 Gramm.

[Shardan]

Yes i am sure.

If you get a reading of 1 in units 100µg it represents 100 µg. To get a 1µg reading you have to multiply by 100.

If you divide a KG by 100 you get tons, not grams.

[vader]

Sorry, I still do not agree. Let's take your example from above: I have a piece of e.g. meat and I share this with 99 people. Everyone get the same amount. How can a person have more than the original weight of 1kg? That's 10g what each person get. 10g x 100 persons = 1000g = 1kg. How do you come on tons?

[Shardan]

What is a unit in physics?
It is a predefined value as 1g or 1m.

A unit is multiplied by a numer to give a size, weight or whatever.
If a tube is 50 x 1cm we say it is 50cm long.

The dust level gives a unit of 100µg.
So if the reading from sensor is "1" this says it is 1 x the unit of 100µg = 100µg

If i want to send "100" to my controller instead of the "1" i should really divide the 1 by 100?
Not seriously......

[vader]

That would be 100 x 0.1mg = 10mg, not µg. Right so far?

[Shardan]

No. Definitely not.

If you have a reading of 1 and a unit 100µg this is 100µg. If you multiply the read "1" by 10 you have to divide the unit by 10 to get the same value.
Same as with equations where you have to multiply/divide both sides.
If you multiply the read value by 100 you have to divide the unit by 100 to get the same value.

So if i have a reading of 1 in units of 100µg and multiply the reading by 100 i have to divide the unit by 100.

So 1 * 100 µg multiplied by 100 gives 100 * 1µg or 100 µg

[vader]

Ok, I see. But I decided not to mount this sensor into my SH network, because the values it gives are too "random". Thanks anyway.

[Shardan]

Ok, I see. But I decided not to mount this sensor into my SH network, because the values it gives are too "random". Thanks anyway.

You may try to use a "Testing" or "Development" Version of ESPEasy-2.0.0 you might use a sensor from the PMS-x003 series.
It is a laser based sensor giving detailed values of PM1.0, PM2.5 and PM10.

It should be mentioned that even these sensors are cheap (around 15..20 USD) compared to professional sensoring devices (> 500 USD).
The laser based sensors will give more exact values then the GP2Y10. Anyways, as always you get what you pay for: I won't rank them as a scientific grade sensor......

Regards
Shardan

[vader]

I'm allready using a self-compiled version 2.0.0-dev11, because some moduls have to be modified for my own needs. And yes, i've read about that sensors, but they are too expensive (>$10) for just a hobby experiment. If there is too much smoke, I open a window. And if there is too much dust, I let my wife clean it. All without any sensors, hehe....

[grovkillen]

I'm allready using a self-compiled version 2.0.0-dev11, because some moduls have to be modified for my own needs. And yes, i've read about that sensors, but they are too expensive (>$10) for just a hobby experiment. If there is too much smoke, I open a window. And if there is too much dust, I let my wife clean it. All without any sensors, hehe....

What if you or your wife (shame on you for not cleaning) is not there to open the window?

[vader]

It's better they have all the time something to do, so they won't come on stupid ideas...