Skip to contentSkip to conversation

iio: ppm: Add IIO_PPM channel type

RFC In review Matt Ranostay 3 participants 11 messages Aug 27, 2015 – Sep 4, 2015 IIO

lore
Share & tools
lore
#0 lore raw

RFC 1/1 iio: ppm: Add IIO_PPM channel type

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

---
 Documentation/ABI/testing/sysfs-bus-iio | 7 +++++++
 drivers/iio/industrialio-core.c         | 1 +
 include/uapi/linux/iio/types.h          | 1 +
 3 files changed, 9 insertions(+)
Documentation/ABI/testing/sysfs-bus-iio +7 -0
index 42d360f..79366a3 100644
@@ -1459,3 +1459,10 @@ Description:
 		measurements and return the average value as output data. Each
 		value resulted from <type>[_name]_oversampling_ratio measurements
 		is considered as one sample for <type>[_name]_sampling_frequency.
+
+What:		/sys/bus/iio/devices/iio:deviceX/in_ppm_raw
+KernelVersion:	4.3
+Contact:	[email protected]
+Description:
+		Raw (unscaled no offset etc.) parts per million measure reading
+		of data like CO2 or VOC (Volatile Organic Compounds) substances.
drivers/iio/industrialio-core.c +1 -0
index b3fcc2c..65285c7 100644
@@ -75,6 +75,7 @@ static const char * const iio_chan_type_name_spec[] = {
 	[IIO_ENERGY] = "energy",
 	[IIO_DISTANCE] = "distance",
 	[IIO_VELOCITY] = "velocity",
+	[IIO_PPM] = "ppm",
 };
 
 static const char * const iio_modifier_names[] = {
include/uapi/linux/iio/types.h +1 -0
index 2f8b117..7113f9b 100644
@@ -35,6 +35,7 @@ enum iio_chan_type {
 	IIO_ENERGY,
 	IIO_DISTANCE,
 	IIO_VELOCITY,
+	IIO_PPM,
 };
 
 enum iio_modifier {
Signature · 1 line
1.9.1
#1 lore raw
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration. But that still doesn't tell us what is actually being measured. Concentration is always something of unit Y per unit of X. Knowing Y and X is in my opinion important to makes sense of the data. I'm not sure how to express these though.

Maybe a modifier for what is measured and then define that the rest of the units cancel out. Like CO2 concentration is CO2 molecules per molecule so you'd just end up with CO2 for the modiifer and use the scale to define whether it is %, ‰, ppm, ppb, ppt.

Given that concentration will not be larger than 1 it makes sense to use something small as the standard unit and then use the scale factor to scale up or down.

E.g. lets say we make ppm the standard unit, then a scale 1 means ppm, 0.0001 means ppb, 1000 means ‰, 10000 means % and so on.

- Lars

Quoted 48 lines
---
 Documentation/ABI/testing/sysfs-bus-iio | 7 +++++++
 drivers/iio/industrialio-core.c         | 1 +
 include/uapi/linux/iio/types.h          | 1 +
 3 files changed, 9 insertions(+)
Documentation/ABI/testing/sysfs-bus-iio +7 -0
index 42d360f..79366a3 100644
@@ -1459,3 +1459,10 @@ Description:
 		measurements and return the average value as output data. Each
 		value resulted from <type>[_name]_oversampling_ratio measurements
 		is considered as one sample for <type>[_name]_sampling_frequency.
+
+What:		/sys/bus/iio/devices/iio:deviceX/in_ppm_raw
+KernelVersion:	4.3
+Contact:	[email protected]
+Description:
+		Raw (unscaled no offset etc.) parts per million measure reading
+		of data like CO2 or VOC (Volatile Organic Compounds) substances.
drivers/iio/industrialio-core.c +1 -0
index b3fcc2c..65285c7 100644
@@ -75,6 +75,7 @@ static const char * const iio_chan_type_name_spec[] = {
 	[IIO_ENERGY] = "energy",
 	[IIO_DISTANCE] = "distance",
 	[IIO_VELOCITY] = "velocity",
+	[IIO_PPM] = "ppm",
 };
 
 static const char * const iio_modifier_names[] = {
include/uapi/linux/iio/types.h +1 -0
index 2f8b117..7113f9b 100644
@@ -35,6 +35,7 @@ enum iio_chan_type {
 	IIO_ENERGY,
 	IIO_DISTANCE,
 	IIO_VELOCITY,
+	IIO_PPM,
 };
 
 enum iio_modifier {
#2 lore raw
Quoted 10 lines · Lars-Peter Clausen
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

- Lars

[1] https://en.wikipedia.org/wiki/Concentration

#3 lore raw
Quoted 23 lines · Lars-Peter Clausen
Quoted 2 lines · Lars-Peter Clausen
Quoted 1 line · Matt Ranostay

There are air quality sensors that report data back in parts per

million

Quoted 1 line
Quoted 1 line

of VOC (Volatile Organic Compounds) which are usually indexed from

CO2

Quoted 3 lines
Quoted 3 lines

or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types

fit

Quoted 3 lines
Quoted 1 line

this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's

more of a

Quoted 1 line

scale. The type would be concentration.[...]

Reading a bit more[1], concentration doesn't actually seem to be the right
term in this case, the correct term is mole fraction. Maybe we can use that
as the type. That also makes it clear that the unit is molecules per molecule.

Quoted 3 lines

- Lars

[1] https://en.wikipedia.org/wiki/Concentration

Works for me.

I'm tending to prefer base units that don't have an arbitrary scale factor and use the
IIO types to define. Molecules per molecule then use scale would be my preference.

Agreed on using modifiers for what we are measuring. Guessing there aren't that many sensor types out there. If it gets really silly we can add a means of listing chemicals detected!

J

Signature · 1 line
Sent from my Android device with K-9 Mail. Please excuse my brevity.
#4 lore raw
In reply to Lars-Peter Clausen
Quoted 9 lines · Lars-Peter Clausen
Quoted 8 lines · Lars-Peter Clausen
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Quoted 7 lines
Quoted 2 lines

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

Actually we can't use mole fraction for this because we aren't in a chemistry lab, and know the other compounds that make up the local atmosphere. Besides you'd have to include some insane lookup table for molar mass of carbon or whatever VOC being measured :)

IIO_CONCENTRATION could be useful since you could just change the scale to make it ppm or ppb or whatever.

Quoted 3 lines

- Lars

[1] https://en.wikipedia.org/wiki/Concentration

#5 lore raw
Quoted 13 lines · Matt Ranostay
Quoted 12 lines · Lars-Peter Clausen
Quoted 10 lines · Lars-Peter Clausen
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Quoted 8 lines
Quoted 4 lines

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

Actually we can't use mole fraction for this because we aren't in a chemistry lab, and know the other compounds that make up the local atmosphere. Besides you'd have to include some insane lookup table for molar mass of carbon or whatever VOC being measured :)

I don't think you'd need that. Mole fraction tells you the number of molecules of something per total number of molecules. You don't need the mass for this.

But what exactly is the sensor measuring? CO2 (or VOC) molecules per total number of molecules or number of CO2 molecules in a particular volume?

Quoted 12 lines

IIO_CONCENTRATION could be useful since you could just change the scale to make it ppm or ppb or whatever.

Quoted 3 lines

- Lars

[1] https://en.wikipedia.org/wiki/Concentration

--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#6 lore raw
Quoted 23 lines · Lars-Peter Clausen
Quoted 13 lines · Matt Ranostay
Quoted 12 lines · Lars-Peter Clausen
Quoted 10 lines · Lars-Peter Clausen
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Quoted 8 lines
Quoted 4 lines

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

Actually we can't use mole fraction for this because we aren't in a chemistry lab, and know the other compounds that make up the local atmosphere. Besides you'd have to include some insane lookup table for molar mass of carbon or whatever VOC being measured :)

Quoted 7 lines

I don't think you'd need that. Mole fraction tells you the number of molecules of something per total number of molecules. You don't need the mass for this.

But what exactly is the sensor measuring? CO2 (or VOC) molecules per total number of molecules or number of CO2 molecules in a particular volume?

CC'ed my Swiss colleague on this because he knows much on the interworkings of VOC sensors than I could hope to.
So simply these sensors are finding VOCs (which the sensor in question does CO2 and tVOC indexes.. probably not too independent of each other)

But molecules are not parts.. think of taking an X volume of air and figuring what is precent of oxygen is verses nitrogen, argon, etc, etc. Sure the highest is nitrogen at atomic weight 14.01 with ~78% of the "air", oxygen is ~20% at the weight of 16.00, and etc.

Think about cutting cube into a millions of pieces and figuring an X% is Y substance which you can detect, but can't detect X, Y, and Z (think any particles in 'air' that aren't bonded with carbon). So mole fractions are impossible here, and you could only take a parts in a known volume.

Most VOCs (if not all, bit of newbie here) use UV LEDs to ionize particles, and with some maths calculate the parts-per-million.

Quoted 13 lines
Quoted 12 lines

IIO_CONCENTRATION could be useful since you could just change the scale to make it ppm or ppb or whatever.

Quoted 3 lines

- Lars

[1] https://en.wikipedia.org/wiki/Concentration

--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#7 lore raw
Quoted 42 lines · Matt Ranostay
Quoted 23 lines · Lars-Peter Clausen
Quoted 13 lines · Matt Ranostay
Quoted 12 lines · Lars-Peter Clausen
Quoted 10 lines · Lars-Peter Clausen
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Quoted 8 lines
Quoted 4 lines

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

Actually we can't use mole fraction for this because we aren't in a chemistry lab, and know the other compounds that make up the local atmosphere. Besides you'd have to include some insane lookup table for molar mass of carbon or whatever VOC being measured :)

Quoted 7 lines

I don't think you'd need that. Mole fraction tells you the number of molecules of something per total number of molecules. You don't need the mass for this.

But what exactly is the sensor measuring? CO2 (or VOC) molecules per total number of molecules or number of CO2 molecules in a particular volume?

CC'ed my Swiss colleague on this because he knows much on the interworkings of VOC sensors than I could hope to.
So simply these sensors are finding VOCs (which the sensor in question does CO2 and tVOC indexes.. probably not too independent of each other)

But molecules are not parts.. think of taking an X volume of air and figuring what is precent of oxygen is verses nitrogen, argon, etc, etc. Sure the highest is nitrogen at atomic weight 14.01 with ~78% of the "air", oxygen is ~20% at the weight of 16.00, and etc.

Quoted 8 lines

Think about cutting cube into a millions of pieces and figuring an X% is Y substance which you can detect, but can't detect X, Y, and Z (think any particles in 'air' that aren't bonded with carbon). So mole fractions are impossible here, and you could only take a parts in a known volume.

Most VOCs (if not all, bit of newbie here) use UV LEDs to ionize particles, and with some maths calculate the parts-per-million.

But what per what? PPM is a completely ambiguous unit if you don't specify parts of what per million of what.

#8 lore raw

Ah true but we could add a few IIO_MOD* modifiers for CO2, tVOC, VOC, etc, etc. I think Jonathan was suggesting that.

Thanks,

Matt

Quoted 55 lines · Lars-Peter Clausen
Quoted 42 lines · Matt Ranostay
Quoted 23 lines · Lars-Peter Clausen
Quoted 13 lines · Matt Ranostay
Quoted 12 lines · Lars-Peter Clausen
Quoted 10 lines · Lars-Peter Clausen
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Quoted 8 lines
Quoted 4 lines

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

Actually we can't use mole fraction for this because we aren't in a chemistry lab, and know the other compounds that make up the local atmosphere. Besides you'd have to include some insane lookup table for molar mass of carbon or whatever VOC being measured :)

Quoted 7 lines

I don't think you'd need that. Mole fraction tells you the number of molecules of something per total number of molecules. You don't need the mass for this.

But what exactly is the sensor measuring? CO2 (or VOC) molecules per total number of molecules or number of CO2 molecules in a particular volume?

CC'ed my Swiss colleague on this because he knows much on the interworkings of VOC sensors than I could hope to.
So simply these sensors are finding VOCs (which the sensor in question does CO2 and tVOC indexes.. probably not too independent of each other)

But molecules are not parts.. think of taking an X volume of air and figuring what is precent of oxygen is verses nitrogen, argon, etc, etc. Sure the highest is nitrogen at atomic weight 14.01 with ~78% of the "air", oxygen is ~20% at the weight of 16.00, and etc.

Quoted 8 lines

Think about cutting cube into a millions of pieces and figuring an X% is Y substance which you can detect, but can't detect X, Y, and Z (think any particles in 'air' that aren't bonded with carbon). So mole fractions are impossible here, and you could only take a parts in a known volume.

Most VOCs (if not all, bit of newbie here) use UV LEDs to ionize particles, and with some maths calculate the parts-per-million.

But what per what? PPM is a completely ambiguous unit if you don't specify parts of what per million of what.

#9 lore raw
Quoted 2 lines · Matt Ranostay

Ah true but we could add a few IIO_MOD* modifiers for CO2, tVOC, VOC, etc, etc. I think Jonathan was suggesting that.

What I meant was CO2 measured in what unit per what unit.

Like is it weight per volume or volume per volume or weight per weight or molecules per volume ...

- Lars

Quoted 61 lines

Thanks,

Matt

Quoted 55 lines · Lars-Peter Clausen
Quoted 42 lines · Matt Ranostay
Quoted 23 lines · Lars-Peter Clausen
Quoted 13 lines · Matt Ranostay
Quoted 12 lines · Lars-Peter Clausen
Quoted 10 lines · Lars-Peter Clausen
Quoted 6 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Quoted 8 lines
Quoted 4 lines

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

Actually we can't use mole fraction for this because we aren't in a chemistry lab, and know the other compounds that make up the local atmosphere. Besides you'd have to include some insane lookup table for molar mass of carbon or whatever VOC being measured :)

Quoted 7 lines

I don't think you'd need that. Mole fraction tells you the number of molecules of something per total number of molecules. You don't need the mass for this.

But what exactly is the sensor measuring? CO2 (or VOC) molecules per total number of molecules or number of CO2 molecules in a particular volume?

CC'ed my Swiss colleague on this because he knows much on the interworkings of VOC sensors than I could hope to.
So simply these sensors are finding VOCs (which the sensor in question does CO2 and tVOC indexes.. probably not too independent of each other)

But molecules are not parts.. think of taking an X volume of air and figuring what is precent of oxygen is verses nitrogen, argon, etc, etc. Sure the highest is nitrogen at atomic weight 14.01 with ~78% of the "air", oxygen is ~20% at the weight of 16.00, and etc.

Quoted 8 lines

Think about cutting cube into a millions of pieces and figuring an X% is Y substance which you can detect, but can't detect X, Y, and Z (think any particles in 'air' that aren't bonded with carbon). So mole fractions are impossible here, and you could only take a parts in a known volume.

Most VOCs (if not all, bit of newbie here) use UV LEDs to ionize particles, and with some maths calculate the parts-per-million.

But what per what? PPM is a completely ambiguous unit if you don't specify parts of what per million of what.

#10 lore raw

Ok that is pretty sane. I'll submit my full patchset for the driver marked RFC to see if my use of types and modifiers makes sense. Well once I get the module off the slow boat from Taiwan and test it

Sent from my iPhone

Quoted 72 lines · Lars-Peter Clausen
Quoted 4 lines · Matt Ranostay

Ah true but we could add a few IIO_MOD* modifiers for CO2, tVOC, VOC, etc, etc. I think Jonathan was suggesting that.

What I meant was CO2 measured in what unit per what unit.

Like is it weight per volume or volume per volume or weight per weight or molecules per volume ...

- Lars

Quoted 58 lines

Thanks,

Matt

Quoted 53 lines · Lars-Peter Clausen
Quoted 41 lines · Matt Ranostay
Quoted 23 lines · Lars-Peter Clausen
Quoted 14 lines · Matt Ranostay
Quoted 13 lines · Lars-Peter Clausen
Quoted 11 lines · Lars-Peter Clausen
Quoted 7 lines · Matt Ranostay

There are air quality sensors that report data back in parts per million of VOC (Volatile Organic Compounds) which are usually indexed from CO2 or another common pollutant.

This patchset adds an IIO_PPM type because no other channels types fit this use case.

Hm, I'm not sure if parts-per-million is a good channel type. It's more of a scale. The type would be concentration.[...]

Quoted 7 lines
Quoted 3 lines

Reading a bit more[1], concentration doesn't actually seem to be the right term in this case, the correct term is mole fraction. Maybe we can use that as the type. That also makes it clear that the unit is molecules per molecule.

Actually we can't use mole fraction for this because we aren't in a chemistry lab, and know the other compounds that make up the local atmosphere. Besides you'd have to include some insane lookup table for molar mass of carbon or whatever VOC being measured :)

Quoted 6 lines

I don't think you'd need that. Mole fraction tells you the number of molecules of something per total number of molecules. You don't need the mass for this.

But what exactly is the sensor measuring? CO2 (or VOC) molecules per total number of molecules or number of CO2 molecules in a particular volume?

CC'ed my Swiss colleague on this because he knows much on the interworkings of VOC sensors than I could hope to.
So simply these sensors are finding VOCs (which the sensor in question does CO2 and tVOC indexes.. probably not too independent of each other)

But molecules are not parts.. think of taking an X volume of air and figuring what is precent of oxygen is verses nitrogen, argon, etc, etc. Sure the highest is nitrogen at atomic weight 14.01 with ~78% of the "air", oxygen is ~20% at the weight of 16.00, and etc.

Quoted 8 lines

Think about cutting cube into a millions of pieces and figuring an X% is Y substance which you can detect, but can't detect X, Y, and Z (think any particles in 'air' that aren't bonded with carbon). So mole fractions are impossible here, and you could only take a parts in a known volume.

Most VOCs (if not all, bit of newbie here) use UV LEDs to ionize particles, and with some maths calculate the parts-per-million.

But what per what? PPM is a completely ambiguous unit if you don't specify parts of what per million of what.