Portable display PM2.5
Mini device which display PM2.5 with temperateure and humidity. Work on battery for weeks, display last measure, hourly and AQI value.
This is my 3rd or 4th device for PM2.5 display.
I developed this portable value as I want to know everywhere I go what are the condition of pollution for running inside or outside.
The condition of this device was to be essential, meaning just measure the value, calculate the hourly average and the AQI. To be portable, it must be power on battery and keep charged for days and weeks.
For simplicity, I add a LED display color for easy reading.
Technical description: Hardware :
- Sensor is the PMS7000. this choice because it is small and fan embedded
- The uC is a PIC18F2520
- The sensor is powerd by a real shutdown DC/DC boost: Vbat=>5v (100mA)
- When a measurement need to be done, need power 30 sec before the first measurement is valid
- The LED are low consumption (2ma max)
- Display with 8*2 cars. This display is B&W, very low consumption and easy to buy on internet
- I add a sensor SHT21 for temp and humidity (not necessary just nice to have)
- LDR for night detection
- Battery charging with USB. Option for solar power recharge
- Overall the device consumption is 2mA, except during measurement phase which is 100mA
Technical description: Software :
- Develop with MikroC for PIC18F2520
- Several display mode:
* Last measure and AQI
* Hourly average and AQI
* Battery voltage and Last measure, Hourly average and AQI
* Instant measurement: Sensor always ON and real time display value
- Choice to display the LED based on last measyre, hourly measure or AQI
- 2 choice for the AQI calculation: USA standard one (24h average rolling basis) or the new calculation method, which is 12h basis rolling with some weight on the latest measurement
- Sampling particule measurement choice, from every 10min to every hour. this is the major impact on the autonomy. Every hour, an average of all the measurement made during this hour is recorded and used for the AQI calculation
- Choice to keep or remove the LED display during the night (LDR light detection)
After mock up develop and validated, create a small PCB with SMT components, 5cm*4cm.
All the components can feet into a standard small casing 9.5cm*6cm.2.5cm. I'm designing a dedicated small casing for 3D print to optimize more the space.
Some consumption optimisation need to be done but overall, this device is working fine, maybe too fine as it can show some bad environment. At least, we can see that an air purifier in a kids's room is working very well.
PCB need to be improve as connecters are too big and some layout error.
Laurent
I developed this portable value as I want to know everywhere I go what are the condition of pollution for running inside or outside.
The condition of this device was to be essential, meaning just measure the value, calculate the hourly average and the AQI. To be portable, it must be power on battery and keep charged for days and weeks.
For simplicity, I add a LED display color for easy reading.
Technical description: Hardware :
- Sensor is the PMS7000. this choice because it is small and fan embedded
- The uC is a PIC18F2520
- The sensor is powerd by a real shutdown DC/DC boost: Vbat=>5v (100mA)
- When a measurement need to be done, need power 30 sec before the first measurement is valid
- The LED are low consumption (2ma max)
- Display with 8*2 cars. This display is B&W, very low consumption and easy to buy on internet
- I add a sensor SHT21 for temp and humidity (not necessary just nice to have)
- LDR for night detection
- Battery charging with USB. Option for solar power recharge
- Overall the device consumption is 2mA, except during measurement phase which is 100mA
Technical description: Software :
- Develop with MikroC for PIC18F2520
- Several display mode:
* Last measure and AQI
* Hourly average and AQI
* Battery voltage and Last measure, Hourly average and AQI
* Instant measurement: Sensor always ON and real time display value
- Choice to display the LED based on last measyre, hourly measure or AQI
- 2 choice for the AQI calculation: USA standard one (24h average rolling basis) or the new calculation method, which is 12h basis rolling with some weight on the latest measurement
- Sampling particule measurement choice, from every 10min to every hour. this is the major impact on the autonomy. Every hour, an average of all the measurement made during this hour is recorded and used for the AQI calculation
- Choice to keep or remove the LED display during the night (LDR light detection)
After mock up develop and validated, create a small PCB with SMT components, 5cm*4cm.
All the components can feet into a standard small casing 9.5cm*6cm.2.5cm. I'm designing a dedicated small casing for 3D print to optimize more the space.
Some consumption optimisation need to be done but overall, this device is working fine, maybe too fine as it can show some bad environment. At least, we can see that an air purifier in a kids's room is working very well.
PCB need to be improve as connecters are too big and some layout error.
Laurent
Discussie (4 opmerking(en))
HaSch 3 jaar geleden
Hi Laurent,
meanwhile, I got all parts including PCB and soldered. Flashing of PIC was successful. Powering using battery seems to fail, I have to check. If I power using USB it comes to life. Not all LEDs shine at initialisation, I have to check, too.
First of all I noticed black segments on display, is it possible to adjust contrast?
Best regards,
Hans
meanwhile, I got all parts including PCB and soldered. Flashing of PIC was successful. Powering using battery seems to fail, I have to check. If I power using USB it comes to life. Not all LEDs shine at initialisation, I have to check, too.
First of all I noticed black segments on display, is it possible to adjust contrast?
Best regards,
Hans
Antwoord
gfaman 3 jaar geleden
Hi Hans,
can you send me an email or whatsapp contact for easy discussion ? You can use this email to contact me: laurent.elektor@gmail.com
This display never requested contrast adjustment. I used hundreds of it. However, it might be possible by SW, need to check.
If the contrast is bad, either the power supply has an issue (3.3v) or the ceramic capacitor to boost the voltage has problem. Also, I had several time some short circuit on the connector (not easy to solder) or dry solder (not connected). Please send a picture of the display to see if I had this issue before.
When the PCBA is power on, you can check the 3.3v level and also the 5V (during the first 30 sec), as the Sw will launch immediate particle measurement. However, during the init phase, all the leds must light on during short time. If not, the SW didn t start well.
BR,
Laurent
can you send me an email or whatsapp contact for easy discussion ? You can use this email to contact me: laurent.elektor@gmail.com
This display never requested contrast adjustment. I used hundreds of it. However, it might be possible by SW, need to check.
If the contrast is bad, either the power supply has an issue (3.3v) or the ceramic capacitor to boost the voltage has problem. Also, I had several time some short circuit on the connector (not easy to solder) or dry solder (not connected). Please send a picture of the display to see if I had this issue before.
When the PCBA is power on, you can check the 3.3v level and also the 5V (during the first 30 sec), as the Sw will launch immediate particle measurement. However, during the init phase, all the leds must light on during short time. If not, the SW didn t start well.
BR,
Laurent
Antwoord
Toon meer
1 Opmerking(en)
HaSch 3 jaar geleden
I tried to create a BOM by myself using information in schematic. Doing so, I noticed that most of the capacitors seems to be polarized capacitors. But not on all symbols is a "+" marker. Are really all capacitors polarized types? Are all 1uF/10V, 10uF/10V and 10uF/6V3 polarized and 100nF not polarized?
Antwoord
gfaman 3 jaar geleden
Hi Hans,
first thanks for the idea on the labels changed for the schematic, works well.
All the capacitors on this projects are ceramic, so, no polarized. I seldom use now polarized (even if they are more stable) as often it takes more space. So, no all capacitors will be ok with ceramics and no worry about the +/-.
Laurent
first thanks for the idea on the labels changed for the schematic, works well.
All the capacitors on this projects are ceramic, so, no polarized. I seldom use now polarized (even if they are more stable) as often it takes more space. So, no all capacitors will be ok with ceramics and no worry about the +/-.
Laurent
Antwoord
Toon meer
1 Opmerking(en)
HaSch 3 jaar geleden
Hi Laurent,
I really recommend to provide a correct Bill Of Material as the circuit seems to include wrong parts. You listed e.g. an LP2980IM5-X5.0 (5V-type) which might be a LP2980IM5-X3.0 (3V-type). Maybe there are other mistakes on it.
Best regards,
Hans
I really recommend to provide a correct Bill Of Material as the circuit seems to include wrong parts. You listed e.g. an LP2980IM5-X5.0 (5V-type) which might be a LP2980IM5-X3.0 (3V-type). Maybe there are other mistakes on it.
Best regards,
Hans
Antwoord
gfaman 3 jaar geleden
Hi Hans,
sorry for the inconvenience. Yes, on the schematics, sometimes I forgot to update some values and the BOM may be wrong, even if in the Elektor article, it is indicated.
I'm updating the schematic with some adjustment on the BOM. However, some comments:
- The LM2980IM5 is 3.3v: LP2980IM5-X3.3
- The uC is 18LF2520 and not F (which is not recommended for 3.3v)
- On the schematic, the LTC4054 seems wrong, based on the drawing. I didn t find the right figure. The signal on the pin number are ok but the name on the signal are wrong. Anyway, the layout are ok.
Do not forget that all the resistance of the LED depend of the LED you can find. Most of time I need adjust with real LED I have, taking account that I always look for high brightness and low consumption. The value on the schematic are indicative.
Will post the latest file when I finish to check
BR,
Laurent
sorry for the inconvenience. Yes, on the schematics, sometimes I forgot to update some values and the BOM may be wrong, even if in the Elektor article, it is indicated.
I'm updating the schematic with some adjustment on the BOM. However, some comments:
- The LM2980IM5 is 3.3v: LP2980IM5-X3.3
- The uC is 18LF2520 and not F (which is not recommended for 3.3v)
- On the schematic, the LTC4054 seems wrong, based on the drawing. I didn t find the right figure. The signal on the pin number are ok but the name on the signal are wrong. Anyway, the layout are ok.
Do not forget that all the resistance of the LED depend of the LED you can find. Most of time I need adjust with real LED I have, taking account that I always look for high brightness and low consumption. The value on the schematic are indicative.
Will post the latest file when I finish to check
BR,
Laurent
Antwoord
HaSch 3 jaar geleden
Hi Laurent,
you said:
"On the schematic, the LTC4054 seems wrong, based on the drawing. I didn t find the right figure. The signal on the pin number are ok but the name on the signal are wrong. Anyway, the layout are ok."
I looked at the datasheet of LTC4054-4.2 <https://www.analog.com/media/en/technical-documentation/data-sheets/405442xf.pdf> and the signals and pins seems to be different.
In your schematic there are:
1: USB
2: GND
3: CHG
4: DC
5: BAT
In datasheet:
1: CHG
2: GND
3: BAT
4: VCC
5: PRG
That's quite different. E.g. in your schematic pin 1 is connected to USB but is the CHG output. What is wrong? Did you use another LTC4054?
you said:
"On the schematic, the LTC4054 seems wrong, based on the drawing. I didn t find the right figure. The signal on the pin number are ok but the name on the signal are wrong. Anyway, the layout are ok."
I looked at the datasheet of LTC4054-4.2 <https://www.analog.com/media/en/technical-documentation/data-sheets/405442xf.pdf> and the signals and pins seems to be different.
In your schematic there are:
1: USB
2: GND
3: CHG
4: DC
5: BAT
In datasheet:
1: CHG
2: GND
3: BAT
4: VCC
5: PRG
That's quite different. E.g. in your schematic pin 1 is connected to USB but is the CHG output. What is wrong? Did you use another LTC4054?
Antwoord
gfaman 3 jaar geleden
Hi Hans, sorry for the delay.
As I mentioned to you before, the figure of the LTC4054 is wrong (label wrong) on my schematic (I didn t find it on Diptrace) but the pin number are right. Before I was using EUP4054 but I found LTC4054 was compatible and easier to find.
On the schematic, the pin 1 is CHG which is the output for the LED charging indicator
Pin 2 is ground
Pin3 is battery and the pin3 of the IC is connected to the battery (after the switch ON/OFF
Pin4 is the VCC or DC input, coming from the solar panel or the USB
Pin5 is the current adjustment
So, yes, the labels inside the component figure are wrong but the pin number are well connected to the right function
Sorry for the confusion. I should have create the component.
Hope it is more clear now.
As I mentioned to you before, the figure of the LTC4054 is wrong (label wrong) on my schematic (I didn t find it on Diptrace) but the pin number are right. Before I was using EUP4054 but I found LTC4054 was compatible and easier to find.
On the schematic, the pin 1 is CHG which is the output for the LED charging indicator
Pin 2 is ground
Pin3 is battery and the pin3 of the IC is connected to the battery (after the switch ON/OFF
Pin4 is the VCC or DC input, coming from the solar panel or the USB
Pin5 is the current adjustment
So, yes, the labels inside the component figure are wrong but the pin number are well connected to the right function
Sorry for the confusion. I should have create the component.
Hope it is more clear now.
Antwoord
HaSch 3 jaar geleden
Hi Laurent,
you don't need to find correct parts in DipTrace library. It's easy to change labels on pin numbers or set right part names in schematic and even in PCB layout. You only have to choose the right footprint. Then you can highlight the part in schematic or PCB layout and rightclick in "Design Manager" or "Part Pins". Choose "Properties" and edit all you want. That's it!
Best regards,
Hans
you don't need to find correct parts in DipTrace library. It's easy to change labels on pin numbers or set right part names in schematic and even in PCB layout. You only have to choose the right footprint. Then you can highlight the part in schematic or PCB layout and rightclick in "Design Manager" or "Part Pins". Choose "Properties" and edit all you want. That's it!
Best regards,
Hans
Antwoord
gfaman 3 jaar geleden
Hi Hans,
yes, it is a 5V version, 175mA. This is to power the dust sensor which need 5V / 100mA. In fact, in the datasheet of the sensor, they wrote that all the sensor works in 3.3v but they need 5V for the embedded fan. So, we have no choice than to use this Dc/Dc 5v. The most difficult is to solder it. I tried several ways and the best one is still hot air gun with some solder past.
For the inductance, 10uH, several models can be used. I tried randomly several reference (10uH / Isat > 400mA) as I cannot find here the one recommended in the datasheet:
Some recommended inductor examples are Murata LQH32C and Coilcraft LPO4812, LPO3310, DO3314, DS1608 and MSS4020
Laurent
yes, it is a 5V version, 175mA. This is to power the dust sensor which need 5V / 100mA. In fact, in the datasheet of the sensor, they wrote that all the sensor works in 3.3v but they need 5V for the embedded fan. So, we have no choice than to use this Dc/Dc 5v. The most difficult is to solder it. I tried several ways and the best one is still hot air gun with some solder past.
For the inductance, 10uH, several models can be used. I tried randomly several reference (10uH / Isat > 400mA) as I cannot find here the one recommended in the datasheet:
Some recommended inductor examples are Murata LQH32C and Coilcraft LPO4812, LPO3310, DO3314, DS1608 and MSS4020
Laurent
Antwoord
HaSch 3 jaar geleden
Is it really a PIC18LF2580 and not a PIC 18LF2520? In the project article one can read PIC18F2520.
Antwoord
gfaman 3 jaar geleden
Yes, typing mistake, 18LF2520. )-: Corrected...
However, I used several time the F version under 3.3v and always working well if clock is low (used 4Mhz). But the datasheet is clear that under 3.3v, it must be the LF version.
Please note also that all these PIC family are close in term of pinout. Recently, I have to use again the 18F2520 and I cannot get it. So, I replaced by a 18F24K22. Sw need be adjust but working.
However, I used several time the F version under 3.3v and always working well if clock is low (used 4Mhz). But the datasheet is clear that under 3.3v, it must be the LF version.
Please note also that all these PIC family are close in term of pinout. Recently, I have to use again the 18F2520 and I cannot get it. So, I replaced by a 18F24K22. Sw need be adjust but working.
Antwoord
Toon meer
1 Opmerking(en)
Updates van de auteur
gfaman 4 jaar geleden
Pollut_sw.zip (502kb)
gfaman 5 jaar geleden
I made an excel file with all the consumption and see how to optimize it. There are still some area which can be optimize, in the software (using IT for button per example) or hardware (lower IQ for regulator). However, the simulation with the excel file shows that the main contributor for the consumption is the particle sensor, which need 30 sec ON before taking measurement. 30 sec with 100mA (due to the fan). this 30sec are in the spec. Reducing it will increase seriously the autonomy.
So, based on this simulation, I didn t spend time to optimze the consumption in SW or HW. However, 2 to 3 weeks is already a good results. No air quality sensor can do that on the market.
gfaman 5 jaar geleden
The pictures showed how small it can be. The withdraw is the battery is smaller, can be around 700mA/h.
Some software improvement are made:
- Launch a measurement at switch on
- Each measurement are now made with 3 consecutive sensor data
- Optimize button management for more fluent in menu moving
gfaman 5 jaar geleden
- Made one support for outdoor fix, with external solar panel and add some recharge from the sun. Fixing it on a window outside, we can see imediatly for air quality only by looking the LED. Very very convenient...
- Adjust the total thickness of the casing to be able to add a 1200mAh battery, giving more than 2 weeks autonomy with 20min measurement period.
- Made a new PCB to correct some mistakes on the first one, and with a graphic display connector possibility (but SW not yet develop)
- I put now 4 devices at home, one living room, one outside, and 2 in the bedroom. I launch the aircleaner with it is bad status.
- Made a simple aircleaner with a portable fan and HEPA filter. It is enough to clean the air of kids's bedroom within 1h.
Wood 3D printer color (1922kb)
gfaman 5 jaar geleden
from inside view (2643kb)
From outside view (2319kb)
Quite thin (1737kb)