Question regarding 5.2v, an Arduino Nano and a MQ135 sensor

Other Hardware Sensors
1

Good evening everyone,

I have a rather stupid question which makes me feel like a newbie…

Scenario:

an Arduino Nano, a i2c LCD and a MQ135 sensor are linked together in order to show the outside Air Quality…

I know that the MQ135 is not a fantastic sensor… but for what I am doing, it would be sufficient.

The sensor has been (more or less) calibrated, — do not ask how mich time I took to do that, and readings “seem” to be valid.

Now…,

when plugged in to the USB port of my desktop, readings are fine and correspond pretty much to what another Air Quality station is spitting out.

(The other Air Quality station is located nearby (2.5 Km) and maintained by a big hospital; hence, they have the means to install quality stuff which works correctly.)

Then… ,

when I power the setup via a mobile phone charger, readings are completely wrong… (they are really too low and stay low)

I tried different phone chargers… output 5V 2W, 5V 10W and even 5V 33W…

The readings are not as expected, even if they were more or less valid a couple of seconds before when powered by the desktop computer, and this, irrelevant with which charger the whole setup is powered.

I have a 5V 120W phone charger, but I have not tried it yet… I fear that it could fry everything.

Here again, I repeat, everything seems to work fine with the desktop computer, but I am not willing to power up the whole computer whenever I want to have a quick look at the Air Quality…

Hence, my question…

What kind of power supply would be ideal for my project…

Do I need a separate power supply to power the MQ135? - have not tried this yet…

Thanks for any hint…

Cheers

Eric

2

Are you sure that your phone charger is giving exact 5V in real life? Did you measure with a multimeter?

3

Show the datasheet.
Are you supplying heater power AND logic power?
These devices sometimes need warm-up time... many hours.

4

Yes…

If I remember correctly…

the 2W gives 5.21v

the 10W gives 5.15v

and the 33W… exactly 5.01V

I have not tested the 120W charger (it is the charger of my wife’s phone… if she discovers that I use her charger to power my things, I am fried… Ha Ha Ha

But if I ask her nicely, she might lend it to me for 24 hours…

5

“Are you supplying heater power AND logic power?”

Yes, everything with the same power supply…

and I think that this is the problem - the sensor needs probably needs a separate P.S..

In that case, how come the PSU in my desktop powers everything as expected?

6

Is this an official Arduino Nano with the USB micro connector (not the USB-C) ?

7

I think it is a Chinese made NANO…

But it works fine (I had it 24/24 / 6 month working on another project which is now dismanteled)

the connection is …

Mini-B USB

image
image260×148 13.3 KB

image
image1800×1200 158 KB

8

Yes, I meant Mini.
OK well that rules out USB-C problems with some of the clones.

9

Oh, so it is a clone.
Tried pushing the reset button ?

10

Which Arduino?
How are you connecting the phone chargers to the Arduino?

11

Those chargers are switch mode power supplies and cheap ones may introduce considerable high frequency ripple on the output voltage.

12

VCC of a classic Nano on USB is about 4.7volt.
In theory too low for an MQ135 that needs 5V±0.1V

13

Maybe you should move to a more modern digital sensor that works on a lower voltage, draws less current and has a shorter warm up time.
I'm about to test the Sensirion SGP41.
Leo..

14

AHA !!!

So that might be the problem.

While the Computer PSU certainly gives a 5.xx volt, it could be that the voltage regulation system on the Arduino Nano reduces the available current to 4.8 volt or 4.9 volt.

And … the phone chargers, while pushing out a 5.xx volt (checked with a multimeter), their real output might drop under “heavy” load (33 ohm heater in the sniffer module).

Hence, I have to find a solution to bypass the Arduino power regulator.

image
image239×211 7.23 KB

maybe this is the solution (picture from the internet)

OR…

Power the module with its own PSU.

Now, I can see another problem…

How to ensure that the 2 PSU’s do not create a shortcut situation?

Means: 1 PSU on the Arduino and another PSU on the module, both linked together via a “data” cable from the module’s Analog “pin out” to PIN A(x) of the Arduino… (in my case: Pin A1)

Certainly, somebody out there has done this already … any experience to share?

mq1350
mq13501198×750 130 KB

mq1351
mq13511198×980 176 KB

mq1352
mq13521198×980 161 KB

15

To explain further,
When powering a classic Nano through the USB socket, no regulator is involved.
USB power passes through a simple backflow protection diode, with a drop of about 0.3volt.
VCC (MQ135 power) is "USB voltage - diode drop".
USB voltage, and thus MQ calibration could be all over the place.
It seems you have calibrated the sensor for average PC/laptop voltage.

Things get better when you power the Nano externally (9volt recommended).
Only then will VCC be a stable 5volt. But you need to re-calibrate for that voltage.
Digital sensors don't have these problems.
Leo..

16

Oh, I see !!!

So best is to power the whole thing with a 9v battery…

By the way, I just tried this configuration (picture below)

image
image1198×980 176 KB

Readings go all over the place. No consistency… from -25000’ish to +25000’ish. - any number, just like a random generator.

And

I also tried to combine all the “GND” together while keeping 2 PSU…

Readings flat zero. Nothing at all.

Hence, I revert to my old setup and get a 9V battery at the next 7/11 later during the day.

17

Yes, you can power the sensor from PSU 2, but there also must be a common ground from the sensor to the Arduino, which is currently missing.

If you mean a 9volt smoke alarm battery, then forget about it.
These sensors draw 150mA...

18

Something is not right. It’s hard to believe that all the phone chargers you have are bad.
You can do some further debugging with the nano to see if the chargers are actually the problem. Let me know

19

Working on it…

and I have to agree, it is puzzling…

Will keep posting.

20

Yep, indeed, I checked the factory description for that module and concluded that a 9V battery would not last long… (especially because this module needs a long warming up time to spit out… some more or less correct data)…

Hence, I did not get a 9v battery…

And I tried this version too… (as below)

Readings were flat zero.

image
image611×500 69 KB