MH-Z19B FALSE CO2 Sensors

Last modified February 1, 2021

We are seeing, in recent times, a growing number of MH-Z19B sensors more fake than a 3 euro coin. It's about imitations low quality, which are being sold both on AliExpress and on Amazon and other businesses.

In this article I'm going to try to shed some light on these fake sensors. Some tips for prevent your purchase, or identify if the sensor we received is fake (or copy).

HELP ME HELPING TO HELP!

Right now there are many Makers all over the world creating CO2 meters based on Arduino, ESP8266, ESP32 and the like, to try to help reduce the transmission of COVID through the air, especially in colleges, schools, institutes and universities.

These Makers do not suspect that the MH-Z19B CO2 sensors they are using (this is by far the most popular model among hobbyists) give some readings that are worthless. It is important that everyone knows.

I consider the problem of the fake MH-Z19B sensors to be important enough to ask for your help.

Please share this page to spread this information.

You just have to click on these buttons:

What's the downside of the sensor being fake?

The biggest problem with receiving a fake sensor is not only the fact, already unpleasant enough that they sell us one thing for another, but that we can't trust nothing of that sensor.

The underlying problem is that Most these fake sensors measure, measure. And since the sensor gives measurements, and most people only have one sensor, they can't compare your measurements with others, and in this way know if you are measuring correctly.

Luckily or unfortunately (I say "luck" because it allowed me to write this article), I received one of these fake sensors a few days ago and since I have several CO2 sensors, it has allowed me to do some comparisons that I will share with you in this article.

Of course yes a normal user receives your sensor, connects it and sees a CO2 concentration of say 1500 ppm, thinks (logically) that your sensor measures correctly. But yes, as is my case, when I turn it on and see 1500 ppm I have two other sensors next to me, I trust, giving a measure of 2600 ppm the thing changes.

In addition to this, many people think that a measurement of 1500 ppm or 2000 ppm is already very high, and this It is not like this. Where I work in my house, which is a normal room, of normal size, as it could be in any house, it's nothing weird reach concentrations of 2500 and 3000 ppm if I spend many hours in a row with the door and window closed (even if I am alone).

In the following graph you can compare the measurements of three sensors: The two above are a MH-Z19B original (in orange) and a Senseair S8 LP (in green). The line below (in blue), with much lower measurements is a Fake MH-Z19B (with black plate). All of them have been calibrated by me at the same time, outdoors, after having stayed there for an hour to stabilize them.

Comparison MH-Z19B original vs Senseair S8 LP vs MH-Z19B fake (copy)

AS YOU SEE, IF THEY WERE USING THIS SENSOR IN A CLASSROOM, THEY WOULD THINK THEY HAVE ACCEPTABLE CO2 LEVELS, AND THEY WOULD NOT VENTILATE ENOUGHWHEN THE REALITY IS THAT THEY WOULD BE VERY ABOVE THOSE VALUES

Please share this page to spread this information.

You just have to click on these buttons:

How can I tell if the sensor I have is fake?

Although there is nothing verified at 100%, and I can be wrong, I am going to risk telling you how can you tell if your sensor is fake or not quite safely.

Fake MH-Z19B (copy, black plate)

Summarized it would look like this (until I see tests of other sensors with my eyes):

Rule 1: If your sensor board is black and your sensor has the smooth box, your sensor is fake.

Rule 2: In all other cases, rule 1 applies.

I'm going to order this list more certain that it is false unless:

MH-Z19B sensor with black printed circuit board

Everything I know so far makes me think that you have a fake sensor with a security of 99.999%.

All the sensors with a black plate that I have seen so far (or of which I have received a reference) have the box completely smooth, without the raised contours that the original sensors have.

They also have a box of a slightly more color «yellowish»(Although it is difficult to know which one it is if you don't have another to compare).

In short, do not complicate yourself, in my experience, sí has the black plate is false.

IMPORTANT:

If you have received one of these sensors, be careful before connecting because, from what I know so far, everyone come with the wrong cable and if you connect it relying on the cables (red to positive, black to negative) it will probably break (apart from not working, of course). In the video you can see how the cable should be, so you can connect it without risk of spoiling it.

MH-Z19B sensor with completely smooth housing

Fake MH-Z19B (copy, plain box)

From what I know so far, MH-Z19B sensors with smooth box (without the raised contours that the original sensors have) they are false.

As with other variations. I have only been able to verify sensors with a completely smooth box with a black plate.

If you have a sensor with the smooth box and green platePlease send me a photograph to document it.

Sensor MH-Z19C

Original MH-Z19C (textured case)

In the case of the MH-Z19C, which are still quite new, it seems that they are original.

The pictures officers from the manufacturer's website show a sensor with a textured box, with some curves, but in the photographs of some AliExpress sellers, the box is smooth (although those curves are maintained).

Of course, this sensor has the problem that is very sensitive with food And it can be problematic if you don't have a measuring instrument to make sure you're feeding it correctly. You can find more information about this problem here.

At the moment, the MH-Z19C sensor I do not recommend your purchase (at least for the project of the «Homemade CO2 Meter with Wifi»Of this blog), unless you have a multimeter and you understand the problem and how to solve it, as indicated here.

Other sensors MH-Z19B

Although in the photos of AliExpress sellers there is a whole collection of varieties and variations, the reality is that I have only been able to verify the existence of the above.

I will be completing this article with new variants that appear.

If you have a different sensor, tell us about it in the comments.

How do I avoid buying a fake sensor?

Although it is not foolproof, at the 100%, there are some tips that can help you reduce the chance of receiving a fake sensor.

  • If the price is too good, the sensor is probably not that good. Right now, a sensor of this type should cost you between 18 and 22 euros (they have risen a bit lately, I suppose as a result of the high demand).
  • You will have more chances that the sensor is original if you see that customers have put pictures, when making their evaluations, and in those pictures you can see that they have the green plate and reliefs on the box.
  • See that the ratings are good and that customers do not say that the sensor is false (not always all ratings can be five stars, and it is normal). If they say that the sensor is false, better if they accompany it with a photograph (there are people who have had problems with their project and blame it on the fact that the sensor was false or was damaged, and it does not have to be that).
  • Make sure that the seller is of a certain age, and that their overall assessment is good. I usually assume that an overall rating higher than the 97% is good enough to give me confidence.

How do I return an MH-Z19B sensor purchased from AliExpress?

If you have received a fake sensor, you will have to open a dispute on AliExpress to get your money back.

There are two reasons why you can request a refund:

  • The sensor is false
  • The sensor it does not work correctly

My suggestion, because it has worked well for me, is that you select as a reason that the sensor does not work correctly, but in the description make it very clear that, Besides, the sensor is not working well.

Depending on your particular case, you will have to use your own weapons. I used the comparison with two other sensors that I had previously.

The message that I have used to open the dispute is the following:

«Measurements are unusable This is reporting a CO2 concentration 1266 PPM. Next to it a real MH-Z19B and a Senseair S8 LP are reading 2419 PPM and 2389 PPM respectively. Attached photos of received product (black PCB, RoHS is missing, Pins are labeled differently and a different physical connection to the one offered). Attached is a video with the full unboxing and explanation. More information and explanation here: https://revspace.nl/MH-Z19B#Fake_MH-Z19B_.28black_PCB.29 »

Along with the text, I have sent the following photographs of the received sensor:

In addition, I have sent a video with the unpacking of the product, in which it is clearly seen shipping label (this is very important for AliExpress, here you will see it with that part blurred for privacy) and the explanation of the comparison.

In less than 24 hours had response from AliExpress, giving me the reason and telling me that They made me refund the money on the same card with which I had paid it and that I didn't need to send the product back.

But beware, they didn't give me the reason why the product was fake (which was what I had claimed), but because the product it didn't work properly.

This means that, if you can't prove the sensor malfunction (like me, who had two other sensors to compare with and the means to document it in a video) maybe it's not so simple that they agree with you. This is something you will have to assess in order to guide your dispute in one direction or another.

The video I have posted is, basically, practically same as the video that I leave you below, without the central explanation of what a fake sensor is, and very summarized (1:30 of unboxing and 3:45 of measurements and comparisons and with a basic edition, much simpler). I also took the time to subtitle all video in English.

How do I return an MH-Z19B sensor purchased from Amazon?

Yes, buying from Amazon is more expensive than doing it in China directly, but, in addition to usually you receive it much earlier, has the great advantage that returning a product is extremely easy.

In this case, as I see it, you have three options:

  1. Return it because if, without further explanation
  2. Return it because it is false
  3. Return it Because it does not work

The difference between the first option and the other two is that, in the first case, you have to pay the return costs, while in the other two cases the return will be free for you.

On amazon you can return (You pay the return shipping costs) any item purchased for a period of 30 days, unless otherwise stated.

In cases where the return is due to the product not work, not as advertised, etc. you do not have to pay the return costs.

This means that it is better to try to return the product claiming that it is false and does not work well and, only in the event that they do not give you the reason, make the return because if, even if you have to pay the shipping costs.

The return shipping costs, in cases where you have to pay them yourself, are € 2.99 if you deliver it in a Post Office or Celeritas or € 5.50 if you request home collection.

In any case, If you bought the sensor on Amazon, you are guaranteed a return.

Fake MH-Z19B sensor in video, full explanation, tests and data

In this video you can find additional information about the fake MH-Z19B sensors with explanation, measurements, graphs, comparisons, etc.

HELP ME HELPING TO HELP!

I consider the problem of the fake MH-Z19B sensors to be important enough to ask for your help.

Please share this page to spread this information.

You just have to click on these buttons:

5 thoughts on “Sensores de CO2 MH-Z19B FALSOS”

  1. Hi, as I mentioned in the original post, AliExpress has given me a 19C (given away because the uncle canceled false shipments so many times that when he finally sent it, I was able to win the dispute over the last cancellation).

    What I'm going to. It looks exactly like the one in winsen's photo, with the ribbed top and only one diffusion window. Measure measures, but I have no idea if it is accurate or not. I have mounted it in the living room and it actually rises when we are, if you blow it it hits a peak of the milk. But then if we go down alone. And in general the measurement varies a lot. I have attributed it to the fact that we have a chimney and how the Monse draft can close, because clean air enters there continuously. But I don't know.

    Is there a way to indirectly measure precision?

    The temperature gives more or less good, a degree at most different from the fetén sensor I have in the living room.

    The only electronic difference is that I have to feed it with just 5v, not with the esp32. Otherwise…

    Here are the sensor measurements, in case you want to take a look:
    https://thingspeak.com/channels/1279169/charts/1?bgcolor=%23ffffff&color=%23d62020&dynamic=true&results=300&title=Co2+PPM&type=line

    I am waiting for the screen to mount it in a box. I want it to manage the opening of the window in the job. That I am dying of cold. Then maybe I'll buy another sensor. And I build one for home.

    Reply
    • Hello Jesus.

      That looks very good.

      In my house the measurements would be very low for the living room, with several people in it, but they may be correct in yours, especially taking into account what you say that the chimney draft is always open.

      These are the CO2 values of my living room on the day with the lowest measurements that I have found in the last seven days:

      I am looking at some homemade ways to measure precision (rather than to measure the precision of if you measure, for example, with a precision 5%, it would be to know if you are measuring 1000ppm when it should be 1200 or 1300ppm (relatively large deviations). When I do tests, and if it works, I will prepare a video and instructions so that everyone can do it. The idea is, above all, to be able to rule out sensors that are not measuring well, rather than check the accuracy of the sensor.

      I hope that in not too long the testing of our sensors will be easier. When the sensors are more widespread in bars, restaurants, waiting rooms and others (they say they are already beginning to be seen) it will be a matter of taking our sensor to the bar with a powerbank, have a beer and check 😉.

      Reply
      • Mw like that about beer…. Well, let's see what you can think of.

        I just put it in a smaller closed room and it seems to go up…. Although sometimes it seems that it gives him the hiccups. Of course, my house has quite a few "leaks" and is poorly insulated.

        I'm also going to walk you around to see how to check it. I read someone who took CO2 out of a soda and then injected it into a known volume ... Let's see ... With a closed box ... Filled with clean external air ...

        Reply
  2. Hi.
    I have the 10000ppm. They have a sticker on the side and it says so.
    It has all the relief and looks original. He wears the green badge.
    In the Datasheet v1.5 they already include the three models: 2 thousand, 5 thousand and 10 thousand ppm.
    I'm running it with an Rpi and python3 in staging mode and haven't been graphing for so long or testing with values so high that it doesn't seem to work well. I think it is the original.
    The temperature is correct compared to a BME280.
    Thanks for the article.
    a greeting

    Reply
    • Hi Lluis.

      Thank you very much for confirming.

      Indeed, in the last datasheet (the one linked in the article "The MH-Z19B Bible") it already appears. Now I update the article.

      Regards, and thank you very much.

      Reply

Leave a Comment