SparkFun Forums 

Where electronics enthusiasts find answers.

General suggestions or questions about the SparkFun Electronics website
By rootbeer
I just received the alcohol gas sensor and breakout board. Soldered it together connected H1 to vcc and ground to ground of course. A1 to vcc and A2 to a 100k resister, and the resistor then to ground. Then I connect a wire from A2 to the analog input. This setup is identical to the datasheet recommendation and the example link on the sparkfun page. Btw this is connected to my bit whacker. So i configure a port to analog input with c,1,0,0,1 and then watch it with t,500,1. All i get is the max 1023 with _slight_ changes no matter how much booze i blow at it. What gives? It is also my understanding that it should idle somewhere around 160 then go up and up as it gets more alcohol.
By JamesC
I had the EXACT same problem.

I'm beginning to suspect that the breakout board itself is designed incorrectly. I would have expected A1 and A2 on the breakout board to be wired to the "A" and "B" pins respectively. It appears they are both wired to "A" (or both to "B", but not both)

So, since I had followed the diagram on the whitepaper and put 5v on "A1", and tried to get a reading on A2... I got 5v always because there is continuity between those pads.

Perhaps someone from SparkFun can check this (HINT HINT... ANYONE THERE READING THIS?). I would not have expected continuity between A1 and A2, but I have that after soldering the sensor to the board. Meanwhile two of the pins appear to be electrically unattached to the breakout pins.

Either that or I've done something quite wrong.

In my case, to workaround, I determined which set of A/B pins were NOT wired to the breakout, and jumpered one of them to H1. This sort of works, however, the range I'm getting right at the moment is limited. I'm letting the heater run for a while, and as I do the non-alcohol reading is getting lower and lower (right now 850 out of 1024, or ~4.1 volts). If I take a swig of beer and blow across the sensor, it jumps up to 1000 (~4.8 volts) , then slowly comes back down.

These readings are with with a 100k pulldown on the output.

Why is the range still so limited for me? Not sure. I may in fact still be doing something wrong. The white paper is hard to follow, its not even clear to me how to read the resistance charts on page 2.

Also, I'm wondering if the heater is supposed to be on all the time, or is it just used to "reset" the sensor. Hmmm.. I'll have to unjumper that now to see... drat. Then again, the whitepaper has a line that says "Preheat time" is 24 hour.

Hope this helps. At the very least, investigate the A1-A2 pins on your breakout, if have continuity there and you're not expecting it.... thats probably your problem.
By bradb667

Have you guys gotten anywhere with this?

I'm having the same problem only I'm not using the breakout board, so I suspect the problem is something else (well, I started using the board but after reading this thread - just in case - I soldered directly to the MQ-3).

I've based my wiring off of this blog:

By JamesC
Sorry, nothing new to report. I was hoping someone from Sparkfun would reexamine the breakout board, as I it appears to be incorrectly designed, as I noted above.

Even after working around this, I have no explanation for the reduced voltage range I'm seeing (~ 4.1 to 4.8 volts)... Maybe I have a damaged sensor?

Are you getting any kind of variation between alcohol present/ not present?
By MacFlecknoe
I too am having problems with this sensor... I am trying to read the resistance between the lower A and B pins on the sensor and i am getting a steady readout floating around 200 or so no matter how much alcohol i blow over the sensor. I am using a 100k resistor.

I thought perhaps I damaged the sensor somehow when i was first setting it up as i shorted the a pin to the H pin... maybe i did? I still get continuity between the A and B pins i am testing over however so i definitely did not burn out the filament.

Any ideas? This is getting frustrating as I want this built for our companys holiday party :-/
By MacFlecknoe
I think I solved the problem and am getting good results (after 10 minutes of testing). I scraped the sample diagram and worked from scratch. I created a simple voltage divider and measured one side:

If anyone is interested here is what I did:

I connected VCC to a 16K resister. I then connected that resister to both the analog-in pin of my arduino and the A post of the MQ-3 sensor. Finally, I connected the B post of the MQ-3 sensor to ground. (the h posts are connected to vcc and ground on a separate circuit)

I get readings from 900 down to around 200 where 200 indicates a high percentage of alcohol and 900 represents none. Ill work more to refine it this week; Ill post follow ups if anyone is interested.

As an aside: I am guessing the heat generated by the coil in the sensor acts as a catalyst to the chemical reaction the sensor depends upon to generate resistance. so I presume the H posts need to have current whenever the unit is in use.
By MacFlecknoe
I do have a question however.

Has anyone been able to make sense of the conversion table in the datasheet? I have no idea how to convert resistence to ppm... Does the relationship looks like a hyperboa to anyone else? What is the equation?
By kurf
So in conclusion the breakout board is wired correctly. Sparkfun has just forgotten to include a B1 and B2.
User avatar
By FartingMonkey92
Seems they have missed out one/two of the vital connections on the breakout board... :roll:
They also haven't re-stocked it since... (so i think they could be making a new version)
By tintenfisch
I have the same problem.. did anyone solve this? I think I'm not reading the schematic correct..

(Sparkfun breakout board)
* 5v -> H1 and A1
* GND -> GND
* A2 -> Arduino analog 0 AND A2 -> 200K resistor -> GND

(I've tried an array of different resistors, always the same results)

This gives me a steady 1023 reading..

How sensitive is this sensor? Could it be that I destroy it while soldering? If I hook it up wrong, would that kill it?

I have tried 3 units so far.. think I've killed them all :/
Does anyone have a good schematic or pictures of a working set up?


By filmo
Hey guys,

I'm also having problems with the MQ-3 sensor. I've posted my set up on the link below:

It includes images of the new (?) breakout board which shows A1 and B1 instead of A1/A1.


When set up as detailed on the link (which is very similar to other posts on this thread), I get ranges from 820 to 1023 instead of something closer to 0 to 1023.

Also the sensor is heating up like a mofo, so I'm concerned that if I leave it running for more than a few minutes, it's going to melt??

That said, as it gets 'hotter', the bottom end of the range drops?

Has anybody come up with the definitive answer on how to wire up these gas sensors?
By NickH
MacFlecknoe wrote:As an aside: I am guessing the heat generated by the coil in the sensor acts as a catalyst to the chemical reaction the sensor depends upon to generate resistance. so I presume the H posts need to have current whenever the unit is in use.
I think you're almost right. The dsheet says "The heater provides necessary work conditions for work of sensitive components" ... not very helpful.

Doing a bit of reading [1], it sounds like the tin dioxide sensor looses sensitivity to ethanol if water is adsorbed on its surface. The heater evaporates the water out of the sensor.

You're right that the heater needs current whenever the sensor is being used. The dsheet even mentions a 24 hour preheat time. (!) It makes sense that the sensor would take time to drive out any water and stabilize.

The sensor is not going to melt: the sensor and heater are combined in a ceramic element, and the whole thing is mounted on a resin base. It's meant to be on for long periods of time.

[1] ... 9878301323