How a breathalyzer works (alcohol sensor)

I teardown a commercial, handheld breathalyzer and discuss how it works, then test a standalone ethanol sensor.
www.seeedstudio.com/depot/images/product/MQ303A.pdf
www.kosmodrom.com.ua/pdf/MQ303A.pdf
www.sparkfun.com/products/8880
www-personal.umich.edu/~mswool/publications/Gas_Sensors_preprints.pd.pdf

[@WillowEpp]

The casual delivery at the beginning really makes it, for me. "Someone gave me a thing, so I'm going to reverse-engineer it and make my own." No hesitation or uncertainty; just the confidence of a guy who does this kind of stuff all the time.

[@piranha031091]

At 5:39 , you say "this isn't pulling oxygen out of the lattice". I don't see why it wouldn't. It is known as the Mars Van Krevelen mechanism, and is a well known catalytic process that occurs on metal oxides, including tin oxide. (See M. Batzill, U. Diebold / Progress in Surface Science 79 (2005) 47-154. It is mentionned in both the introduction and part 2.2). What happens is not that you reduce the oxide to tin metal, you just introduce defects in the lattice, where the tin is partially reduced. In doing so, you change the doping of the tin oxide (making it n-doped), explaining the change in its electrical properties. Those lattice defects are then filled by oxygen from the air, regenerating the metal oxide. I guess it is safe to assume that the concentration of those defects depends directly on the oxygen/ethanol ratio in the gas been fed to the analyser, so that is why you can deduce how much ethanol is present in someone's breath this way. (No, I'm not just pulling that out of Wikipedia, I've been working for two years now on titanium dioxide, which can also take part in this kind of reaction.)

[@AppliedScience]

Good information! Thanks!

[@piranha031091]

Applied Science By the way, I was wondering, do you have access to scientific publications online? Or do you have to pay each time you want to download a paper? (Oh, and why did you change your channel's name?)

[@WeAreGRID]

piranha031091 he finally did that google plus thing and decided that his channel should be different from his personal G+ page, same as mine is different from my G+ page, but run off the same google account (i have no idea why google makes us use more than one account for a google account for a google service that runs on a google account login for a different google service though)

[@piranha031091]

Applied Science Hey, If you're still working on Tin dioxide sensors, you may want to take a look at that document I just stumbled on, it may be useful to you: Consumption measurements on SnO2 sensors in low and normal oxygen concentration tobias-lib.uni-tuebingen.de/volltexte/2004/1201/

[@BrettW]

Implement this into your cookie machine, the more drunk you are the more cookies it makes.

[@WeAreGRID]

No, include a sensor that detects how STONED you are, therefore making you more cookies :P

[@warywolfen]

Wow, breathalyzer technology sure has advanced in recent years. The traditional kind of device employed a tube filled with potassium dichromate. The dichromate ion has an intense orange color. It's an oxidizer. It oxidizes the ethanol in the air passing through it, and is reduced to chromate, which is green, in the process, The BAC was determined by the depth of the color change. Of course, a fresh tube had to be used for each test.

[@isaackalashnikov3681]

I have had a test tube full of potassium dichromate laying around for about two years and just now i am learning that

[@jamesfenton7338]

First video I watched was about your lathe, but now I am hooked. I thought I was an applied scientist, but in reality I am just a student. The wealth of knowledge contained in your videos is invaluable. If school was this much fun, I would have remained in class.

[@RimstarOrg]

Interesting. I wonder if you can test the alcohol content of a drink by vaporizing some of it and sending it through this device.

[@aserta]

I think it would be over saturated. You have to consider that there's a lot of other things besides alcohol in a breath.

[@RyuSujin]

That would be a very cool experiment to try!

[@thatguyb3rt]

While a cool concept, I don't think it'd work due to the way alcohol evaporates at a lower temp, Thus the process of distillation to make higher strength alcohol.

[@thatguyb3rt]

***** Huh?

[@ThingEngineer]

Besides, a hydrometer is already used to determine that by measuring the specific gravity of liquids. Plus the vapor would be missing the O2 component.

[@askassk]

Interesting video! I've got an idea for a little experiment you could do: How about you hold your breath as long as you can before you blow into the breathalyzer? That should decrease the amount of oxygen in the exhaled air, which then may lead to a lower reading. I'm quite curious if that would work and I would really appreciate an answer!

[@sciencoking]

8:10 For science!

[@tonyellen_]

LOL I'm 5 years too late to comment the same thing.

[@RealationGames]

Hah, for the calibration they have a specific guy in the factory being wasted all the time to give reference values. Probably the best job in Asia.

[@jskratnyarlathotep8411]

asians are very bad at metabolizing alcohol. How it is called when you have headache after the drink?

[@alfrednzala7988]

Lol

[@Ma_X64]

@@jskratnyarlathotep8411 In Russian it calls pokhmel'ye)

[@ufi9540]

@@Ma_X64 Нет. У них не симптомы похмелья а именно интоксикация.

[@Ma_X64]

@@ufi9540 ты разберись, дружок, что такое симптом, и что такое синдром и не путай впредь. И учебник логики ещё прочти.

[@EMandMORE]

The videos are getting more and more awesome. Thank you for all the great job. I love when you analyze circuits so much in-depth. It makes really understand the whole circuit. Thank you

[@HeilmanHackatronics]

Quite interesting, Chemical Sensors are fascinating.

[@ЛукаХренов]

☺!

[@Sedokun]

"It's just happens to be that You don't exhale Butane or Hydrogen..."

[@comkaosstime]

I've only seen two of your videos but really great clear concise very well paced and organized okay the moral of the story here is to suck on these things and not blow in to them.......... 1 ounce is not a true test.... a true test is test to destruction.... looking forward to a follow up video on ethanol production.... thanks for the vid

[@bormisha]

A lot of catalytic gas sensors use temperature regulation. While supplying power to the heating element, they also measure its resistance and thus determine its temperature by thermoresistive effect. The regulator is programmed so as to keep the resistance constant. The power necessary to maintain constant resistance is a measure of medium flow and/or catalytic reaction. Methods exist to distinguish one influence from the other.

[@MissionFitnessCTC]

Really impressed with the production quality of the videos. Thank you for taking the time to make these videos clear and concise. So much garbage out there

[@ttkoh123]

I am always fascinated with his detailed explanation. thanks;

[@UberAlphaSirus]

Actually, I could have really done with one of these today! I have spent around 3 Hours removing vynle decals from my van ready for the newer updated ones. I used methylated spirits (denatured ethanol) to remove the glue residue from the paintwork, loads of it. My hand absolutey stank for 2 hours after and I wondered just how much went into my blood stream. I thought this would be tough to explain if I was pulled over. I also had the same problem when I done this last year then had to go to a parent evening at school, I thought I should explain I haven't downed a bottle of vodka before they made presumptions. BTW vodka is great for removing Biro ink, I used to be in the soft furnishing trade for 20 years, so if you see a bottle in and upholsters workshop, you can still trust him ;P

[@iirelu]

"So to test it out I drank exactly one and a half ounces of vodka 20 minutes ago." SCIENCE!

[@peppeddu]

Kinda remind me of the documentary about the Japanese sex dolls factory where the boss asked some of the employees to take one home and see what kind of improvements could be made.

[@jonevans2687]

I think you description of the sensing mechanism is correct. SnO2 is naturally n-type due to oxygen vacancies acting as shallow donors. Surface oxygen removes electrons from the conduction band creating a depletion region at the film surface. Removing the surface oxygen by reacting it with the test gas will then result in a measured decrease in resistance.

[@DrSomeGuy]

How significantly would it effect the resistance if you were to hold your breath for about a minute to exchange most of the oxygen in your lungs with CO2 before blowing on the sensor?

[@hellterminator]

You can't use up most of the oxygen you breath in. You'll pass out from CO2 buildup way before you'll have used up even half the oxygen (assuming normal atmosphere with ~20% oxygen).

[@LEXTHE4]

hellterminator yeah but even if you could reduce the oxygen concentration to below 15% id imagine it would still have some effect.

[@markt3287]

Alex Boehm ya it makes ur b.a.c. higher cuz u have less oxygen coming out

[@ChristmasEve777]

Mark, it's the other way around? So the trick is to inhale a pure oxygen balloon just before blowing? Unfortunately, the cop doesn't make you blow right away. He'll have you get out of the car and walk the line, etc, first. So you can't just suck in oxygen while you're still in your car. Oh well....

[@stargazer7644]

@@ChristmasEve777 there’s a sure fire way to get around this test. Stop driving drunk before you kill someone.

[@AsymptoteInverse]

I recently bought one of the larger-model BACtrack breathalyzers. My knowledge of electronics and engineering is somewhat limited, but of interest, there were some components inside whose purpose I wasn't certain of. There was something that looked like an electret microphone, which I'm guessing, based on this video, is the alcohol sensor. But there's also a solenoid controlling a silicone bellows which draws a small amount of air into a rectangular device also covered in silicone.

[@GeorgeGraves]

Nice! Also, the consumer BAC devices like you tested will only display a value up to a preset limit. I assume it's to keep people from seeing how high of a BAC they can blow. That of course, take all the fun out of it for a drinking game.

[@swsephy]

One of my old roommates bought one but I think he got screwed on the deal because it worked for one party and then shit itself permanently. Sad times.

[@USWaterRockets]

Could you quickly cool the sensor and see if it thinks it is being blown into? That would confirm your theory about the constant current sense for air being blown in.

[@laurdy]

it would be interesting to try different products such as breath freshener or different types of mouthwash to see how they affect it

[@RODALCO2007]

Great video, thanks for posting.

[@rkankare]

the sacrifices you do for science

[@alexandersteen6533]

nice look inside! these sensors can also be used for controlling butane powered spudguns by the way, these sensors surprisingly survive beeing right in the explosion chamber, or even more interesting also, for measuring farts!

[@3D_Printing]

8:08 I was waiting for the Booze Up :) Merry Christmas

[@abdalghanikhamis5276]

Great work! Believe it or not, your work is actually useful even after nearly 7 years. I just have a question, hopefully, you have the answer after this long time, Where did you get the converting table from?

[@Brickcaster]

In the final test you blow directly on the sensor. The resistance likely drops considerably due to the cool your breath provides. That's why the keychain device has the breath moving across. Try blowing directly over it when you haven't had anything to drink and it will likely drop in the reading. Presumably the mesh helps prevent turbulence and further rate-limits incoming air.

[@jerryg50]

Excellent explanation! Very well presented!

[@breadfan262]

Seems like a fun positive control.

[@ligius3]

You can overload it by gargling a mouthful of high-concentration alcohol just before breathing on/in it. I've noticed that in my tests the softer you breathe the smaller the number. This seems to be the contrary of what I have expected with the O2 concentration.

[@ScienceSeance]

Thank you dude. Great explanation.

[@stoneslice]

Outstanding video. Interesting stuff. Thanks Ben

[@andyspark5192]

Some people, who are on ketogenic diet, can measure their ketone level with that. There only problem is, the display isn't adjust to it. A company, which produce 'Ketonix', the ketone analyzer, possibly uses the same sensor. Do you have any sources on that ?

[@FizzlNet]

"And then later I build my own" And that's what takes Ben's videos to whole another level of unfair :D

[@NinjaOnANinja]

8:11 Now that is a job! I drink in the name of science! Oh yeah!

[@wyattsheffield6130]

Fascinating as always!

[@toolhog10]

Very interesting. I always wondered how those devices measure the concentrations.

[@ISmellBurning]

Thanks Ben.... Interesting stuff. "So to test it out I drank....." haha, of course!

[@giageorge2549]

what model is that? Is it a fuel cell type? I'm using an alcosense stealth , a fuel cell type sensor and works great :)

[@xanokothe]

Hey Ben! Could you make a animation about the alcohol interacting with the sensor? Thanks!

[@toolthoughts]

very interesting, you really looked into this!

[@katid9941]

love watching the videos ben! not that i ever understand anything in them (with the exception of the cookie machine, i know all about chocolate chip cookies) still they are very interesting keep up the good work!

[@Sqwince23]

Does it know the difference between blowing into the tube and sucking air out of it? I would think both would cool the sensor. It would be interesting to know if it could tell the difference between the two though.

[@tokeeptrackofrandomsubs5899]

Based on what I saw on this one, probably not but this is most likely not how a police one is constructed. They probably have some way of determining the airflow direction and perhaps even keeps track of the volume. Wouldn't be surprised if they'd also take the extra factors such as humidity into account. Also in my country these tests are only the indicators, if you don't pass them then a more accurate test is done at the station.

[@0x4C4A]

Thank you very much for the informative video!

[@markbell9742]

What change in resistance (if any) do you get from the big sensor if you blow an ETOH free breath. May be try some different gases. Cheers, Mark

[@sharadthakur7]

plzz help me with a thing that do we need to remove the power supply after pre heating it or reduce the voltage to 0.9 from 2.2 volts? and do we need to apply this pre heating voltage before every blow of alcoholic air??

[@christheother9088]

Wondering if this sensor, when inserted into the airspace above bottled alcohol, could be used to estimate alcohol content of the liquid.

[@BenNBuilds]

What sort of effect does water vapor from your breath play? maybe drying out your mouth or blowing through an absorbent material would change the reading? I'm sure there is a large number of sketchy people wondering how to beat these things, haha

[@superdau]

Your lungs have a few liters of air in them and you blow most of it through the breathalizer, so whatever is in your mouth is negligible. That's why all those "tips" how to fake a test (take a mint, lick copper, whatever) are just stupid. EDIT: just checked. Residual alcohol in your mouth can show higher readings if the test is taken in the minutes after drinking.

[@hackeritalics]

They sell canned oxygen in outdoors stores and I'd be really interested to know what effect that would have. You know taking a big hit of oxygen before using a breathalyzer .

[@therealmaximus]

YES! new krawsnow video!

[@HuMaHuMHuM]

My favorite part wast the overlook on the sercet bord. I've alway whonder what all the components rey do with the current, and why that's important.

[@AsymptoteInverse]

Correct me if I'm wrong, but don't they use tin dioxide sensors to detect carbon monoxide as well?

[@rosssmith3072]

Love your videos.

[@UberAlphaSirus]

I don't condone drunk driving whatsoever, however I have driven after a drink within the law. I saw these bac testers in maplin for around £5 and I almost bought one, just to check if I am OK to drive in the morning after a night out, something a lot of people don't consider. I wouldn't trust one to honest, without doing one of my own cal tests on it, say 2 beers after an hour, but then there are more variables to. I have been breathalyzed maybe 5-6 times, picking friends up from a night out, or having a night out myself but only had 2 beers or non and in the morning on police stings to catch people driving to work after a night out. Obviously I have always passed within the limit or no alcohol reading. I have always been asked how many, what and when I have had a drink and if I have smoked and when. they wait 20 mins if you have smoked. I find it quite interesting that there are so many variables to consider for the mobile unit they use, but I have seen on the tv cop shows that if you are over they have some massive piece of kit to blow into at the station. That must have atmospheric pressure, oxygen sensors and all sort of top spec stuff to be spot on so as to prosecute with firm evidence. I also heard that sucking a copper coin might work in your favour, I ponder what the chemical reaction might be in that.

[@jackthmp]

is this the same kind of mechanism used in all or most BAC measuring devices?

[@robmckennie4203]

if the sensors are quite sensitive to the level of oxygen in the air, does that mean you could beat a breathalyzer by taking a long full of oxygen depleted air? or a gas containing no oxygen at all?

[@robmckennie4203]

lung* full damnit

[@Shambler101]

if you feel the need to cheat a bretho then you probably deserve the punishment

[@robmckennie4203]

Shambler101 I don't feel that need, I don't drink, I'm just wondering

[@bpark10001]

Regarding the tin oxide sensor, here's the explanation I got from Figaro. Tin oxide is a semiconductor, like silicon. Its conductivity is influenced by the N -or P type doping it has. It is doped N-type, so normally has low resistance due to the high concentration of free electrons. When heated in presence of oxygen, the heat lowers the barrier between the O2 gas and the SnO2 allowing electron exchange. The oxygen steals electrons to fill its 2 empty spots in its shell from the N-type SnO2, raising its resistance. Alcohol, hydrogen, and hydrocarbon fuels donate electrons, lowering resistance. The older sensors needed to run for days hot before stabilizing at high resistance. With a strong dose of alcohol/hydrogen/fuels, it required hours to regain high resistance. Obviously, the newer sensors have solved the lag problem. The oxygen is required to get the sensor in its "zero" state.

[@markhodgson2348]

Very interesting technology

[@joeneilson]

You know why it knows if there is breath coming in. your breath changes the resistance in the tin dioxide a little bit, (because of the decreased temp and also because of the change in oxygen) just enough to give a ,00 reading, If it does not detect any change it gives the FL reading. You were drinking though I suppose.

[@Lorecastapendragon]

thanks I found that interesting, saw a commercial version of your cookie machine not long ago, don't know whether it became popular because of you or not but I hope someone didn't rip your idea >.

[@markm1514]

"...and then, later, I'll build my own breathalyzer." For, you know, reasons.

[@ElectronFunCom]

I wonder how accurate this breathalyzer is.

[@AZOffRoadster]

+/- 5 DUIs.

[@checkdown3774]

Do a video on an Interlock ignition device, please.

[@elverman]

What's your best guess on how to get an invalid reading on one of these? Either too low or too high.

[@TheIdeanator]

What effect does increased O2 have on the system? Logically it would seem that, if the sensor works by loss of absorbed(adsorbed?) O2, an increased O2 avilibility would reduce the effect the alcohol has on the sensor by either replenishing the sensor quicker or just reacting the atmospheric O2.

[@superdau]

Yeah, I wondered about that, too. With all the "how to beat a breathalizer" "tips", that might be one of the few things that could change a result. I don't know how long it takes the oxygen to be exchanged with CO2 in the lungs after inhaling, but it has to be less than a second. Otherwise breathing heavily would be useless. On the other hand you have to blow for a few seconds anyway, so taking a deep fresh breath before a test doesn't matter anyway.

[@TignMeg]

Increased O2 causes higher readings. See en.wikipedia.org/wiki/Breathalyzer I guess that begs the question does lower O2 give lower readings?

[@markbell9742]

You might need to be careful not to overload the detector element as the SnO2 is an oxidation catalysis and could overheat with a big load. Cheers, Mark

[@gingerfeest]

I would like to see a follow up to test your hypothesis. Maybe breath in and out of a paper bag for a few minutes to lower your oxygen level and see if it reads a lower bac.

[@omegasalmonfish]

Exhaled hydrogen is used for some medical tests, so maybe they sell the same sensor for that too.

[@HUsoldier171]

Most of what is in your breathe wouldnt be alcohol. You dont inhale alcoholic drinks its acetalhyde the oxidation product of ethanol after it goes through your liver thats why straight ethanol on your breathe would give an off reading jn these things ethanol is a bigger molecule requiring more oxygen to react so smaller amounts make a bigger change. Ive never seen a sensor like that then again never drank and never taalk to cops but the ones ive seen have a potassium dichromate solution which reacts with the acetalhyde turning it into a green potassium chromate solution and they are either calibrated to turn green only if your above the .08 limit or they have a sensor able to tell the concentration of chromate and dichromate and tell you what amount of acetalhyde caused that amount of reaction. But otherwise really interesting video hope i provided a bit of better insight to help even a tiny part of your research into these things.

[@axelmermans3067]

If i were to hold my breath for some time before blowing into the machine, would there be less oxygen in my breath reacting with the ethanol and therefore the machine would read a lower alcohol percentage?

[@burnhamaj]

I was wondering the same thing. It seems like that would be an easy way to cheat the system.

[@misiek97012387]

"You don't exhale Butane or Hydrogen"- actually you can ;) I had a patient, who was breathing in butane to get intoxicated. He sad it was alkohol-like state for few minutes. He was brawling in liquor store, arrested by police, and they performed breathayizer test. Result was 3.3mg/l ~ 7 per mil. He was sober, talking normally, denying drinking any alcohol. It was few hours after breathing in butane. Also detection of methane and hydrogen in breath is used to diagnose SIBO- presence and activity of specyfic bacteria in bowel. Patients drinks lactulose, and you take measurements every 15 min. Bacteria break down this sugar, as a product of this reaction comes hydrogen or methane (depends on type of bacteria).

[@AdamBechtol]

If i were making this vid on my channel, I'd have done the whole thing drunk. haha. Interesting video, thanks

[@pong301275cmu]

Hi, How can I get the document that you show in video?

[@johansenjuwp]

Just plain breath will cause a change because of the humidity change caused by your breath.

[@goskascience]

Where did you get the paper/graph shown at 6:47 ?

[@s0nnyburnett]

Try to explain this a jury, about accuracy at varying humidity and temperature.

[@tokeeptrackofrandomsubs5899]

That's why these devices in my country are used as indicators and if it indicates it a blood sample test is taken at the station (or some kind of other more accurate testing method). I wouldn't be surprised if the same situation applies there, they usually have to separate the drunk driver from the vehicle anyway. I'd imagine taking it to the station is as good a place to store it as anywhere else.

[@RonJohn63]

+s0nnyburnett This probably isn't the same model that the police use. But if it's the same technology... well, clever lawyers figured out long ago that they could challenge the calibration of radar guns, so why not also breathalyzers? Of course, the police chief (should) know that, and ensure that they are (1) calibrated properly, and (2) paperwork retained showing the calibration.

[@dougfairweather]

Nice one, Ben. All in the name of Science, eh?

[@falcon02012]

what model of multimeter are you using?

[@Thomcat]

I was really hoping that it showed you'd gotten drunk for the video in the beggining :P

[@wa4aos]

Hello Ben, Thanks for another well done video! Do you know anything about the life specs of the sensor? It would be interesting to see if an RF source at some frequency TBD could influence the sensor to force it to read less than consumption should reveal.. LOL Breathalyzer Jamer, sounds like $$$$$$$$$'s

[@AureliusR]

How could RF possibly influence this?

[@JacobChrist]

It would nice to see a follow up video where you consume the 1.5 oz of vodka every 20 minutes until you hit 1.0 on the sensor to see if you can get the decimal point to move.

[@sivalley]

+Jacob Christ No! Bad Person, no biscuit for you! Back to watching KZbin videos.

[@nonchip]

the part about oxygen levels seems interesting: could one trick the device into measuring a lower concentration by breathing in pure oxygen before blowing into the device?

[@rajeshcharan29]

what is alchohal normal percentage in breathe enal...

[@ibrahimmohamed6858]

Did you ever build your own?

[@ConcreteBombDeep]

Just realized Ben's voice is alot like Crispin Glovers.

[@krisztianszirtes5414]

I wonder what does it do with other alcohols

[@redtails]

hmm it's quite interesting that this breath tester is also sensitive to hydrogen.. There's a whole array of gastrointestinal conditions where people have quite significant levels of hydrogen in their breath. It'll most likely be too low to be detected by this specific test equipment, though still something to keep in the back of your mind that this tester isn't specific to ethanol

[@Bunjee98]

never ending loading screen

[@TheFlacker99]

Interesting

[@adrienperie6119]

So, in theory, if you hold your breath long enough before blowing in you might read a lower alcohol concentration ?

[@DeviousMalcontent2]

I'm going to go get something to drink. :^)

[@IllSkillz]

its like im watching eev blog =)

[@karlisbikis5013]

Great video! Next time I get stopped by the cops, I will make multiple fast breaths to increase oxygen concentration in my lungs before blowing in their breathalyzer. Then I will hope that more oxygen leads to lower reading :D

[@lekoman]

One wonders if holding your breath for a period of time before blowing into it (theoretically decreasing the oxygen content and increasing the CO2 in your breath) will torque the reading.

[@Cory989]

Then you wouldn't have enough air to blow the device long enough to show a reading.

[@guillermocastillo8488]

make some Carbon based gum?

[@starrychloe]

lekoman But that would also increase the ethanol content of your breath as well, since the gas in your lungs would have more time to absorb the ethanol escaping from your lungs.

[@RonJohn63]

+starrychloe Correct. Hyperventilating is the trick. But the cops know that, and so trying it is evidence of guilt.

[@johnsmith-gk3ek]

RonJohn63 how about a mouth full of spit blown down the tube? Copzi: Your not blowing! Slave: Bu!! shit! I'm blowing Copzi: Now i'm going to kidnap you, rob you, etc. "Land of the free and the home of the brave" lol

[@estrelladelnorte1308]

so this works like a wide band o2 sensor on a car?

[@Astinsan]

can the sensor be tricked with compressed air?

[@Superdan187211]

a long tome ago I had to pass one. BUT I was drinking so I used all my brain power to come up with sucking on the tube....and it WORKED ...but now they have valves to stop that ....I was not driving by the way.. then they made me pull over four the test

[@GeorgeTsilikas]

Drinking Vodka for science. Nice :)