Home-built/DIY smoke machines for intake/evap leak testing (bad ideas inside).

[phucheneh]

Hey AT Garage. Been a while since I've been here. I got aggravated by ATOT and quit posting...I dunno, a couple years ago? I forgot my login info and email had changed, so I had registered a new account a little while ago. Hopefully that's okay and we can just leave this little intro at that.

I wanted to post and see if anyone else has ever tried to build their own device for 'smoking' vehicle intake and evap systems. I've recently quit working for the stealerships, and I think I'll find some success on my own. But sometimes you REALLY miss having a particular tool, especially when that tool is ~$1000 or more to buy new.

The commercial smoke machines I've used always has some kind of propriety solution stored in them, and you'd simply hook a couple leads up to the battery, and hit a pushbutton. Smoke would roll out of the attached vacuum hose fairly quickly. Most of them simply consist of a glow plug (like in a diesel) or similar heating element, somewhere for the smoke to acculate, and something to push the smoke through the hose and into the the area you're leak testing.

Google found a lot of discussion and a few working examples, which I used, together with my own intuition and whatever crap I had lying around, to build my own smoke generator. I'm almost ready to start testing...and then, surely, tweaking my design.

Anyhow, behold my ghetto masterpiece. Ignore the destroyed carpet and dirty room, please...there's a reason I do these things in the ugliest room in the house. And it stays ugly.

This is the remote. It is carefully fabricated from some cheap radioshack switches, some wire, a couple bolts, and tape. Classy, right?

The commercial units have a single pushbutton, like a remote starter (of the diagnostic variety). AFAIK, this immediately turns on both the heating element and whatever the machine uses to move air. Since I have no idea how either of these things will perform, I'm using two seperate switches. One is momentary, the other toggles.

And it leads down about ten feet of wire to this stuff. This is so I can move around while continuing to pump smoke as needed (since evap leaks can be at either end of the vehicle). Also, this is important because most 'home-built'/DIY machines simply use switches or valves to turn on the air supply and heating element. I don't believe I want either to run constantly.

Glow plugs get over 1000*F. You don't want to just...leave that on. And you don't want to build significant pressure inside of the vessel holding hot mineral oil (or some kind of glycol solution...whatever makes the best smoke without being bad for the vehicle- havn't decided).

The glow plug is in the bottom of that paint can, so I can control how much of the liquid is covering the the heating element. Fitting in top is the smoke outlet, the one in the side is the air inlet. Those air hose fittings are kind of placeholders until i test this and see how hot it gets. Also I'll seal them up around the base with some high-temp epoxy.

The cables at the top would go to battery. Yes, the negative is a pair of ~12g speaker wire using all the copper for the same function. Positive is 8ga from an old car audio install. With a fuse, of course.

The negative goes straight to a clamp that fits snugly on the body of the glow plug (that whole can will be a ground). Another small wire ground the relay coil. The positive goes to the high side input of the relay, with the output having a spade terminal for the glow plug. It also has a small wire that leads to the switch, which will trigger power for the relay coil.

The other switch leads to that molex connector, which connects to my modified portable compressor. I could've just used the little compressor bit, but I figured I'd keep it together and just tap into the on/off switch.

There's also a computer PSU I may put with all this so that I can plug it into a wall, instead of using the car battery.

Total investment right now is like...20 bucks, maybe. I bought the relay and the glow plug.

Yes, this will need testing and tweaks. That will be tomorrow.

Hey, it beats paying a grand to have someone essentially engineer the same device and put everything in a pretty case.

edit- Oh, and nothing is 'done.' Bare connectors need to be covered, everything should be on a sturdy base, ect.

 

[Throckmorton]

I believe smoke machines use propylene glycol. Ethylene glycol is very toxic. Propylene glycol is toxic too but not nearly as bad. It's used in makeup.

Apparently you can also use glycerine http://chemistry.about.com/od/howthingswork/a/smokemachines_3.htm

 

[herm0016]

we used "food grade mineral oil" in our smoke machines. interested to see if this works.

 

[franksta]

we used "food grade mineral oil" in our smoke machines. interested to see if this works.

This is what I have seen used before too.

 

[phucheneh]

I finished this today. Sort of. Going to have to run to the grocery store and get mineral oil. I would assume I can buy the right kind there? What's 'food grade' imply?

I've got my PC power supply rigged up. Glow plug gets up to temp at the tip almost immediately (...it glows). Might try using 5v power so I don't have to keep turning the plug on and off. Have the side air inlet set up, just need a hose for the top. And possibly an accumulator/cooling tank of sorts, depending on how fast I can produce smoke and how hot it is.

Will post a youtube video later. To include any fire department visits.

 

[manimal]

Great project man, keep us updated.

 

[Squisher]

Son's car threw an EVAP code today, something like this could come in handy. I wonder what I'd look like with a quart of mineral oil aflame on my clothes? Richard Pryor always said people get out of your way when you're running and you're on fire.

 

[phucheneh]

Argh, damn you quick reply. I typed way too much then accidentally hit the 'back' button. Post gone.

I was trying to post a question for the more scientifically-inclined. Chemistry, more specifically, I guess...with regards to flaming mineral oil.

1) Mineral oil, like any oil, needs an oxidizer to burn, yes?
2) Is the reaction that creates smoke different from the reaction that would cause actual combustion (or just 'fire.' I know there will be no exploding).
3) The temp at which most oils smoke, and the temp of autoignition are pretty close together. I'm wondering how commercial machines eliminate the risk of fire...inert gas? The oil can't light without oxygen, but can it smoke?

This is assuming actual 'smoke' and not vapor. I could be wrong, but I don't think the latter is very visible. Plus oil would condense inside the evap system.

 

[phucheneh]

Damn. Success. The worst kind of success. Everything works as planned...no leaks, no electrical problems. Things I was worried about are non-issues.

...but you don't get the results. Maybe the 'glow plug in the bottom' idea wasn't the best. It seems like the volume of oil required to partially immerse the heating element is...well, too much oil. I think it's just working too damn well as a cooling medium. The paint can is not even very warm. As much as I don't want a toggle for the glow plug (in case I forget it's on), I think that may be the way to go. Switching my...switches, that is. Only occasional bursts are needed from the compressor to keep the air (and smoke) moving.

Glow plug triggered:

Little old tire inflator, connected to a valve stem with the rubber cut off (exposing a metal tube). That's attached to some other random piping in an attempt to keep heat away.

Another overview:

...cake pan is a temporary safety measure. I think it's time to take this outside and see if I can't make fuckin' fire. The 'heating element' is obviously not going to work as quickly as I want unless I switch to a different container, I think.

 

[phucheneh]

Think I figured it out. Need to blow air directly onto the glow plug, then it smokes much better. So that air inlet pipe just needs tweaking, and the compressor should indeed stay on all the time. I think it's low volume enough that I shouldn't even have to worry about a pressure relief.

 

[phucheneh]

Welp, I've yet to 'optimize' this. As in, 'make it work right,' but I do need to test a car tomorrow.

I may have to resort to the 'soak a rag in gas, light it on fire, and shove it in the can' method.

On the plus side, this little 300w Enermax PSU I've wired up for 12v/5v supply is a hoss. Been using it for various 12v stuff, and I've yet to encounter a problem.

There are a lot of people who say 'omg that's dangerous, you can't do that,' but...feh, I don't see an issue. I've used factory-provided switching power supplies as battery 'maintainers' during software flashes and the like. They're really not any different...just lack the extra rails, and probably have a small internal load to keep them stable.

All I did to this one was stick a 10 ohm resistor on the 5v rail (so it burns off about .5a as heat) as a dummy load, and wired a switch onto the remote turn on wire. And I connected the 3.3v sensing wire to the worthless chopped off 3.3v power wires (not sure if neccesary).

 

[phucheneh]

I got this working well today. The flow of smoke was very similar to the commercial machines I've used. But I just could NOT get it dense enough.

So I did end up switching mediums from 'mineral oil' to 'gas-soaked rag.' It worked fine...going though the evap test port behind the purge valve, I was able to observe smoke rolling out of the canister vent.

Hot mineral oil = nothing. Flaming rag...slightly harder to handle. Couldn't get the container closed up without the fire going out. Then I realized...oh wait, you've got a glow plug in there to light the rag without accidentally getting your eyebrows removed by a backdraft.