$8 Smoke Test, with pics (and also high idle issues)

[Timp]

Hi guys,

A couple of months ago I did a valve adjustment on my '87 ETA, and my idle was fine. A week later it started idling at 1300 to 1500 RPM's. I called around to local shops and the cheapest place I could find to do a smoke test was $40. I didn't want to deal with this so I looked up some DIY options. Most used fog machines that you'd use during halloween or a dance party. I did not want another machine around, so I looked into other options.

I found this video where a cigar is used to fill the intake system with smoke. However, I have no interest in smoking. But after thinking a bit, it turns out a $7 transfer pump from Harbor Freight I purchased to do a tranny fluid change worked perfectly. Here's how I set this up (descriptions are below each picture):



These were the cheapest things I could find at Walgreens.



Here it is all set up. The cigarillos just so happened to fit perfectly in the tubing that comes with the transfer pump. Once you start pumping, it burns quickly. I tried this before connecting it up to the car and it produces a lot of smoke. Was this the right tube to plug it into? I don't know - but it fit perfectly as well.



After burning this much off, there was absolutely no sign of smoke coming from the bay. I opened the oil cap to see if smoke was getting down there to my suspected leak area. As soon as I opened it, it began spewing smoke - at least it was actually getting down there.



Here's a closeup. It was hard to get the amount of smoke on camera. Much more than you can see here. For higher res pics, you can go here.

After all of this, I still have a high idle issue. Is this as good as a smoke test at a shop? I don't think so. Is it sufficient? Well, the only way to know is if I take it in and compare this to an actual machine meant for these things. From what I know, the ones they use at shops are pressurized (just a few psi).

If anyone has any suggestions on how I can improve this, let me know. I'd like feedback. I've since used my pump for the tranny fluid change, so I don't know if I'll be using it again for this purpose but at $7, I wouldn't mind buying another.

Also, any ideas on where to look next for my high idle issue? If I unplug the ICV, the RPM's shoot up to 3K. From what I understand, this should only be at 2K. In addition to that, the resistance on my coolant temperature sensor was 65 ohms 15 minutes after returning from a 10 minute long drive. The Bentley manual states that if the sensor temperate is 176° F, the resistance should be 270 to 400 ohms. I had no idea what the temperature of this is, so I have no idea if this is faulty or not. Any tips with this?

[jlevie]

Critical parts of a smoke test are plugging the exhaust (and the intake boot if you use one of the hoses to feed in smoke) and achieving and holding a smoke pressure of 2-4psi for several minutes. While smoke can be introduced via any fitting/hose, I prefer to use an adapter that replaces the AFM. That means that I am testing the way things are in normal operation.

If you didn't plug the exhaust & intake boot or didn't hold smoke pressure long enough there could be leaks you didn't find.

Might you have checked the gauge sensor rather than the ECT sensor? The gauge sensor will have a much lower resistance.

$40 for a properly run smoke test is a really good deal. About $100 would be a more normal charge.

[Timp]

Right, I didn't think of the exhaust but I did think of the sealing the intake boot. Holding that much pressure for that long with this method doesn't seem possible. But... perhaps my lungs would be able to fill the intake system with a little bit more than atmospheric pressure assuming both the exhaust and boot are well-sealed off? Any ideas on what I can use to cap off the boot?

I will still probably take this in to get it smoked just for the purpose of knowing for sure. Also, I may very well have tested the gauge sensor instead of the ECT sensor. I will look at this again.

[Gregs///M]

Critical parts of a smoke test are plugging the exhaust (and the intake boot if you use one of the hoses to feed in smoke) and achieving and holding a smoke pressure of 2-4psi for several minutes. While smoke can be introduced via any fitting/hose, I prefer to use an adapter that replaces the AFM. That means that I am testing the way things are in normal operation.

If you didn't plug the exhaust & intake boot or didn't hold smoke pressure long enough there could be leaks you didn't find.

Might you have checked the gauge sensor rather than the ECT sensor? The gauge sensor will have a much lower resistance.

$40 for a properly run smoke test is a really good deal. About $100 would be a more normal charge.

What do you mean by plug the exhaust? Like plug the muffler tips? Is there another, more local way to plugging exhaust?

[LJ851]

What do you mean by plug the exhaust? Like plug the muffler tips? Is there another, more local way to plugging exhaust?

No.


Well, i guess you could remove your cam or rocker arms so you wouldn't have any valves open, but that seems like a lot of work.

[Gregs///M]

I think we should establish a correct or "best practice" way of performing a smoke test. Then hopefully someone can figure out a cheap, efficient way to achieving good results.

[MoreMayhem]

Smoke test with cheap Walgreens cigars may be habit forming.

[Timp]

Thank you for the input everyone. After reading the comments, I went to a shop to compare my results. They found no leaks. Even though the method I outlined originally can definitely be improved in a number of ways, it at least yielded the same results.

Now, what to tackle for this idling issue...

[Gregs///M]

Without a doubt, the first thing you must check is the Throttle Position Sensor. located under the throttle body, it is a 3 pin sensor. Pins 1 and 2 should show continuity with throttle at idle position, pins 2 and 3 should show continuity with throttle near 70% mark. Continuity between the contacts should always be one or the other, never at the same time. Check with a multi meter. GL

[jlevie]

I think we should establish a correct or "best practice" way of performing a smoke test. Then hopefully someone can figure out a cheap, efficient way to achieving good results.

That is pretty easy. The exhaust has to be plugged and the AFM replaced with an adapter for introduction of smoke. The smoke needs to be a long persistence oil smoke. Smoke is then fed in until a pressure of 2-4psi is attained and that pressure held for a while. If no leaks are immediately found, hold the smoke pressure for 10-15 minutes. But if leaks are immediately found, fix those and repeat the test.

[scottinAZ]

using this method is known to the state of California to cause cars to rust (cancer)

[Timp]

Without a doubt, the first thing you must check is the Throttle Position Sensor. located under the throttle body, it is a 3 pin sensor. Pins 1 and 2 should show continuity with throttle at idle position, pins 2 and 3 should show continuity with throttle near 70% mark. Continuity between the contacts should always be one or the other, never at the same time. Check with a multi meter. GL

Thank you for the recommendation. Did you suspect this was the issue because of your experience or because of my specific problems I listed? Just curious.

I ended up checking the TPS - it was fine. There was continuity in the positions you mentioned. Also, I ended up finding the coolant temperature sensor, (which had two separate wires going into the prongs, unlike the picture in the Bentley manual) and tested resistance between the two after the car sat for an hour and a half. The resistance jumped anywhere from 0 to 1000 ohms. There was some play in the two prongs coming out, and this would affect the resistance. The engine was warm to the touch.

This is what is happening with my idle: In the mornings, after a cold start the engine will idle at around 900 RPM. By the time I get to work after about 12 minutes of driving, it is higher - around 1300 RPM. When I unplug the ICV, when the engine is warm, it idles at 3K. The idle is about 90% consistent, and occasionally moves up and down, but not drastically.

[jlevie]

You really need a properly run smoke test. There are other possibilities, but it is pointless to investigate those until intake leaks are ruled out.

[Timp]

You really need a properly run smoke test. There are other possibilities, but it is pointless to investigate those until intake leaks are ruled out.

Not sure if you've read the whole thread, but I have already done this. I'm not sure they left it pressurized for 10-15 minutes as stated above, but they told me they found nothing. I suppose I could get a second opinion.

[jlevie]

Not sure if you've read the whole thread, but I have already done this. I'm not sure they left it pressurized for 10-15 minutes as stated above, but they told me they found nothing. I suppose I could get a second opinion.

I see that now, but the question is whether the test was run properly.