3L smoking when engine heats up (1 Viewer)

[Shyrock]

We have a Toyota 3L engine in a 1997 Dyna that makes a lot of white smoke once hot. It only has approximately 48,000 miles on it. The engine does not overheat, and there are no signs of coolant loss. We have also checked to rule out air leaks into the fuel system. Ran it fueling direct through a couple of bottles of LiquiMoli Diesel Purge. The engine never missed a beat and produced no smoke. Switched back to the fuel tank, and it initially ran with no smoke. Took it for a road test and upon return there is white smoke when idling, and the engine runs rougher with what sounds like a miss. In the past I read that issues with the injection pumps can arise when hot, so we poured water on the pump to cool it down. With the engine operating at about 1500 rpm, through about 1/2 gallon of water and the white smoke disappeared. After a couple of minutes at idle, the white smoke reappeared. Then operated the engine at bout 2800 rpm and lots of white smoke - the smoke disappeared after slowly pouring about a gallon of water on the injection pump. Then again, a couple of minutes idling and the white smoke reappeared. With cold starts in the morning, it starts immediately and runs smooth - until the engine heats up and the white smoke starts again.
I would appreciate comments, suggestions, and questions. We can pull the pump and send for rebuild, I guess; but I thought I'd ask if any experts have other ideas - there are such low miles on the pump.

 

[Eurasiaoverland]

Does it have an exhaust brake?

 

[Shyrock]

It does not have an exhaust brake.

 

[Shyrock]

I think we will try to better rule out air leaks into the fuel system. Does anyone have troubleshooting suggestions for locating air leaks?

 

[Eurasiaoverland]

Bypass the whole system by running a hose from a clean jar of diesel to the injection pump. If you do this when it is hot you can see if the smoke disappears. If not, the pump may want re-sealing.

 

[Shyrock]

Bypass the whole system by running a hose from a clean jar of diesel to the injection pump. If you do this when it is hot you can see if the smoke disappears. If not, the pump may want re-sealing.

We basically did that with the 2 cans of LiquiMoli Diesel Purge, but it was from a cold start. There was no smoke so I don't know if the temperature got hot enough to cause the problem. We will try it with diesel as you suggest.
I spoke with a guy from Diesel Fuel Injection Service out of Portland, OR; and he said there is a cold temperature mechanism that increases the fuel until things heat up. He suspected that may be sticking open and overfueling once hot. He also said with the age of the pump that seals and o-rings degrade and likely need replacement - even with the low miles on it.

 

[Eurasiaoverland]

You can easily check the cold start advance by seeing if the piston moves out (cancelling the advance) when the engine is up to temperature. A high idle when hot would be indicative of the device having failed.

However, it does not 'overfuel' the engine, it advances injection timing. And under load at higher engine speeds, the dynamic advance should 'catch up' with the cold start advance.

It's not a bad idea to get the pump resealed and recalibrated, but it's not cheap to do.

Have you checked the static timing of the pump by checking the alignment of matchmarks on the pump mounting flange and timing case?

 

[Squash]

We basically did that with the 2 cans of LiquiMoli Diesel Purge, but it was from a cold start. There was no smoke so I don't know if the temperature got hot enough to cause the problem. We will try it with diesel as you suggest.
I spoke with a guy from Diesel Fuel Injection Service out of Portland, OR; and he said there is a cold temperature mechanism that increases the fuel until things heat up. He suspected that may be sticking open and overfueling once hot. He also said with the age of the pump that seals and o-rings degrade and likely need replacement - even with the low miles on it.

I use a clear fuel line to check for vaccum leaks. A bad line allows air into the system. Air shows up as bubbles...the more bubbles the bigger the leak.
Note: Does the primer remain stiff?
A soft primer also indicates a vaccum leak.
I generally install the clear hose between the filter/primer and IP. It's then just a matter of working upstream towards your fuel tank.
Generally it's the soft hoses or primer O ring which require attention because the remaining supply is metal.

 

[Shyrock]

You can easily check the cold start advance by seeing if the piston moves out (cancelling the advance) when the engine is up to temperature. A high idle when hot would be indicative of the device having failed.

However, it does not 'overfuel' the engine, it advances injection timing. And under load at higher engine speeds, the dynamic advance should 'catch up' with the cold start advance.

It's not a bad idea to get the pump resealed and recalibrated, but it's not cheap to do.

Have you checked the static timing of the pump by checking the alignment of matchmarks on the pump mounting flange and timing case?

Is this the piston you reference? It's the only item identified as a piston in the parts diagram I have access to.

You can easily check the cold start advance by seeing if the piston moves out (cancelling the advance) when the engine is up to temperature. A high idle when hot would be indicative of the device having failed.

However, it does not 'overfuel' the engine, it advances injection timing. And under load at higher engine speeds, the dynamic advance should 'catch up' with the cold start advance.

It's not a bad idea to get the pump resealed and recalibrated, but it's not cheap to do.

Have you checked the static timing of the pump by checking the alignment of matchmarks on the pump mounting flange and timing case?

Is this the piston you reference? It's the only item identified as a piston in the online parts diagrams.

 

[Eurasiaoverland]

What you have highlighted is the dynamic timing advance piston. 22790 (bottom right) is the cold start advance mechanism which your diesel shop was talking about (and which I don't think would be the cause of smoke unless your static timing was off).

 

[Shyrock]

You can easily check the cold start advance by seeing if the piston moves out (cancelling the advance) when the engine is up to temperature. A high idle when hot would be indicative of the device having failed.

However, it does not 'overfuel' the engine, it advances injection timing. And under load at higher engine speeds, the dynamic advance should 'catch up' with the cold start advance.

It's not a bad idea to get the pump resealed and recalibrated, but it's not cheap to do.

Have you checked the static timing of the pump by checking the alignment of matchmarks on the pump mounting flange and timing case?

The cold start advance is not cancelled when the engine heats up. The piston does not move out.
I'm not familiar with these pumps, but slowly understanding a bit more as I dig into it with the information you and others have provided. Thanks so much.
Is there a common location for the alignment match marks? Access is tough for this cabover design, so a starting point would be helpful.

 

[Eurasiaoverland]

Yes, I can imagine it's a pain to access. If you look at the front of the injection pump there is a flange in the housing where two unts hold it to the rear timing belt case. On the saide of the flange away from the engine there should be a scored line on the pump flange and timing case which indicate when the pump is properly aligned.

With the ACSD, most people remove them and add a blanking plate, but you might be able to buy a new one.

 

[Eurasiaoverland]

Different engine but the details of the injection pump flange are the same as on a 3L.

 

[Shyrock]

Updates:
1. Ran a heavy dose of diesel fuel addititve through a full tank of fuel. Smoking has been reduced drastically, but the exhaust still smells odd - maybe still quite a bit of unburnt fuel.
2. The automatic cold start device (ACSD) piston does not move once the engine heats up. It stays retracted at all times.
3. The fuel pump timing marks are not aligned. See photo.

Would this indicate advanced or retarded timing?
From what I understand, with the ACSD not functioning, the ACSD keeps the timing adjusted for cold start operation continuously. Could someone have tried to compensate for that by altering the pump timing?
I think we should delete the ACSD, blank it off, and adjust the pump to line up timing marks. Our weather is moderate, 20 deg F (-7 deg C) a few times each winter.
Is this a proper path forward. What am I missing?
How many bolts hold the injection pump? We are battling access to everything with the cabover design.

 

[Eurasiaoverland]

Good progress. The misaligned marks indicate that the injection timing has been retarded. The failed ACSD will be keeping the engine advanced, though only at idle and low speeds. As engne speed increases, the dynamic timer advance (based on internal fuel pressure) will advance the timing and so the effect of the engaged ACSD will be reduced/negated. So once the engine is running at higher speeds, the injection will be too retarded.

Deleting the ACSD is a legitimate way forward, just be sure to properly block off/loop the hoses.

The injection pump is held on by two bolts on a stay at the back of the pump, and two nuts on the front mounting flange of the pump body (one of which has the matchmark on your photo). Getting to the nut on the engine side of the pump is hard at the best of times, so I can imagine it will be a real joy to try to reach it in your vehicle. Usually you can get it directly with a 12 mm spanner, or from the back with a spindly 1/4" ratchet extension and universal joint.

 

[Shyrock]

@Eurasiaoverland thanks so much for your input. Now to have a go at accessing the mounting bolts.

 

[mudgudgeon]

The timing marks on the pump flange are vague. Not acurate.
Use then as a guide only.
You really need to use a SST and dial gauge to measure plunger stroke at TDC to be sure of the timing. Then small tweaks can be done from there.

Retarded timing could cause some smoke.

Did you verify no air in the fuel system with a clear hose?
Air in the system messes with injection timing as air bubbles compress, fuel does not. Air bubbles in an injector line means there may not be enough pressure to overcome the springs in the injectors, as a result injection may not happen, or may happen late. This will mean the engine runs rough, or down on power when under load/accelerating etc.

 

[Shyrock]

The timing marks on the pump flange are vague. Not acurate.
Use then as a guide only.
You really need to use a SST and dial gauge to measure plunger stroke at TDC to be sure of the timing. Then small tweaks can be done from there.

Retarded timing could cause some smoke.

Did you verify no air in the fuel system with a clear hose?
Air in the system messes with injection timing as air bubbles compress, fuel does not. Air bubbles in an injector line means there may not be enough pressure to overcome the springs in the injectors, as a result injection may not happen, or may happen late. This will mean the engine runs rough, or down on power when under load/accelerating etc.

We did not check for air bubbles w/ a clear hose yet. We will do that for confirmation.
The drastic difference in smoke from cooling the injection pump as indicated in the original post led us to look for some temperature related issue. Air issues kind of became secondary then, but I need to make sure to rule air leaks out.

 

[mudgudgeon]

Air in the fuel is a really common cause of white smoke.
It could be the cause of your issues, and is really easy to diagnose.
Also, if air is the problem, its usually a cheap fix, $10 for a piece of clear hose, a few bits of rubber hose, maybe a new primer pump, maybe some new hose clamps and you're good.