Modification for D-Jet M116 and M117

[Greg in Oz]

I have just performed this simple modification to my 1973 350SLC (with 3.5 M116) and am happy with the results. See the thread at Improvements to M116 in early 107 (true cold air intake)

[tosbt]

Would really like to see some pictures, I just did a slight modification to my 4.5 because I think the motor is starved for air or cold air.

[Greg in Oz]

Pictures are on the thread at Improvements to M116 in early 107 (true cold air intake)

[GaryF]

Nice neat job, it'll be interesting to see if you find more crud around the airfilter after a few hundred miles.
I presume rain water couldnt be sucked up through the intake when driving through a heavy shower....although probably less of a problem in Oz unless you're in the North.

[Greg in Oz]

Gary,

Even in Sydney we do get occasional heavy rain. I had considered the possibility of the cold air intake ingesting water however I don't believe it will be a problem.

Firstly, the air velocity into the intake pipe is only high at wide throttle openings and high engine speeds, an unlikely situation in heavy rain.

Secondly, looking at the pictures I posted, you will notice that water would have to get through the grill, then the holes in the panel between the headlight and radiator (oil cooler on that side actually), then the intake pipe which runs uphill to the flexible hose. If water does actually get beyond the flexible hose, there are a couple of small diameter holes at the bottom of the original intake pipe of the air filter which should allow it to drain. If water could even get past here and reach the air filter element, the air velocity is lower through the much larger cross-sectional area of the element itself and I doubt the water would get beyond this point, instead lying at this low point within the filter housing and eventually evaporating.

To reinforce my reasoning, both my 190E and 300TE have original factory fitted cold air intakes behind the grill with little to suggest they would prevent rain water being ingested any better than my DIY cold air intake on my 350SLC.

As for additional material being collected in the air filter that may be possible, particularly larger objects such as insects. Again, I don't expect this to be a great problem since the air filter element has a very long life and the filters on my other cars with original intakes similarly located don't have a problem with the air filter becoming contaminated quickly.

Greg

[tosbt]

What a great idea! Well done and thanks for the photo's.

[Tomguy]

It's a good idea, but there's a major problem with it in my mind:

The D-Jet operates off of the theory that the air intake pressure is the same as the hood pressure - the MAP sensor is at the firewall on a 108 and 109, not sure about a 107 (It was moved on the 116, presumably because of this). When you moved the air intake's inlet to an area where the air pressure is much higher, you're now letting more air to the engine (which is good), but the engine's computer doesn't know it. In other words, you're going to be leaner at highway speeds. I think what you need to do now is make a modification so that the MAP sensor is in the same area as where the air is coming in, or at least make it so that is the area from which it senses the air pressure. possibly drill a hole in the air filter housing, attatch a hose to it, and run the hose to the allen-screw end of the MAP sensor and clamp it on (If memory serves right, it is round and nothing would interfere with this idea). Then the D-Jet ECU, primitive as it may be, can more accuratley judge the intake air pressure and richen your mixture accordingly (for more power!)

[Greg in Oz]

Originally Posted by Tomguy

It's a good idea, but there's a major problem with it in my mind:

The D-Jet operates off of the theory that the air intake pressure is the same as the hood pressure - the MAP sensor is at the firewall on a 108 and 109, not sure about a 107 (It was moved on the 116, presumably because of this). When you moved the air intake's inlet to an area where the air pressure is much higher, you're now letting more air to the engine (which is good), but the engine's computer doesn't know it. In other words, you're going to be leaner at highway speeds. I think what you need to do now is make a modification so that the MAP sensor is in the same area as where the air is coming in, or at least make it so that is the area from which it senses the air pressure. possibly drill a hole in the air filter housing, attatch a hose to it, and run the hose to the allen-screw end of the MAP sensor and clamp it on (If memory serves right, it is round and nothing would interfere with this idea). Then the D-Jet ECU, primitive as it may be, can more accuratley judge the intake air pressure and richen your mixture accordingly (for more power!)

A valid point and one I had considered, but not a "major problem" in my opinion for the following reasons:
1. I don't believe the air pressure forward of the radiator would be "much higher" than that in the engine bay. I believe the radiator is reasonably "transparent" to air movement.
2. At highway speed under normal cruising conditions the throttle would only be partiatially open and therefore be the major restriction to air flow. Under these conditions the manifold pressure is very low (ie. high vacuum) and any slight difference in air pressure at the mouth of the air intake versus that of the engine bay would be negligible compared to the low manifold pressure.
3. My understanding of the operation of the MAP sensor is that it causes the mixture to become richer as the manifold pressure approaches ambient pressure. Would this not suggest that any higher manifold pressure at wide throttle openings resulting from slightly higher air pressure at the air intake would actually cause richer (rather than leaner) mixture with the MAP sensor referenced to the air pressure within the engine bay?
4. Under conditions of wide throttle opening air flow is high. I would suspect that air pressure within the air filter (which you suggest as alternate reference for the ambient side of the MAP sensor) would be as low if not lower than that in the engine bay under these conditions.

This is of course all speculative. Measurements by way of exhaust gas analysis under real driving conditions would be a means of backing up any suspicions. D-jet being an open loop system tends towards rich running anyway. Unless a modification caused significantly leaner running, I would not expect any problems.

[Tomguy]

1) If this is the case, the rest would be moot. But you must think there's some advantage, thus why you did it

2) At partial throttle, if the air pressure is greater because of the cold air intake sucking air from a higher pressure area, the engine will be receiving more air than the MAP is telling the ECU it is.

3) Yes, you're increasing the manifold pressure, which would (should?) cause the engine to be richer because of the way the MAP works. The DJet ECU for the 3.5, from my understanding, doesn't have the 10% enrichment at full throttle like the 4.5 one does. With a 4.5, this would cause black smoke to POUR out the exhaust!

4) I'm unsure of this as the cold air intake seems like it'd be in a higher pressure area, where full throttle at high engine RPMs would also mean high speed, in which case, pressure would be quite high in the air filter. Although, I don't know, and I WAS thinking about this earlier (I'm trying to devise a way to get more air into my engine as well, but have yet to modify the "Snorkel" of the air cleaner assembly, might tomorrow)

edit: Actually, I just thought of a way to measure the air pressure at different spots under the hood while running in comparison to the outside air pressure. Hard to explain, I'll just take pics while I do it!

[Greg in Oz]

Tomguy,

I appreciate your thoughts and I believe you raise an interesting question. I've been thinking some more about this so I'll try to put it to words. See what you think of this reasoning.

With the D-Jet, the only inputs are (I hope I haven't missed any):
1. Timing signal from the trigger points,
2. Coolant temperature,
3. Air temperature,
4. Manifold pressure,
5. Opening throttle (temporary enrichment only while throttle opening is increasing),
6. Closed throttle (for fuel cut on overrun),
7. Wide open throttle (4.5 only, you are correct the 3.5 does not have this).

From this I deduce that for any throttle opening apart from closed (or wide open on the 4.5), the ECU does not actually know what the physical throttle opening is. The mixture is therefore determined entirely by manifold pressure (vacuum), coolant temperature, air temperature and engine speed. At all but wide open throttle at high engine speeds, the only significant restriction to airflow will be the throttle itself. Is it therefore reasonable to say that it matters little what the ambient pressure is at the air intake? Under steady cruise conditions, even at high speed, the relatively small throttle opening will be the only significant restriction to airflow and the manifold pressure will be determined by the throttle opening. For a given cruising speed, a certain amount of engine power will be required. This power will be dependent upon the amount of fuel and air consumed. If the air pressure at the air intake is slightly higher, then a correspondingly (if only slightly) smaller throttle opening will result in the same manifold pressure giving the same signal from the MAP sensor and hence the same fuel mixture.

At wide open throttle, then I agree that other restrictions apart from the throttle, such as the filter element and intake pipe (snorkel) become more significant. You are correct in suggesting that I felt there must be some benefit to improving air flow. I have made use of an intake pipe with a larger cross sectional area in the hope of reducing restriction to the airflow at wide open throttle and high engine speed. My primary reason for taking that air from ahead of the radiator however, was to allow cold air to be inhaled rather than any possible "ram" effect. If there is any ram effect then I believe it will only be of any significance at wide open throttle and at high road speed. Hopefully, should this situation occur, the resulting manifold pressure will still be within the working range of the MAP sensor and that the resulting fuel mixture will be correct. The colder air will be accounted for by the air temperature sensor on the intake pipe and the mixture slightly enrichened as necessary.

The greatest benefit I have realised (and this would be due to colder air) is the elimination of pinging (knocking or detonation) under load and during acceleration without having made any other changes. I have also noticed improved throttle response (probably primarily due to ellimination of pinging) and the engine also seems more willing to pull under load in the upper rev range at wide open throttle. Previously it appeared to be somewhat "strangled" above about 5,500 rpm (even though it could still pull to the 6,500 rpm recommended limit). Now it continues to pull strongly and quickly to this limit with no indication of "running out of breath". Whether this is due to less restriction to airflow, lower air temperature, or a combination of both, I cannot be sure. Possibly it was previously pinging in the upper rev range as well (but not audible due to masking by other engine noise).

Hopefully you can follow my logic and it offers further food for thought.

Greg

PS. You could make use of a water manometer to measure air pressure at various locations in the engine bay as well as forward of the radiator, just as a water manometer can be used to measure pressure drop at various points in an air intake and air filter system ahead of the throttle body. I am just not sure that it is very significant due to the reasons given above. Readings from an exhaust gas analyser under real driving conditions before and after the modification would be worth seeing. Unfortunately, I do not have access to this equipment to try it.

[yogiysh]

Ok, I couldn't wait any longer and tried a very crude test over the weekend to see if there are benefits of cold air induction on Tobias (280se 3.5L W108).

Greg, Thank you for the advice of not doing anything drastic or irreversable cutting and drilling so I simply opted for a single 80mm flexitube (AUS$25.00) from the local auto parts store to run a test.

I haven't cut the snorkel but simply inserted it into the flexitube until the cold air sensor and found that the ouside circumferance of the snorkel is still way too small for a snug fit, so I used a strip of cork (about 4mm thick) around it to fill the difference prior to clamping.

Then I fed the flexitube down beside the alternator and then vertically down behind the sway bar then with a right angle bend straight to the front of the vehicle attaching the open end to right next to the towing point with cable ties. Yes, very crude indeed !!
I did put a fine wire mesh across the openning to stop large particles getting sucked in.

Anyway, with no changes really apart from a constant feed of cold air, I was really suprised at the results. I performed the test in city driving, freeway driving and up a very steep roads up Mt Keira - Wollongong.

I second Greg's findings in the above linked thread and did find that there's remarkable increase in power and quick accelleration, lots of power and pull all the way through to the high revs and no pinging or black smoke even under load up very steep roads. I was using 98 Octane fuel.

I just couldn't believe it !! So I took it all off and did the steep road test again and back to the old symptoms and performance. I tried it on the freeway going back home and got the same results. I reconnected the flexitube again at a road stop on the freeway and what do you know - back to better performance!

To experience such difference in performance with such a crude set up and simple testing, for me warrants the next stage.

So, I guess the next step for me is to address the snorkel and design a very inconspiquious airscoop.

I'm also hoping that the nearly two feet of vertical section of flexitube addresses the water issue as well as having drainage holes in the air filter housing - I'm not planning to drive in flooded areas and do any river crossings anyway...

Any thoughts on ramming the air in to an extent?

Tomguy, I'm looking forward to the results of your air pressure tests.

Best regards,

Ysmael

[Greg in Oz]

Ysmael,

I'm pleased you gave this a try and also noticed some improvement. Mount Kiera is certainly a good test!

I would expect even better results if like me you eliminate the narrow end of the intake pipe to increas the cross-sectional area to match that of the 80mm flex pipe. I would also expect that the mesh you fitted at the intake would also cause some restriction.

I probably would not be too concerned about fitting mesh since the filter element is designed to stop everything from bugs down to dust. I would be a little concerned with having the intake down as low as the towing point though. You never know when you may encounter a deep puddle or a flooded road and an intake this low is risking serious water ingestion. I would also expect more dust would be encountered this low down.

I have not studied a 108 with a view to fitting a cold air intake, but is there either sufficient space next to the radiator behind the grill or alternatively is there a gap under the leading edge of the bonnet where it closes down onto the panel between the grill and headlight? Maybe an intake could be fabricated with a narrow oval shape. Just so long as the cross-sectional area is no less than an 80mm pipe.

As to any benefit from a ram effect at speed, I am not sure. The issues Tomguy raised may well come into play here. It really depends upon whether the fuel mixture remains correct when the manifold pressure is slightly higher as a result of any slight ram effect combined with the enlarged cross-sectional area of the intake. Hopefully it will still be within the range of the MAP sensor and the injection system can provide any additional fuel required. I feel that only exhaust gas analysis under true driving conditions would confirm this. The tendency for D-Jet to run somewhat rich at least provides some safety margin.

Greg

[j9fd3s]

couple things that come to mind reading this.

1. the percentage of airflow change might not be very big. we tuned a friends rx7 with an aftermarket computer a couple weeks ago, and the difference between running with a 3" pipe in place of the factory air flow meter and the 2" air flow meter in place was very very small, when looking at the mixture with a wideband oxygen sensor. the haltech is also a map sensor based system.

2. we talked to the guy who runs the volvo in the scca touring car class, and they had the air intake in the headlight. they also had a duct to the bottom of the bumper. we asked him about it and he said the headlight duct alone had positive pressure until 80-90mph and they the pressure went away and became vaccum, so they added the second duct. to reiterate, if you picture a volvo, they cut a hole in the front of the headlight and stuck the intake duct on the rear of it, and found that a forward facing hole didnt always get positive pressure.

mike

[Tomguy]

Got around to doing those tests today. Here is what I did and the results:

Picture 1 shows the 3 locations I chose to measure the pressure from: 1) The air intake horn, 2) The MAP sensor, 3) The cowl. I chose the cowl because I think this area, by looking at the front of the 108, would provide the most positive air pressure while being the best spot to do a cold air intake at (by going through the firewall into the engine compartment).

Picture 2 shows my archaic, simple setup. Each 3/16" diameter hose has exactly 3mL of water in it (measured with a disposable syringe). I put dye in the water to make it easier to see.

Now, I couldn't do any more than glance at the tubes while running but I do think the results are surprising.

Picture 3 shows the air pressures at 40MPH. The cowl has the largest pressure increase.

Picture 4 shows the pressures at 80MPH. The air intake is actually becoming a vacuum, while the MAP is a good deal higher. Again, the pressure at the cowl is much higher.

Picture 5 shows 100MPH. The intake is a definite vacuum, which must be robbing engine power. The MAP is fairly higher, and there is a LOT more positive pressure at the cowl.

Now, you might think that perhaps, due to the engine sucking in a lot of air at speed, this has something to do with the air pressure being a vacuum where I measured it. However, I revved my engine up pretty high in the driveway with no change on any of the 3 locations. This leads me to believe that at speed, air passes over the front of the car at the hood seam, instead of into it. This passing over creates a lower pressure area inside the hood, at least at the air intake horn, much like the effect on the top side of an airplane wing.

This is probably why you notice a much better throttle response at speed. Perhaps later, I will compare the cowl to the tow hook. For now, these are some interesting results I think!

[DieselBone]

Nice job on the test. I really liked the home-made set up, very smart!