L Jetronic Tuning -- Surging Idle and High RPM Lean Out

[1975cs]

I am hoping to get some L Jetronic insight and also provide some information that may help others tuning theirs. Two issues for discussion:

1). In Closed Loop Mode (Oxygen Sensor connected) as my engine warms from Cold Start idle speed grows from 550 RPM to 800 RPM as the sensor becomes active. With slight throttle above idle, the engine RPM oscillates up and down a bit. This goes away with increased pedal.

2). I am concerned that my engine may be leaning out excessively at high RPM and I am looking for insight on proper tuning for high RPM as well as idle.

As background I have a stock 1978 USA 528i L Jetronic system running a 3.5 liter BMW M30 with larger 320i intake runners, Metric Mechanic 275 degree sport cam, ported head, b35 exhaust manifold and oversize exhaust. Engine was rated at 260 HP when rebuilt 25 years ago and has 125K miles on the rebuild. L Jetronic is stock except for an adjustable fuel regulator (currently set a 2.5 bar without vacuum which is factory spec for 528i and 530i).

Surging idle is usually caused by a vacuum leak, which I do not have. My system pulls up to 20-22 IN HG. If I add air and reduce vacuum to 14-15 IN HG the oscillation at light throttle turns into an oscillation at idle (Closed Loop Mode only). If I disconnect the O2 sensor to move to Open Loop Mode, the idle is stable (about 425 RPM), and it does not increase to 800 as engine warms. RPM oscillation at light throttle (or idle) is not experienced.

Anticipating your suggestions: Aux Air Valve works properly (I have swapped this with another BMW unit, exchanged with a new Benz unit and also ensured there is proper function hot and cold); Coolant Temperature Sensor has been exchanged with new; Thermo Time Switch works properly and Cold Start Injection has good spray and shuts off properly; fuel pressure is correct; Throttle Body is properly adjusted as are Throttle Switches and Butterfly; AFM seems fine and is at factory spec (hasn't been adjusted); AC Bypass Valve works properly. Originally I thought the Oxygen Sensor might be a fault so it was replaced it with a new one. None of these seem to effect the Closed Loop problem I am experiencing.

I have previously experienced some stumble with a cold engine which was usually fixed with idle adjustment or a new Oxygen Sensor. Also I have wondered if my idle might be too rich (spouse complains of gasoline smell). It was only when I replaced old rubber hoses, rebuilt my brake booster, AC system and power steering all as part of an engine compartment restoration, that I encountered this idle speed instability in earnest. So I bought a Heathkit Exhaust Analyzer to help with the diagnosis and tuning.

I believe spec on a 528i is 0.2 - 0.8 %CO at idle speed (Haynes) while a 530i (Open Loop Mode - no oxygen sensor) is 1.5 - 3.0% CO. In Closed Loop Mode my system is rich versus 528i at 1.9 % CO at cold start (13.9 Air Fuel Mixture) at 550 RPM. I then warms to 3.0% CO (13.4 AFM) at 800 RPM. In Open Loop Mode my idle is steady at 425 RPM 2.2% CO and 13.7 AFM with mixture levels nicely in the middle of the 530i spec range.

If I adjust the Idle Valve on the Throttle Body to raise Cold Start idle speed I move the oscillations into idle from light throttle and the idle RPM then surges to cutoff at about 1400 RPM then drops abruptly oscillating (Closed Loop Mode only).

The Bosch Injection book recommends adjustment of the AFM bypass air valve to deal with surging idle. The procedure recommends adjusting from a lean position to richer until surging stops. Also I could: move to a 3 or 4 wire Oxygen Sensor to minimize the time at low Cold Start idle speed; or eliminate the Oxygen Sensor (Open Loop Mode only).

Separately, I had heard that older big six BMW L Jetronic engines tend to lean out a high RPM. At 3450 RPM, no load and in Closed Loop Mode I register 0.0 %CO and 14.8 AFM. At 3000 RPM and no load Open Loop, I register 0.14% CO and 14.6 AFM. So interestingly, Open Loop Mode has a both a stable idle and light throttle and seems to have slightly less lean out at high RPM.

If I only used Open Loop Mode I could use the Throttle Body Idle Adjustment to raise the Idle Speed to 800-850 RPM spec and perhaps also adjust fuel pressure up slightly to potentially reduce lean out at high RPM. Open Loop Mode seems a bit more powerful.

But I have to believe getting the Closed Loop Mode with the Oxygen Sensor to work properly would be the best solution with benefits in mileage and clean air (the car has no Cat) if I can figure out how to make this work.

I have attached graphs of AFM and % CO as RPM increases in both Open and Closed Loop Mode with no load. And also a graph of how RPM increases from Cold Start to Operating Temperature at Idle in Closed Loop Mode with Vacuum and AFM measurements.

My car is an e9 but I am posting here in appreciation of Peter Florance who's work among others introduced me to the L Jet system which has served me so well over these years.

[1975cs]

Here is an old post of mine from the e9 site on fuel pressure adjustments I had made previously. I think maybe I should go back to 2.8 bar and see how it affects my Closed Loop Idle issue. I also have an O2 sensor with heater on order…

Lean L Jet 3.5 Liter & Adjustable Fuel Pressure Regulator

My original 3.0 liter engine was converted to a 3.5 liter with a sport cam and ported heads by metric mechanic. It has a 1981 USA L Jetronic ignition from a 528i. I have low restriction stainless exhaust. Over the years I have tried a number of fuel pressure regulators trying to get the best balance of high end power with low end OEM idle.

With the OEM set up, at maximum throttle from a standing start through third gear, my engine would lose its power band above 5,000 rpm and seemed to be running lean at high RPM. I found that the rising fuel rate fuel pressure regulator from Bavarian Auto with pressure slightly higher than normal helped the high end but then the low end seemed suboptimal. I had the version that fastened directly on the fuel rail. Eventually the Fuel Pressure Regulator broke at its flare fitting and began leaking fuel.

After some research I read the the OEM fuel pressure was 2.5 bar. I had been running 2.8 bar. After doing some research I found the Saab 3.0 bar unit that is bolt on compatible. This is item Bosch 0 280 160 256 and is common on many older Saabs. The power output with this set up was wonderful. But I can be a bit of a perfectionist, and it seemed to run a rich on idle particulary before the 02 sensor was warmed. After some research and calculation i found that flow is a function of the square of the system pressure and a 3.5 liter should be about 2.8 bar versus 2.5 OEM for the 528i all else being equal. I identified a Saab part fixed at 2.8 bar, but this was Euro only, uncommon and no longer seems to be available anywhere.

I began looking for adjustable fuel pressure regulators, but wanted an OEM look. Bosch has the ideal solution in a unit sold by its Motorsport division (FPR 15-50). But it was 300 euros plus shipping.

I decided to go with the Bosch unit 0280 160 001 as modified by Mallory for vacuum pressure adjustment and sold as MSD 2222. I sourced this for $99. But it does not fit directly to the fuel rail as it has a barb fuel inlet. So I needed a barb adapter to fit the fuel rail's 17mm barrel nut bubble flare fitting. After some research I determined the internal flare thread size was 14M x 1.5. From this I found a couple of adapters. I went with a Parker 68HB-6-M114. To make this work I needed to shave down the barb height a bit as it is over sized for the BMW fuel injection hose. I also honed the bore to an angle to better fit the bubble flare. I also found a two piece adapter at Autozone by searching for 14 x 1.5 fuel injection adapter. It looked less sturdy, but in retrospect would probably have been a quicker install.

I love this new setup. 2.8 bar at a fixed pressure indeed seems perfect. I have a great wide power band to the RPM limit. And the idle is OEM perfect. Bosch's fuel pressure regulators are highly regarded versus aftermarket units and the whole set up was $108, one half to one third the cost of other adjustable solutions and only slightly more than the OEM fuel pressure regulator.

[1975cs]

Given my experience with this engine I decided my first tuning adjustment would be to raise the fuel pressure. I hoped that AFM measurements would confirm prior guesses based on feel that this was correct. I also I hoped this might alleviate surging at idle or low throttle and reduce risk of excessive lean out at high RPM.

I moved the engine to 2.8 bar fuel pressure (without vacuum on the fuel regulator). This was a positive step:

Surging was eliminated at low throttle or idle. This makes sense as the usual cause of this symptom is vacuum leakage which leads to leanness. I was happy to see that surging was gone in both Closed and Open Loop modes.

Lean out risk at high RPM seems to have been mitigated somewhat as shown in the attached graphs.

However there is (still) a large fluctuation in AFM at the start with a hot engine. As next steps, I will do some driving, evaluate cold start and install a narrow band Oxygen Sensor with a heater and retest start-up.

[1975cs]

As promised, I ordered and installed a Bosch four wire, heated 02 Sensor. Then I calibrated my Heathkit Air Fuel Mixture monitor. Here are the AFM results from cold start by minute, first at Idle, then at Higher RPMs, then back to Idle. Overall, I could not quite believe how much more stable the engine is with the heated oxygen sensor than it was with the unheated one. It doesn't look like I have any worry with excessive lean out at high RPM with this set up and settings.

The Cold Start idle is very stable in the 750-800 RPM range while the Hot Idle is stable at 1000. I still don't understand why the Hot Idle is faster, but I can live with these differences. Perhaps the Auxiliary Air Valve is not opening quite as much as it should when cold. In any event, I am in temperate California and the cold start is just fine.

The Cold Start Idle and Hot Idle both look a bit rich perhaps. I could try reducing fuel pressure from the 2.8 bar (increased due engine displacement at 3.5 liters), but I am rebuilding my brake booster and that may change vacuum levels somewhat, so I think I will stay with these and try driving for a while until after the new booster is installed.

[Mike W.]

Sorry to be so late to the party, I am impressed with your work on it.

One thing that jumps out at me is what injectors are you running? E12 injectors are rated at a pretty high flow so they only need the 2.5 bar pressure. And the E12 535i used the same injectors as US spec 528i's, so I suspect there would be sufficient capacity with the 2,5 bar regulator. Later US spec motronic 3.5 injectors are rated at much less flow which is surely part of the reason for a higher fuel pressure on E28 3,5s. Now E28 533s used similar spec, but a slightly different form factor than E12s. As far as leaning out at higher speed, I would think that would only be under load, and I've heard at higher RPM less fuel is actually needed per revolution.

Idle speed. 528i's were spec'd with a fairly high idle speed, 950+/- 50 IIRC. So if you're at 4 something you will likely have problems, especially at startup. You mentioned the AAV, which type are you using, water or electric? I've found the electrics to be borderline if everything is perfect, in terms of cold idle speed. There is a slight adjustment on them and many, many different variations within the same basic part. Both fuel pressure and timing are critical to a good cold and warming idle.

You reference the CO reading, but mention the B35 ex manifold, does it have the little stainless testing tube? That's where the CO measurement is supposed to be taken, not at the tailpipe.

Has the AFM been opened or tinkered with? Seems like they've all been opened at some time, and some adjusted in a belief that more fuel is better, which is not the case.

Personally I would work to get it to idle correctly when warmed up, at spec idle speed and go from there.