• Welcome, Guest! We hope you enjoy the excellent technical knowledge, event information and discussions that the BMW MOA forum provides. Some forum content will be hidden from you if you remain logged out. If you want to view all content, please click the 'Log in' button above and enter your BMW MOA username and password.

    If you are not an MOA member, why not take the time to join the club, so you can enjoy posting on the forum, the BMW Owners News magazine, and all of the discounts and benefits the BMW MOA offers?

Oilhead surging

I don't think power output vs mixture is that critical - especially in a minimum power situation. More critical is what the map optimization does to both cylinders when only one cylinder is dumping excess oxygen into the converter.

I gotta study & think about the rest of your post
NRP
Wait a minute! - I should have said what happens when one cylinder dumps unburned fuel into the converter. This would make both cylinders to quickly be remapped leaner - & probably more stumble.

At any rate, The mass flow to each cylinder vs the milli-volts out of the sole TPS has to be within a fairly small range - whatever that is.

Q - At a typical idle (i. e. against the throttle stops), what is the ratio of butterfly opening area versus BBS metering port area? I frankly don't expect anyone to know this, but that's what I'd like to find out without taking my R1100 down.

Another thought - since we are dealing with the very small angle of TPS motion, could it be that the TPS pot linearity at small angles has something to do with surging too? Film pots do wear, of course and are not particularly durable.
 
Wait a minute! - I should have said what happens when one cylinder dumps unburned fuel into the converter. This would make both cylinders to quickly be remapped leaner - & probably more stumble..

It would by more apt to say that the Motronic adjusts the Closed Loop mixture to bracket the lambda value of the O2 sensor. So if you lower the O2 sensor's lambda by, for example, 6% the the Motronic will adjust to that 6% richer mixture. You give the Motronic the new goal by installing the richer O2 sensor and the motronic does all the work.

At any rate, The mass flow to each cylinder vs the milli-volts out of the sole TPS has to be within a fairly small range - whatever that is..

My tests indicate that it resolves a smallest step of about 0.32 degrees. When cold the Motronic adds enough extra fuel that accuracy doesn't seem to be an issue. When warmed up, closed loop learns exactly how much fuel is needed.

Q - At a typical idle (i. e. against the throttle stops), what is the ratio of butterfly opening area versus BBS metering port area? I frankly don't expect anyone to know this, but that's what I'd like to find out without taking my R1100 down..

I don't think I've read any test of it. But we get idle air and 2000 rpm air (vacuum) pretty close. Just switched from twinmax to harmonizer today's. what a good upgrade.

Another thought - since we are dealing with the very small angle of TPS motion, could it be that the TPS pot linearity at small angles has something to do with surging too? Film pots do wear, of course and are not particularly durable.

Two things: The TPS pot has two wipers. One goes from 0 to 23 degrees the other covers 20 to 81 degrees. Once the throttle is steady, closed loop learns the small errors and saves them for open loop fueling too.

In summary,

I think that air balance and fuel balance are the two principle issues. The air balance can be adjusted, but the fuel injector imbalances can't. So you got two options: clean and match the injectors or richen the mixture a bit.
 
In summary,

I think that air balance and fuel balance are the two principle issues. The air balance can be adjusted, but the fuel injector imbalances can't. So you got two options: clean and match the injectors or richen the mixture a bit.
Why not balance the air measurement system? I think we have to quantify the BBS vs throttle stop screw tradeoffs first. Your data seems to me to be based on assumptions that the air supply system is essentially equal on the two sides. I don't think it is.

Again the call - does anyone have a spare throttle body?
 
Why not balance the air measurement system? I think we have to quantify the BBS vs throttle stop screw tradeoffs first. Your data seems to me to be based on assumptions that the air supply system is essentially equal on the two sides. I don't think it is.

Again the call - does anyone have a spare throttle body?

Of course you would balance the air intake, it is one of two necessary step to getting equal power side to side. Several things are related to the Volumetric Efficiency (VE) which is actually what you hope to balance. Just because the intake manifold vacuum is equal does not guarantee that VE is equal. VE has to do with Intake Manifold dynamics, throttle body/throttle plate/BBS) flow, intake valve/combustion chamber/exhaust value timing and performance, exhaust pressure at the moment of exhaust valve closure. Exhaust pressure is interesting as the shape of the exhaust plumbing is different between the two sides. All that said, based on the smoothness at many loads and RPMs, the VE vs RPM/TPS on my motorcycle seems pretty balanced.

Once you have done that, add enough fuel, a few percent, so that you move to the point on fuel vs power curve to be nearer best power. Once you do that, inequalities in fuel flow are much less significant. Looking at the chart below (the thick red vertical line is about where our bikes run, stoic or peak EGT), second set of curves up from the bottom, you want to be nearer point #2 than point #3. At point # two, changes in fuel flow don't change the power outpout. Nearer point 3, the power changes a lot with AFR and it gets worse, the more to the right (leaner) you go.

If you look at the end of my Wideband O2 thread, wsj04oh has noticed that at small throttle angles and rpms above 3500 RPM that the Motronic is Open Loop even with steady throttle. He has also noticed that the fueling is leaner there than expected. This is also an area where his bike was prone to surge before he added his LC-1. Because he is riding and observing AFR, not just inferring what may or may not be happening, we have another data point suggesting that rough running is related to leanness.

pt1fig2.jpg
 
I agree. I have 170K on my 94 RS, still on the original injectors, and I have never had or developed a surging problem. But, I have been using Chevron Techron fuel cleaner (the only one I feel really works well) for well over ten years. I go through at least two bottles of it per season. Never had a fuel system related problem on my RS.
 
I agree. I have 170K on my 94 RS, still on the original injectors, and I have never had or developed a surging problem. But, I have been using Chevron Techron fuel cleaner (the only one I feel really works well) for well over ten years. I go through at least two bottles of it per season. Never had a fuel system related problem on my RS.

Not really surprising that yours is fine. I think it is a situation where some do and some don't surge, exactly what you would expect in a stack-up-of-tolerences of air and fuel.

There are some other benefits of running a few percent more fuel. The biggest of which, in my mind, is the gain in torque between 2000 and 3500 RPM. First let me say that I think that you should only add the fuel you need 4-6% to improve driveability. That said, I installed a WB O2 that I'm evaluating and although it is set at lambda of 0.94 (6% more fuel), it has some issues and the lambda measured is 0.91 (about 9% more fuel). I don't have any intention of running my bike at 0.91 but, wow, does it run nicely.
 
I agree. I have 170K on my 94 RS, still on the original injectors, and I have never had or developed a surging problem.
ANDYVH - What is your bike's present configuration? What sort of throttle body history does it have?

Assuming it is stock (which it sounds like few are), I wonder if perhaps the nature of growing parameter discrepancy between air flow calculations on the two sides is completely random. If so, maybe on half the oilheads, if the right cylinder (the one w/o a TPS and independent oxygen sensor) happens to evolve rich (not lean) if then they do not develop a surge/stumble?
 
Of course you would balance the air intake, it is one of two necessary step to getting equal power side to side. Several things are related to the Volumetric Efficiency (VE) which is actually what you hope to balance. Just because the intake manifold vacuum is equal does not guarantee that VE is equal. VE has to do with Intake Manifold dynamics, throttle body/throttle plate/BBS) flow, intake valve/combustion chamber/exhaust value timing and performance, exhaust pressure at the moment of exhaust valve closure. Exhaust pressure is interesting as the shape of the exhaust plumbing is different between the two sides. All that said, based on the smoothness at many loads and RPMs, the VE vs RPM/TPS on my motorcycle seems pretty balanced.

Once you have done that, add enough fuel, a few percent, so that you move to the point on fuel vs power curve to be nearer best power. Once you do that, inequalities in fuel flow are much less significant. Looking at the chart below (the thick red vertical line is about where our bikes run, stoic or peak EGT), second set of curves up from the bottom, you want to be nearer point #2 than point #3. At point # two, changes in fuel flow don't change the power output. Nearer point 3, the power changes a lot with AFR and it gets worse, the more to the right (leaner) you go.

If you look at the end of my Wideband O2 thread, wsj04oh has noticed that at small throttle angles and rpms above 3500 RPM that the Motronic is Open Loop even with steady throttle. He has also noticed that the fueling is leaner there than expected. This is also an area where his bike was prone to surge before he added his LC-1. Because he is riding and observing AFR, not just inferring what may or may not be happening, we have another data point suggesting that rough running is related to leanness.
If I read you right, your contention is that overall leanness is the cause of surging & that shifting the target Lambda a few percent is the fix. If so, it would seem like both cylinders are participating in the instability via the closed loop mode optimization.

My contention is that only one cylinder is dropping out during our surges, and that stable operation off idle should be achievable with about 14:1 A/F ratio - if it could be made similar on both cylinders.

Does Volumetric Efficiency have a formal definition?

Our Lycoming powered 172M could go a couple hundred degrees lean of peak without getting into any serious stumble. Of course that with a carb and single EGT probe, but it was at ~55% power. I think here we are dealing with misfire in a near over run situation which could be much different from aircraft engine applications around cruise power.
 
Thank you for bearing down on the points I made earlier and testing them.

If I read you right, your contention is that overall leanness is the cause of surging & that shifting the target Lambda a few percent is the fix. If so, it would seem like both cylinders are participating in the instability via the closed loop mode optimization.

Not what I'm contending. Let me go over it again.

I think there are a lot of factors that can contribute to driveability issues in lean-fueled vehicles. Based on a year and a half of data recording and research, left/right power imbalances are
one cause.

For the power to be balanced, there must be equal VE (volume of air and oxygen) in each cylinder and their must be equal air to fuel ratios left to right.

If the left/right AFRs are unequal, it is well known from experience and experiment that the leaner you are relative to the Best Power Mixture, the more susceptible an engine is to roughness, stumbling and vibration (or one form of surging).

No matter how precisely you balance VE, if there is a fuel imbalance and the overall mixture is lean, you are likely to feel it.

You can make an engine much less susceptible to unequal fuel delivery if you balance VE, and richen the mixture toward Best Power.

My contention is that only one cylinder is dropping out during our surges, and that stable operation off idle should be achievable with about 14:1 A/F ratio - if it could be made similar on both cylinders.

At an AFR of 14:1 you are about 5% richer than stoic and nearer Best Power Mixture. If you balance air you will and richen to 14:1, you will get a smooth running engine. I would agree with that.

Looking at the hundred or so hours of test data that I have, there is no data showing one cylinder dropping out or misfiring. You should set up some equipment, measure this and show the data.

Does Volumetric Efficiency have a formal definition?

Yes, you can easily find it by Googling: Volumetric Efficiency

Our Lycoming powered 172M could go a couple hundred degrees lean of peak without getting into any serious stumble. Of course that with a carb and single EGT probe, but it was at ~55% power. I think here we are dealing with misfire in a near over run situation which could be much different from aircraft engine applications around cruise power.

My Continental powered Beech A36 Bonanza could go somewhere in the vicinity of 100 degrees lean of peak and not stumble. But you could lean it to the point where there was no mis-fire but the type of roughness mentioned by Gami in their articles.

As I mentioned earlier, wjg04oh has taken data with his LC-1 that shows there are regions of steady throttle operation (e.g. 3rd gear, 4000 RPM, 7 degrees TPS) where the Motronic is Open Loop and quite lean. I believe this is a candidate for further experiments.

It seems to me that you're trying to convince yourself that if the air can be balanced, that fueling differences don't matter much. That is only true near Best Power Mixture. Lean of Peak EGT or in our case Leaner than Stoic (14.7:1 for gasoline) our boxers, and many other engines, are susceptible to rough running.

I have lots of measurements at this point that show if you balance air (VE) and richen the mixture 4-6%, you get quite a smooth engine. And you get a secondary benefit of a nice boost in torque in the 2000-3500 RPM range. Give it a try.

RB
 
Roger - I only started to wade thru the 10+ pages of plots etc in yr earlier posts but couldn't yet see where you can identify which cylinder is missing - (or if they both are). Can you direct me?
THX
Niel
 
Ok, here is how my 94 RS is set up:
Been using a K&N air filter since I had the bike at 3,000 miles on it. I ran the stock exhaust and fuel system, with Autolite 3923 plugs for years. No issues. Now, my bike is a late 93 build which had NO CAT-Code plug at all. Those early Oilheads ran just a bit more rich. Plug wires, coils, fuel pump, injectors, fuel regulator are all stock, original and untouched. I replaced the fuel filter, finally, at 120k and it looked pretty good when I took it apart.

Then in about 04 the muffler outlet tube cracked (common to some early Oilheads). I had it TIG welded, but noted more "popping" on decel when it was cracked. Then I got another cracked muffler for nothing and my curiosity led me to cut the can open, study it, make my own internals and TIG weld it up. Looks stock, but deeper tone and breathed a bit easier enough to cause more decel popping (it is not backfiring as many incorrectly claim it). To counter that I added a Techlusion. Balanced the TBs myself and got good results of 45 mpg average and clean plugs.

Three years back I rebuilt the TBs myself using the Bing kit. The ticking sound on the TB caused by shaft wear made it harder to balance. After I rebuilt both TBs and installed new throttle cables, I again balanced the TBs using my basic old mercury sticks. Nice and smooth again. Bike is still on the original untouched fuel injectors. Though, again out of curiosity, I am taking the injectors to a local diesel injector repair shop to have them flow tested (curious about the current state flow matching), then cleaned and flow matched again.

But aside from all this, I never had a surging issue. I certainly read up on it when I first got the bike in Oct 94, but can't say I ever experienced it at all, or enough to cause any concern.
 
Ok, here is how my 94 RS is set up:.......

.....But aside from all this, I never had a surging issue. I certainly read up on it when I first got the bike in Oct 94, but can't say I ever experienced it at all, or enough to cause any concern.
THX! I'm more concerned about ways to make the R bikes perform as they did when new and as configured when new.

You have answered the problem with a Techlusion. It would be interesting to swap the fuel injectors (L-R) but I don't expect you to do that.......................

NRP
 
Roger - I only started to wade thru the 10+ pages of plots etc in yr earlier posts but couldn't yet see where you can identify which cylinder is missing - (or if they both are). Can you direct me?
THX
Niel

I've only posted about 10% of the data I collected over the past year and a half. I've laid out my case with detail and supporting documents. Since this is the first time that ordinary riders can both program lamda and record the results of combustion, there is still more insight to be gained. I'll continue to post what I find and try to explain it.

You obviously have a theory, you are skeptically-minded (me too, nothing wrong with that) so you should really try to plan experiments, log data and present it and see where it takes you. With regard to your earlier questions, my post here: http://forums.bmwmoa.org/showthread.php?67397-Oilhead-surging&p=866626&viewfull=1#post866626, is about the best I can do to explain what I've seen. Since a wideband sensor is one of the most accurate measurements of combustion, I'd be happy to help you set one up on an 1100. You could be one of the first to have definitive data on how the MA2.2 behaves with no Coding Plug installed. There are lots of theories but few measurements.

RB
 
Last edited:
...

To counter that I added a Techlusion.

...

A Techlusion is one way to richen the mixture. On many bikes you have to give up closed loop operation when you install it and I find it hard to tell how much fuel you're adding. But it can work.

Some of the early 1100 that had a catalytic converter and O2 sensor coded the Motronic by hard wiring in the harness.

Do you run an O2 sensor?
 
Not any more. My RS did its part for "green" running for over 15 years. When I installed the Techlusion I disconnected the O2 sensor.
 
Not any more. My RS did its part for "green" running for over 15 years. When I installed the Techlusion I disconnected the O2 sensor.

There have to be a lot of 20 year-old cats and oxygen sensors that have given up the ghost.

Just curious, do you have a socket for a Coding Plug with none in it or not even a socket? I remember reading that some bikes had the Motronic coding done in the harness but can't find that in my BMW electrical schematics disk.
 
I don't know how one goes about detecting a misfire with instrumentation - especially which cylinder, short of having pressure transducers in each spark plug port. Maybe with the new lo cost solid state ones? It would be a very revealing piece of data. At least we only have two cylinders....:)

Also - I wonder how the air mass flow rate is calculated by the Motronics. I presume an algorithm based on throttle position pot & rpm (and eventually air temp and altitude etc) since there is no manifold pressure transducer. Is there precedence for this type of abbreviated engine control?

At this point I think it is easiest to look at the flow characteristics of actual throttle bodies (including dirty ones), where there is a very very big valve (the throttle) flowing in parallel with a very small manually adjustable one (the BBSs). That's why I keep suspecting the equality assumptions.

But I agree that anything that could identify a misfire would be a big help.
 
No CAT-code plug, and no socket or wiring for it even if I wanted to install one.

When the Oilhead first came out, Bob Lentini published an article about tuning tweaks on the new engine to boost the midrange torque for the US market. One of the tweaks was to install the Pink versus the Yellow (I think) CAT-code plug. Doing that richened up the fuel to air mix slightly. So I rushed to the dealer to buy the $17 plug. Only to find out when I got home with it, I had no plug at all and no where to put it. So it is resting for 19 years in my parts stash.
 
+1 on the Booster Plug--just installed one this morning and I have to say that I would not have recognized my bike. No more surge, much more power, smoother idle. a great machine before, wonderful now(1994R1100RS with catalytic converter)
Tony
 
I don't know how one goes about detecting a misfire with instrumentation - especially which cylinder, short of having pressure transducers in each spark plug port. Maybe with the new lo cost solid state ones? It would be a very revealing piece of data. At least we only have two cylinders....:)

If you install a Wideband O2 sensor and run it in instantaneous mode you might be able to record misfire. Is show up as a lean spike like in the chart below. I have not detected this in any of my plots, even when I leaned closed loop lambda to 1.06 (AFR 15.3:1). The bike will stumble at times with that AFR but not misfire.

Also - I wonder how the air mass flow rate is calculated by the Motronics. I presume an algorithm based on throttle position pot & rpm (and eventually air temp and altitude etc) since there is no manifold pressure transducer. Is there precedence for this type of abbreviated engine control?

Below is a draft block diagram I'm working on for the MA2.2. The yellow highlighted boxes need further investigation but the overall picture is, I believe, a pretty good representation.

As an alpha/n type ECU, it starts in a VE table with an air mass estimate based on TPS/rpm, with interpolation for TPS/rpm conditions that are between cells. Next it corrects for air temperature and pressure. Then it measure battery voltage to correct the fuel pulse which does vary with voltage. From my measurements this gets it in the ballparkÔÇöa few percent on AFR. Finally, there will be some type of adaptation value correction.

The closed loop program is very accurate and does the fine tuning. When in closed loop it can also compare the table driven result with the closed loop result to create an adaptation value that I mentioned above. I have not measured the R1100 so I can't say how extensive these adaptation values are. But some information on Anton's site describes their existence.

RB

Caravanmisfires-1.jpg


motronic2.2.jpg
 
Back
Top