Closed-loop vs Open-loop

[RacerX]

Folks,

 

The 02-sensored bikes, including the Euro-Californias, World-Spec Brevas, BMWs and the new MZ1000S (which I just had the protoype PCIIIusb built for a few months back), are only partial closed loop. Rather then me state my opinion... I saved a posting from Dynojet U.S.' Director of Fuel Injection Development. His words below summarize why the technology simply isn't there for motorcycle/high-performance closed loop mapping. Below address' the new-to-be-released partial closed loop PCIIIusb for BMWs.

 

Enjoy - Todd

p.s. The PCIIIusb allows individual cylinder mapping on the dyno for up to 6 cylinders... you have to experience it to believe it.

 

"Gentlemen,

There seems to be quite a buzz concerning the new BMW Wideband Power Commander and its possible use on other motorcycle models. There are a number of reasons why the unit was designed for the BMW models the way it was and many reasons why we do not use this setup on other bikes.

 

Let us start with why the BMW uses a wide band O2 sensor as part of the unit. The bike already has a "closed loop" circuit as part of the OEM injection system. It does not "auto map" the entire rpm/throttle position range of the fuel map. Generally speaking, the closed loop system only adjusts the fuel curve below @40% throttle. Above that the system is "open loop". The new Wide Band BMW unit only controls the stock "closed loop" area. Outside of that the bike is mapped in the normal fashion, on the dyno.

 

We would actually prefer not to maintain the closed loop section. Due to the design of the OEM injection system it is not possible to bypass it as we do with other models. Closed Loop systems are not the "magic" that most people believe they are. There are a number of problems that keep it from being the best choice for high performance applications.

 

This brings us to why we do not use this type of system on other applications. Unlike car systems that have a mass airflow sensor, motorcycles cannot monitor the amount of air ingested (as a side note, most automotive application are also not fully closed loop systems). O2 sensors only read the amount of residual oxygen in the exhaust and calculate the air/fuel ratio based on that data. The problem is that the reading at the O2 sensor has already occurred in the engine. This is not a problem when making adjustments during steady throttle riding while on the highway as there is adequate time to change the mixture and then check to make sure it is correct. During times of aggressive throttle movement, both on and off throttle, this data is useless and an "open loop" system must be used.

 

Things such as installing a sensor into your bike, the cost of the sensor (and its replacement), the lack of individual cylinder control, disabling of "air injection" systems, etc., all play a part in the why a Wide Band Power Commander is not practical for all applications. For many, simply finding a shop to weld on a fitting for the O2 sensor onto their brand new chrome exhaust system without damaging it is a problem.

 

A big problem is that if all fuel adjustments are based on the O2 sensor reading items such as an exhaust leak, "air injection" systems, poorly located O2 sensors, and faulty or degraded O2 sensors can have the bike running very poorly. Short exhaust system are notorious for having reversion issues (air pulsing back up the exhaust) which then cause false readings leading to overly rich mixtures.

 

Since the BMW was designed from the start to have a closed loop layout and exhaust styles do not vary to any great degree it is not difficult to implement our Wide Band technology. On other bikes simply adding an O2 sensor is not the best choice on tuning. From a cost standpoint it is less expensive to purchase the standard Power Commander and have a custom map (individual cylinder mapping for best performance) made than to buy the more expensive Wide Band system and have a fitting installed into the exhaust.

 

On the dyno all of the changes can be double checked to make sure that the performance is at its peak. Improper adjustments will be corrected before they cause a running problem out on the road. While at first glance it would appear that a Closed Loop system is the answer to all tuning questions, in actual practice it is just not that simple.

 

Regards,

Michael Belcher

Director of Fuel Injection Development"

[Guest Brian Robson]

Buy Michael a beer please.. .I think I finally understand the difference between open loop and closed loop.

An equally simple explanation of individual cylinder mapping and how it can be done would be a bonus.

[callison]

An equally simple explanation of individual cylinder mapping and how it can be done would be a bonus.

 

Consider two single cylinder ECU's, each with an optimized map for their specific cylinder. The only thing really in common is the housing, the timing signal and the power supply.

 

Simple enough?

[Guest Brian Robson]

But wasn't he talking about a single PC111 allowing indivdual mapping of two or more cylinders?

[beauchemin]

Yes. But the PCIIIusb will only allow individual cylinder map adjustments if the ECU already has individual cylinder mapping.

[RacerX]

Yes.  But the PCIIIusb will only allow individual cylinder map adjustments if the ECU already has individual cylinder mapping.

 

Totally not correct... PCIIIusb individual cylinder mapping has nothing to do with the OEM ECU. It does so independently of the ECU.

 

Brian, as Carl stated... the PCIII will "store" a "map" for each injector/cylinder. Dynojet originally experimented with using one PCIII for each cylinder, and when they saw how well it worked... it was then adapted such that one PCIIIusb will store many maps, one for each cylinder. Ind. cyl. mapping is done on the dyno using Tuning Link for optimum refinement for each (your) bike. Think of it as the ultimate throttle body "balancing".

 

ToddGuzziTech.com

Certified PCIII Tech

[Cliff]

Sorry I tried to ignore this but anyway.

 

First off I'd like to state that there MAY be some bias in the above. Obviously Dynojet have a vested interest in ECUs not doing the right thing but enough of that I'd like to keep on topic.

 

As an electronics engineer with control system experience my interest is from the technical view point. I don't have a interest in maintaining a belief of one system being superior to another and recognise that all systems are a balance of compromises and assumptions. In fact most the my units are run open loop.

 

Firstly, yes it is important to make sure your o2 sensor is giving the correct readings. It also important that all other sensors ( air temp, air pressure etc ) are correct. In fact the our standard "open loop" systems are in fact "closed loop" in a way as they respond to these other sensors and modify their output. The new ECUs are really only adding an addition sensor to the mix.

 

Secondly its a gross oversimplification to treat any 2 closed loop controllers as the same. "Closed Loop" can be done in so many ways.

 

Technically a real closed loop ECU would have no map at all. All its outputs would be based purely on sensor inputs. I don't think there are any controllers of this type out there.

 

The bulk of closed loop ECUs would be operate from a map just as an open loop controller. They would than perform small adjustments based on the sensors. The variation between the controllers would come from a number of areas

1) How quickly they respond to the sensors

2) How much compensation they apply to the map

3) When and under what conditions they decide to be open loop or closed loop

4) Whether they learn ( correct the map over time )

 

The list goes on but probably the most important

 

N) what is the manufacturer's aim in putting in a closed loop controller? Is it to extract the optimal performance from an engine. Perhaps if its the Honda Race team. Its more likely an attempt to reduce the emissions to pass the EPA requirements. This is not optimal as far as the rider is concerned. This in itself will create an impression of closed loop controllers being less than open loop controllers.

 

One question about the PC tuning on the dyno. Are the adjustments to the map derived from the gas sensor?

[Guest Brian Robson]

Thanks Todd for your explanation, and Cliff for confirming why I am not an electrical engineer.

The depth of knowledge on this site is overwhelming at times and I can only chirp in with the odd observation from the frequent miles and maintenance perspective.

[JuhaV]

One question about the PC tuning on the dyno. Are the adjustments to the map derived from the gas sensor?

37937[/snapback]

 

Very good question indeed !

 

I will also throw in here my "old" question : How do you adjust your map correctly in the mid range (partial throttle) if you are using an inertia type (not brake type) dyno ? I still have the impression that inertia type dynos simulate mainly the full throttle acceleration situation. Please advice if this is not correct interpretation.

 

BTW, more advanced ECUs allow the user to decide what parts of the map are used in open-loop or closed-loop mode. The most advanced ECUs even allow to key in different lambda targets for different parts of the map.

 

br, JuhaV

[al_roethlisberger]

Here's a practical question for you all:

 

My old FJ1200 was carbureted, and as such, during extreme altitude changes... it's relatively "fixed map"(jets/needles) would have issues running once the atmospheric inputs varied outside its supported extremes. For example, up in Rocky Mountain National Park above Estes, CO... it barely ran.

 

However, my friend's '98 Honda VFR, with EFI(I have no idea if closed or open loop), would run like a champ as it apparently adjusted automatically to the changing air density.

 

 

With our OEM ECU incorporating an air pressure sensor(the black knob on the ECU case, from what I understand), does the V11 EFI system have the capability to run well in this scenario?

 

Not having ridden my bike at extreme altitude, except up over Yosemite a couple years ago(where it did run fine BTW), I don't have any experience on the V11 in this application... and am curious if it will behave like the FJ or the VFR?

 

al

[luhbo]

I will also throw in here my "old" question : How do you adjust your map correctly in the mid range (partial throttle) if you are using an inertia type (not brake type) dyno ? I still have the impression that inertia type dynos simulate mainly the full throttle acceleration situation. Please advice if this is not correct interpretation.

....

br, JuhaV

37948[/snapback]

 

I think it's just a question of the inertia the dyno has. I have never seen one in action. If it takes two seconds to redline the bike, then it could be quite useless for everything else but full throttle. If it takes 20 seconds, then it should work like a static one and give similar results.

You should also take in account how fast your ecu can store the results or how fast your lambda-probe reacts and so on...

If you ask your dealer, the answer probably depends on what system he has.

 

Hubert

[luhbo]

The ecu has an inbuilt correction mechanism. The pressure sensor will affect the applied correction factor. But wether this and other influences are correct could only be verified by running the ecu in closed loop.

 

I remember having read somewhere, that Marelli f.i. uses first hand the map values, but then when the engine runs static the ecu begins to lean out the mixture to reach the values the traffic authority likes. But I am not sure.

 

Some folks also think you cannot change your original mapping without switching off the O2 sensor, because the ecu seems to notice that something has changed and also seems to be "learning" to reduce the mapped fuel amount to reach the correct values again!

 

The point is always: they tell you so or so and then offer you the best solution the market has, what is surprisingly always their system.

 

Hubert

[roberto]

al,

 

I have climbed to 8000/9000ft in the Alps and noticed no reduction in performance while my companions on various japanese motorcycles reported problems (I did cook the brakes going back down tho )

 

However, I did similar journeys on my 1988 Le Mans and I don't recall having any problem with that either

 

Could be the big block doesn't mind running a bit rich.

[beauchemin]

Totally not correct... PCIIIusb individual cylinder mapping has nothing to do with the OEM ECU. It does so independently of the ECU.

 

ToddGuzziTech.com

Certified PCIII Tech

37936[/snapback]

 

Sorry what I meant to say is that since the PCIII isn't adding wiring, cylinder mapping can only be possible if the ECU has separate injector trigger wires.

 

Chris

[RacerX]

Sorry I tried to ignore this but anyway... First off I'd like to state that there MAY be some bias in the above. Obviously Dynojet have a vested interest in ECUs not doing the right thing...

 

Heh, I felt the same way about the My15M posts, hence a seperate post here... and yes you are right, much like everyone else trying to "sell" something to make bikes run "better".

Dynojet does a tremendous job in building a product that works... quite easily for most folks, and the fact that I witness some of the fastest race teams here in the U.S. use PCIIIs at every AMA Superbike race, has pretty good merits.

Cliff, I'll continue to say that you deserve high praise for your work in building your own ECU. I look forward to being able to test one of your units on the dyno.

 

One question about the PC tuning on the dyno. Are the adjustments to the map derived from the gas sensor?

 

They use a ($2500usd) gas analyzer, often on the factory bungs that the OEM uses (on Guzzi, etc.) to pull from. All technicalities aside, it works very well.

 

How do you adjust your map correctly in the mid range (partial throttle) if you are using an inertia type (not brake type) dyno? I still have the impression that inertia type dynos simulate mainly the full throttle acceleration situation. Please advice if this is not correct interpretation

 

You are correct. You cannot properly map on an inertia dyno. PCIII mapping is only by Eddy current load-cell, air/fuel dynos.

 

ToddGuzziTech.com