Thought about Piggyback tuning ECU (GET, PCV, Bazzaz, Powertronic...)

[chevaliernoir]

For racing purpose, max torque/power is of course a key parameter, and piggyback ECU as PCV, GET, Powertronic, Bazzaz is good for us.


For road usage, we are not interested only in full load, but also partial load.
Removing the calatyst converter, we don't care anymore to be at lambda 1, and we would prefer to tune the engine with the best richness for engine stability, for example, (or consumption if I want...)
Then I feel that piggyback is "dangerous" :

If we increase richness (for example +10%) in the closed loop area, to improve drivability or reduce temperature maybe, then the closed loop control of the ECU will slowly compensate by applying -10% for example.


Even worse : what is learnt by the ECU in closed loop conduitions, is ofter applied also in open loop condition. Then the ECU will apply also -10% in full load.
So the overall effect of +10% richness in partial load could become in fact -10% in full load!


Even worse, as the learning and adaption of the ECU takes some time (typically ~10 minutes in stabilized closed loop condition with engine hot).
So if we tune the piggyback ECU at dyno and test it, we have the expected befavior and we can be proud and happy of it... at the moment
But at the end, after real usage, all the gain could be lost, and the behavior could be worse than the original one.


This is my fear I have about piggy back tuning method that keeps the closed loop active. (is there a system that can de activate the closed loop control?)


I know very well the standrad ECU mounted on bikes and the strategy as this is my job, but I don't know the Bash one used on RC390, and don't have any experience on tuning these ECU, so I could be wrong, and I'm interested in your feedback about it.

Thank you

 

[FSTASFCK]

Isn't that the purpose of the WB2 from dynojet?

For racing purpose, max torque/power is of course a key parameter, and piggyback ECU as PCV, GET, Powertronic, Bazzaz is good for us.


For road usage, we are not interested only in full load, but also partial load.
Removing the calatyst converter, we don't care anymore to be at lambda 1, and we would prefer to tune the engine with the best richness for engine stability, for example, (or consumption if I want...)
Then I feel that piggyback is "dangerous" :

If we increase richness (for example +10%) in the closed loop area, to improve drivability or reduce temperature maybe, then the closed loop control of the ECU will slowly compensate by applying -10% for example.


Even worse : what is learnt by the ECU in closed loop conduitions, is ofter applied also in open loop condition. Then the ECU will apply also -10% in full load.
So the overall effect of +10% richness in partial load could become in fact -10% in full load!


Even worse, as the learning and adaption of the ECU takes some time (typically ~10 minutes in stabilized closed loop condition with engine hot).
So if we tune the piggyback ECU at dyno and test it, we have the expected befavior and we can be proud and happy of it... at the moment
But at the end, after real usage, all the gain could be lost, and the behavior could be worse than the original one.


This is my fear I have about piggy back tuning method that keeps the closed loop active. (is there a system that can de activate the closed loop control?)


I know very well the standrad ECU mounted on bikes and the strategy as this is my job, but I don't know the Bash one used on RC390, and don't have any experience on tuning these ECU, so I could be wrong, and I'm interested in your feedback about it.

Thank you

 

[cornerslider]

Chevalienior- If I'm "hijacking" your thread, I sincerely apologize........ I see you post quite a bit, and I know you work in the FI industry. I don't know that much about fuel tuning. I know how to set-up a fuel tuner at the "base" level on a variety of bikes, get minimal gains, and not wreck anything - I'm one of those guys that wants to be good at electricity/electronics, but I seriously lack the necessary skills . I have a 2016 RC390, with a PCV, Arrow exhaust, with a stock ECU, Chad Wells map, open airbox mod, and K&N filter. My bike runs great, and I could not be happier. If my PCV were to "fail", what would happen? My understanding of a "piggyback" system (PCV) is that it works in tandem with the stock ECU. Would I gain anything by getting my ECU "flashed" to the Akropovic/Cup map? I'm not racing/chasing championships.... I just want to be in the middle the pack "on-track", and not wreck anything. I've heard the term "open-loop", and "closed-loop", but not really sure what that means???? Could you "dumb -it-down" for me (and maybe others on this site) as to what the difference is between open-loop, and closed-loop is? Any help would be greatly appreciated. Thanks-

 

[chevaliernoir]

Hello.
Let's try to be focus on the point with just the details needed (and some shortcut to simplify) :

In order to reduce emission (HC, CO NOX...) the catalyst is fantastic and very very efficient.
Catalyst is the one that reduces emission. Not the injection system.

The 2 issues with the catalyst is that
- it is efficient only when it is hot (so the first minute of usage could generate more emission than the following hour when it is hot)
- it needs to have exactly the stoichiometric richness. If we are slightly richer, then we will need to be slightly poor just after to compensate.
That's why the carburetor is dead : it was not able to guaranty this perfect mix ration Air/Fuel in all conditions.

About the Injection system (EFI), so the target is to be around stoichiometric.
This cannot be calibrated one time at KTM (or other) because with the aging, the air filter dirtiness (?), the valve gap change in time, the atmospheric pressure, the variation from bike to bike... the quantity of air that enters is not always perfectly the same. So the quantity of fuel to be injected has to be corrected in real time.
In order to do so, there is the Lambda probe in the exhaust. This sensor is able to tell to the ECU is the mixture is too rich or too poor, respect the ideal stoichiometricratio needed the catalyst. No more.

So the ECU has a base calibration, from the factory, stored. It uses it to define the injection quantity.
In certain conditions (partial load, engine hot, stabilized TPS or RPM...) it check the info from the lambda probe.
If lambda says "too poor", then the ECU add some fuel, little by little, at each engine cycle, until the lambda says "too rich". Then the ECU decreases the fuel quantity slowly until "too poor" and so and so.
That's how the ECU can guaranty we are always around the best Air/Fuel ratio.
That's also why the lambda signal is like a square going from rich to lean.
That's also why it's a non sense to simulate the lambda signal if it is not phased with the ECU strategy that goes richer/poorer.
So the ECU is learning how much fuel it must add or remove from the base calibration it has from te factory, and it stores this value inside, and update it at each run.

Usually, we need something like 10 minutes of idle in hot condition, or ride in stabilized situation so the ECU moved from the calibration stored in factory and the adapted one.
This is the closed loop way of working : the lambda info is used in order to adapt the fuel injected


This closed loop is used only when engine is hot, in stabilized condition (not in cas of sudden throttle opening or closing), and only in partial load.
When we are in Wide Open Throttle, the ECU does not try the have the catalyst working and inject the fuel needed for power, and to have a safe condition for the engine.
So the injection time is the value stored in the ECU at the factory, and it applies the correction it learn in partial load probably. We say we are in "open loop" as we don't use at this time the lambda information.

I hope these few lines clarify for you what Closed Loop is, and then will give more sense to my initial message

 

[CDN Duke]

I hope these few lines clarify for you what Closed Loop is, and then will give more sense to my initial message

That's helpful, thanks for the explanation.

 

[streetfighter]

is it possible to set a constant AFR value within the stock factory ECU? isn't this a better solution since the bike operates in closed loop anyway when it is below 80% of full throttle?


Sent from my iPhone using Tapatalk

 

[KTMGene390]

is it possible to set a constant AFR value within the stock factory ECU? isn't this a better solution since the bike operates in closed loop anyway when it is below 80% of full throttle?


Sent from my iPhone using Tapatalk

I believe it is possible to reprogram the stock ECU, but a simpler solution is Dynojet's WB2 along with a PCV which is what I'm using. The WB2 can be programmed using only a LCD200 which is discontinued so you would either have to locate one or buy from a vendor that will program it for you, it does not come programmed with a AFR value. If you just plug it in it will do nothing in the closed loop area until program and cannot be accessed via the PCV software other than enabling Auto-Tune and setting values for open loop tuning.

The WB2 can also be used as a standalone AFR monitor with a gauge hooked up and/or for closed-loop tuning, but you will need a PCV if you want to use the auto tune feature.