resetting tcu/tps?

ok, I've heard a lot of people complaining about how their tcu has learned how the previous driver drove. or it's finally learned how you drive. Is there a button or something that can reset the whole system so it can learn from scratch?

I think you just disconnect the battery for overnight.

[BLKSVX92]

I think you just disconnect the battery for overnight.

Yep , that's all you have to do

[AutoTchr]

Try pulling the fuse if you don't want to have to reset your clock. or other electronics.

[Beav]

I've tried to clear this up for everyone before, and I guess it was missed it or it just didn't sink in or...

'Learning' type computers in cars does not mean they adapt to the driver. "Gee, it looks like Bob wants to be a tightwad today so I'll adjust myself to give him excellent mileage" is not a parameter in the computers coding. A learning computer is one that can adjust it's programming to allow for sensor deviations, etc. For example, if the engine coolant temperature sensor should read .xx volts when the engine has run xx amount of time and has seen the throttle at xx% vs. xxxx rpm with xx ambient air temp and xx, xx, xx ..... and the temp sensor is sending a signal a few .ox volts off from what's expected, the computer can 'learn' the new parameter and adjust for the change. As electrics age they normally tend to increase resistance and thus their expected values will change.

Maybe this explanation isn't as fun as thinking your computer is capable of cognizant thought, but it does allow it to optimize the system and not require constant replacement of sensors, etc. (which is probably the real reason for its presence in the first place - federal law requiring conformity for x years and x miles.)

Computers occasionally, as we all know from obviously being able to read this in the first place, catch a 'glitch' (normally voltage fluctuation induced) and create operational errors. Usually the easiest way to correct the problem is to reset the computer by disconnecting its power supply. The older the car and its sensors are, the bigger the difference felt when/as the computer 'learns' the values over again and corrects for deviations from the original programming PIDs.

Whew! That was a kinda short and fast, hopefully it clears up the misconceptions. I know I need another cold and frosty now....

[AutoTchr]

You are absolutly correct about adaptive processors. But because different drivers will create different sensor values for given situations, the processor can be said to adapt to different drivers. Semantics.

All processors are designed to adapt to small deviances. Otherwise the MIL would be lit constantly.

OKee.. Question. Say I fill my empty tank with 89 octane and drive it down like a limo. (knock sensors retard the timing) Then I fill with 93 octane and drive it like a teenager. Do I now get full timing advance, power, shift points, etc? Do I have to reset the ECU/TCU? hehe

[AutoTchr]

I am not an expert on the SVX. Most cars would adapt pretty quickly/ possibly immediatly to the difference in octane. It has been discussed here that the SVX requires 2-3 tankfulls. I kind of doubt it seeing that the car is as advanced as it is. I fired my car up after sitting for two years. I know the gas was not the best. Ran fine because it adapted.

The transmission in my car shifts firmer than it first did. Did it adapt to me, or did it just get better because I serviced the transmission? The difference is so slight that I don't really care.

Also I would like to add that it would be possible for a manufacture to produce a car that was completely adaptive to driving style without having AI or anything of the sort. We not only monitor throttle position, but the speed at which the throttle opens. Not to mention the ability to monitor the speed at which the steering wheel makes turns. (Talking future here) The new SL has pitch yaw and roll sensors. Those could be used to automatically adapt suspension for cornering and braking. Currently they are used for the traction control system. The car has the ability to apply one brake in order to counteract a slide!!!

[Beav]

Originally posted by AutoTchr
You are absolutly correct about adaptive processors. But because different drivers will create different sensor values for given situations, the processor can be said to adapt to different drivers. Semantics.

I can see where you're coming from but I'll agree to disagree for the sake of others' confusion here. My point was that there is no big learning sequence that changes the car into something totally unmanageable for all but the one true driver of the car. Do so many truly believe that their driving skills are so advanced or different from others that their car needs to be dialed in, to the nth degree, for just how they alone drive?

My explanation was not intended to be an exacting, technical report, but rather a quick gloss-over showing that, with the rare exception of a glitch, resetting the computer by disconnecting the battery usually results in nothing more than the loss of radio pre-sets and the proper time displayed on the clock. Oh, and the possibility of a cd getting stuck behind a closed radio door...

[AutoTchr]

Yup, it is all in how you interpret the info. The techy will know that the processor is just adjusting for different inputs. Different drivers=different inputs. But your right about the Nth degree thing. I can notice some small changes, but I don't really care that much about them.

As I was saying with the SL, the everday person would say, "hey this thing is adapting to my driving style" as if there were some nueral interface or something. The tech would say, "hey this thing is jacking up the right front because I am making a hard left and the sensors are telling the processor what is going on." So you see, it is all the same thing just different interpretations. I think anyway.

[Beav]

Agreed! However, when Pop toodles the SVX home from work and junior takes it out drag racing later that evening, junior won't have to reset the computer to keep Pops from figuring out what he was up to.

[AutoTchr]

pops will never know, until he has to change the tires/ tranny

[Trevor]

Beav, as usual your practical approach is spot on.

The ÒadaptableÓ description must have originated from the sales department. The system can do nothing more than correlate and optimize inputs from transducers and adjust variable components accordingly taking into account a calibrated signal based on, resistance variation, or electrical output, constant or pulsed, or of variable polarity. A time constant could also be involved but if this is variable overall, considerable sophistication is required which one would not expect in an automobile control system.

There is no way the system could be arranged to ÒadaptÓ to variations in the set calibrated parameters without physical adjustment as in so doing the objectives could not be maintained in the event of faulty transducer outputs.

Trevor.