The Beloved 3rd Gen Transmission About Got Me Killed This Morning

It's not even friday yet...

 

New cars will still be shit tomorrow. It's the '70s all over again baby.

 

I have, no issues with my 19

 

You have this backwards!

 

Excellent. Trust me, you don’t want to experience the “sudden loss of power” from the early gen3 days. Glad that issue is gone.

 

Got a 3rd Gen 3s-gte MR2 Turbo. CT21 turbo @ 16-17 psi. Now that I have the Taco for a daily driver, I can make it a real track car!

 

Except choose for me was exactly correct . . . .

The ECU detects a different waveform from the OEM pedal position sensor via the Pedal Commander electrickery, thereby opening the throttle plate faster, therefore more linear throttle response compared to OEM drive-by-wire.

The ECU sees a waveform that translates to deeper pedal position so it doesn't fuck around slowing throttle body opening.

 

I step on the gas, my truck moves.


Thank you Toyota.


This thread is pretty damn good also- thanks OP.

 

Ok. Cool.

 

Alright, I'm back. What did I miss? Oh, more of the same... Good news, everyone! I hooked my truck up to my nifty scan tool and took some measurements. I was able to export everything to excel, and after a lot of spreadsheet voodoo, cursing, and goat sacrifices, I came up with this pathetic product. I'll be the first to admit, my sample size is extremely low, but the data is the data. I'm no car scientist, but I'll try to walk you through it:

Initial time correlates to the moment the throttle pedal senses movement and generates signal voltage to the ECU. Based on the analog pedal input from that lead foot of yours, the ECU outputs the percentage it thinks it needs to change the throttle position. At idle, the throttle is open at 14.9% (this is really important in the analysis section). The next time block is when that movement occurs, with TAC reaction time representing the signal delay, which was constant for every single run, even factory. Next we have the absolute (actual) throttle position, basically where the throttle actually is compared to pedal input and ECU debauchery. The next time block records when the absolute throttle position (ATP) measurement was taken, with ATP reaction time showing the delay, which was also constant for all modes. The last column, % variance, is the difference between the throttle signal and the throttle position. I don't know the causal relationship between the two, but the relative percentages change with each mode, thus reinforcing that Pedal Commander does more than just simulate someone mashing the gas pedal.

Now for the fun part... analysis and interpretation. Right out of the gate, the factory output (with Pedal Commander uninstalled) missed the first signal, with the 21% signal being provided, and it remained at 14.9%. See, I told you that number would come in handy. It did not respond until the next measurement cycle, wasting nearly 2 tenths of a second before reporting a change, where 30.9% was provided, and 22.7% was achieved, for a variance of -8.24%.

I won't go through the rest of the sheet for you. Put on your thinking caps and check it out for yourself. One thing that I did find interesting is that every mode registered at -4.31% on the second cycle. Someone figure that out for me, please, or it will keep me up tonight. And sorry to the dude that wanted waveforms or graphs or pretty pictures of rainbows and unicorns. I couldn't save or record the real-time graphs, screenshots didn't work because I couldn't pause the stream, and taking pictures of the laptop screen with my phone didn't pan out (damn screen refresh rates). But it was very apparent, given the slopes on the graphs, that there were differences for each mode that extended beyond just pressing the gas pedal to the floor. If that were true, at some point all of the data would align into the same linear representation, and that didn't happen, even when the engine revved down.

Now for all of the haters out there that told me to get off my ass and prove my point... well, here you go. Now get off of your lazy asses and prove I'm wrong.

 

HaHa! Amateur

 

I don't disagree with you. I did some science on my truck today and posted it here. From my understanding of how the system works, the 5V is just the signal to kick things off. It's either 'on' or 'off' basically. Once the potentiometer, ECU, and throttle sensors get involved, things get wacky. The factory throttle mapping from Toyota takes a lazy slope upwards. It's been designed this way since the Prius pedal debacle as a 'safety' feature. PC simply makes that slope more linear (straight line), resulting in better, more accurate response.

 

Wow, it wasted nearly 2 tenths of a second? That's completely inexcusable.

 

Tell that to guy that almost died. I'm doing this for him, and all of us who almost died in the line of traffic...

 

Op says that during this lag, the trans was hunting through the gears. Where does that fit into the equation?

 

In over three years I have never had my truck when punched from a standstill, or definitely when moving, hunt between 1st and 2nd gear (really ? ) for 3-4 seconds, nor refuse to take off immediately.
Something is wrong because all trucks do not do this.

 

I'm impressed he went and gathered data, nicely done.

 

They all don’t do this. I have a 17 and it never happened to me. Is it a map difference?

 

I'm glad you asked. As the truck accelerates and gains speed/shifts, adjustments to the throttle need to be made to maintain momentum. Since the factory mapping slopes off more compared to using PC, the transmission will hunt unless you give it the juice. Guessing how much it wants has always been an issue for me on my truck, and it did exactly what the OP states. PC fixed it on day one.