A340 Paddle Shifters

Hi all!

This is a long post and probably ambitious undertaking, so I appreciate if you are willing to read the whole thing! If not, there was your warning!

I am very interested in a project to have paddle shifters for part-time, full manual control over the transmission. I have a 2.7L Auto 4x4, so it is a dog, but a lot of the frustration is due to the transmission never knowing the correct gear to select. A lot of people would suggest a manual swap of some kind, but I want to retain the automatic, as I have other vehicles with other types of transmissions and like variety.

I have done a solid amount of digging already, and have found some threads that exist utilizing arduinos and DIY circuitry, but I want something robust enough to trust offroad. Other issues with the DIY threads I have found is most of them do not have good control of the TC lockup, dont allow for engine braking, and critically only control your highest gear (as opposed to your selected gear). This is effectively like using the stick to shift down to 2, where the TCU can still decide it wants to be in 1, since it is lower.

Older threads suggest using a suprastick setup, or latent solutions equivalent (old auto Supras used the same transmission apparently), and while people seemed to have some success with these, they are no longer sold and may present similar issues to the DIY setups in regards to TC lockup. I have also heard of a KDOG setup, but from what I can tell these never made it to the market officially in any major capacity. I have seen a couple youtube videos of both loosely in action, but nothing majorly convincing, even if I could get my hands on them.

This leads me to my current best options that I have found, both of which I believe are standalone TCUs that would require total disconnection of the factory TCU and throw a check engine light:

RAD Designs Winter's shift controller -

https://www.radesignsproducts.com/products/winters-shifter-electric-to-manual-control-kit

This is made directly for the A340E/F, and allows for a switch to change between auto/ manual settings, which is ideal. However, they seem to be designed to work pretty directly with floor "sequential" shifter units (not +/-, but manual select). It seems to me that these work with many momentary switches, but most paddle setups are just a single up and single down momentary closed signal. It may require custom circuitry to integrate this, regardless.

COMPUSHIFT -

https://hgmelectronics.com/pages/transmission-controllers

Also can be configured specifically for the A340. This seems much more highly developed (and pricey) and has supporting software/ apps for adjusting shift points and times in auto, various modes, etc. However this also only works as a standalone. These do come with compatible paddle shift kits, but are not usable with factory wheels or airbags seemingly.

Both of these options suggest that they can properly account for engine braking/ TC lockup, and can be used manually and automatically.


In terms of actual paddles, I have seen multiple people swap out 2014-2019 Corolla steering wheels on their tacomas, tundras, and ravs. One thread for a Tundra even goes through the pinouts/ clocksprings of making a factory Corolla paddle shifter wheel work with the tundra's Sport shift transmission, which is much easier as it natively handles the +/- functionality on the shifter console and can be tapped into electrically. The A340 is solenoid controlled (2 for 4 speeds, 1 for lockup), so it is not so simple in my case.

Ideally I would retain stock-ish look, airbags, and dependability.


So, my questions:

• Does anyone have experience with any of the controllers listed above that they could share with me? Computshift, RAD Designs, KDOG, Suprastick, others?
  
• Has anyone done something similar, and I am just missing the thread?
• I have pinouts for both Corolla and Tacoma wheels, but is there a resource for me to know signal types (voltage or communication type of signals from wheel buttons)
• I believe the ECU and TCU are not very closely intertwined, meaning I can run an ECU tune without it impacting my programmed shift points, etc. And by removing the factory TCU, it should not impact engine performance. This seems obvious, but I dont want to assume. Can anyone verify from experience?
  
• Any other advice? (besides that it isnt worth the effort, because I already know that lol)

Thanks for making it this far!

 

Bumping...

I found another possible option that could be used with custom circuitry. The Nifty Shifter is for Jeep AW4 transmissions, but these seem to basically be the same transmissions rebranded for different manufacturers. Seems like there are more options for the AW4 out there, so maybe I will poke around on jeep forums.

No one has experience with anything like this? Any other sub-forum on TW I should post this on besides 'Gen 2s' that may yield more hits?

 

We've done full control with aftermarket ECUs. This transmission is much more common in first gen tacomas and third gen 4Runners as well as any number of old Toyota cars and vans from the '80s and '90s. A340 has lived a long life, but it was basically only used in one sub model for the second gens, and is likely fairly unknown to that crowd

 

Thanks for the response!

Okay, I may be better off posting in the 1st gen section then to see if I get more hits. I see a few that allow full control, but none so far with native sequential shifting (which makes some sense with how the transmission actually shifts). Seems like id have to create the code to set that up myself in conjunction with any of the aftermarket options I have run across so far.

 

no one puts paddle shifters in a Tacoma because it’s a truck.

 

Paddle shifter? That’s crazy.

 

...except for you know, certain models of Ford Rangers/ mavericks, F150s, Ram 1500s, etc...

Oh, and not to mention Toyota plans to on the new Tacoma X-Runner. The Tundra already comes with a sequential shift function which behaves the same as paddles, just on the console. Heck, I've driven CVTs with paddle shifters, and those dont even have gears! If the implication is that paddles are for fast cars only, I prefer manual on HPSCs unless I am on track.

I will move this thread to the 1st gen section soon to see if I get more responses!

 

The Tacoma will still be slow as heck, dont get me wrong lol but the extra control will be welcome and I like a good challenge! But honestly what it really needs is an extra gear to help keep the 4 cyl in the power band (a wise person would 5 speed swap). Appreciate the encouragement!

 

If you're frustrated with the way it shifts, it would be worth considering a tuning of the engine and trans. I just tuned our 4cyl 3rd gen with auto trans and it's a night and day difference in the way it drives.

 

I have an OTT medium tune and it definitely helped offset the larger tires and weight of the armor, etc. But I believe the 3rd gen autos have a 6 speed instead of 4 speed? Correct me if I am wrong!

 

Not sure.. you are in the 2nd gen forum. Plus @Musubi3 mentioned a 4cyl and not v6

 

I am not sure what you mean by this, but I am in the second gen forum because I own a second gen (2015) 4 cylinder, with a 4 speed auto. He was mentioning he tuned his 4 cylinder Gen 3, and I was saying I believe those dont use the A340 4-speed, but a different 6-speed autobox. I am not sure where a v6 got mentioned?

Earlier I mentioned moving to the gen 1 forum if I dont get any help here because the A340 is more common on gen 1s, apparently.

 

Update:

After talking to a helpful fellow on the gen 1 subforum, I have decided to create this setup with my own microcontroller. I have found some decent source code as a jumping-off point that seems to function virtually with minor modifications. I have all of the hardware components I need to bench test, so the next step is creating the physical circuit. I am using and arduino mega R3 and paddles from a Corolla iM, whose wheel uses the same clockspring config as this gen of tacomas and tundras. Using an LCD screen to display auto/manual mode, gear, and lockout.

Definitely a solid start. My main concern is the possibility of a money shift, which would be much easier with paddles than with a clutch and stick. It looks like engine speed only is accessible via CANbus/ OBDII on these trucks, short of messing with the tach analog/ potentiometer, but that may interfere with the resistance of that circuit in the dash. With a basic OBDII scanner, I can pull engine speed over bluetooth to my phone, but I do not know how I will pull that signal into the controller.

Another much less important benefit to this would be that I can incorporate orange and red shift light LEDs, which will be more helpful if I ever throw a URD supercharger on the truck. As of now, you'd be pretty darn hard pressed to have rpms sneak up on you lol.

So.... anyone have experience pulling engine RPMs signal to an arduino? Not holding up progress yet, but i'd be afraid to street it in manual mode long term.

 

I know at this point I am basically just making this as a project log for myself to keep track, but I want to provide discoveries in case anyone else goes down this obscure and specific rabbit-hole.

I successfully compiled and uploaded my first round of code, and I am able to use the corolla paddle shifters now to toggle the proper output values (using LED lights and simulated feedback for now). I may need to amplify this signal to actually send to the solenoids on the transmission. Right now, I am using the automatic mode as a pass-through for the controller in order to fake out proper feedback and avoid a CEL, but I may end up using a physical switch instead, in case the board/ processor were to fail. Will upload a video on the breadboard soon, and clean up the circuitry for a PCB down the road.

As for the engine speed safety, it looks like the signal between the ECM and tach for my year are not analog, but a digital pulse in the form of a squarewave. I can either attempt to tap this signal and read the frequency, or mess around with a shield attachment for the Arduino called a SparkFun, which can read in CANbus via OBDII/serial. The latter option may be a bit more complicated and would commit my OBDII port semi-permanently, so I think I will explore the waveform equation first.

As I come closer to a more concise design, I will upload my schematics, interface with the FSM, reference documents, source code/libraries, and photos of the actual wires in the dash/ console/ clockspring. Thus far, I have only used about $40 of components, and most of that was the cost of the paddles.

 

Here is the progress so far. The three LEDs are the outputs for the solenoids on the transmission. The three dangling wires on the left are for signals from the ECM, so when I switch from Auto to Manual I can read their last state instead of always shifting to 4th. I apparently cant upload a video.


LCD readout shows mode, gear, and TC lockup. The arduino operates at 3.3/5V, so I will need a boost on the outputs and a buck on the inputs. Since automotive 12V is more like 13.7V and can fluctuate, I also wanted to regulate the voltage in from the ECM signals, so I ordered LM1117 5V regulators. Bit nervous about the heat those might put off on drain. For the boost, I plan to tap into the same power that the Arduino will be running off of and using MOSFET (N style) Transistors, IRLZ44N. These should be much faster and more reliable than a system of relays, which counts for shift speed.
In case there is an issue, someone else on a jeep forum used a 4PDT O-O switch to allow passthrough of the ECM signals directly to the solenoids, allowing a circuit to bypass this setup entirely. I am sketching the circuits for the voltage drop regulator, transistors, and switch now.

For the engine speed, I will attempt to use a pulseIn() function, as it appears the square way from the ECM is a rate of 10ms/DIV, and that function can scan at a max of 10microseconds in the controller with efficient coding. I am developing the engine speed downshift safety lockout calculations now based on the A340 gear ratios, and will test with a simulated pulse using a step motor input.

ECM pinout:
E5-8 - Solenoid 2 signal
E5-9 - Solenoid 1 signal
E5-10/11 - Lockup Solenoid (possible PWM?)
E7-1 - Engine Speed/ Tach signal (definite PWM)

From what I can tell sifting through the FSM, a check engine light will only be generated on an incomplete circuit, but no feedback to confirm position is read back in from the commanded state. I could be wrong about that. Probably will finish these circuit sketches and write some rpm lockout code and take a break until the new parts arrive. As always, advice (and insults) welcome.

 

Progress:

I built and tested a voltage regular to step down the automotive signals to the controller from 12-5V. I was advised not to use a standard buck converter, as those can be vulnerable to sudden spikes, and this should be steady with anything from 7-20V in. I hooked up my 4PDT switch to just the MAN/AUTO master switch, and the signal read into the controller properly, so I just have to duplicate that for my other input signals from the ECU.

I have the parts for the boost circuit, and I will wrap that up and report the results next. Also need to consider final PCB and making heatsinks, since 13.7-5V is a massive drop for those regulators. Will upload the circuit diagrams eventually.

After going down a long rabbithole of accumulator, line pressure, and lockup solenoid programming, I think I determined to go in a different direction. For my specific application, I really will never need the torque converter to lock up. If I am on the highway cruising for a long time, I will just switch back to ECM control. If I am trying to have spirited driving, offroading, or even just downshift to go up a hill, the torque converter is probably disengaged anyway, because with ECM control a) it CANT come on until 45 mph b) it CANT come on unless youre in at least 3rd c) it turns off as soon as you hit the brake. If I go URD supercharger down the line, I will get a v6 transmission cooler anyway.

So for my specific purposes, I am going to plan to just send a steady 18% duty cycle of 300hz PWM at all times in manual control (unfortunately, mine is a 2-wire PWM instead of flat 0-12, like the older ones.) When I haul or tow (I only tow about once every 2 years) I will just make sure to go back to auto once I am cruising. I will just leave this code commented out in my upload, in case anyone else is doing this sort of thing primarily for towing, in which case that could be very helpful. Just be aware that it isnt complete or tested.

From what I can see, line pressure is pretty closely correlated to the accelerator pedal, so in theory if I leave it as is, it should respond well to shift harshness based on my driving style during the shift, rather than the gear itself. If I find it is inconsistent, or too harsh or soft, I will circle back. These transmission used to be pressure by throttle cable until 2005/2006, anyway.

I have programmed in safety checks for engine speed and tested them in simulation, but not with the actual signal from the truck yet. Anyone still reading? My downshift limit point info with a 100 rpm factor of safety:

Max power ~5200, redline 5500

Gear 2 > 1 – 3003 RPM

Gear 3 > 2 – 3591 RPM

Gear 4 > 3 – 3877 RPM

I will most likely program shift LEDs, and have red at 5400 and orange at 5100 rpms, just as a bonus feature since I am reading it in anyway. According to some very helpful folks on another thread, I learned that the tach reads at 3 pulses = 1 rpm, so if I read the gap duration from high to high in milliseconds, the formula should be simple for each scan: 180,000/duration = instantaneous engine RPM.

It looks like there are open circuit codes that cause CEL despite what I thought in my last post, so I think I will try to bogey those with resistors.

 

I'll be honest—I have no idea what you're doing, but I'm all in! Your level of understanding is impressive.

 

Thanks, I appreciate that! I never know what I am doing, I just jump into things until I figure them out.

 

What's your goal for the final form of this device? Are you thinking as far as like a custom printed pcb, putting it in some sort of project box, etc?

 

I am definitely going to create a PCB with much cleaner and more compact circuitry, put I will probably not go far as a custom printed one, just a board I will make and solder on a bench. I am hoping to make it compact enough to fit in a sealed box under the console above the cupholders. I am not planning to make more than one of these, and I'd love to finish the whole thing for under $75!

For now the LCD readout I will probably adhere to the steering wheel column by the cluster, but if I go supercharged and need to get pods, I may make a round LCD pod with shift lights and throw it along the A-pillar with AFR/ boost pressure.