Power Mode Override Switch - Update

[James614]

I tried the #4 connector-ground trick yesterday, and it ended up making sparks and that terminal on my TCU went dead, although everything still works fine (power mode still works and engages normally).

Maybe I did it wrong though; I shorted the terminal to a chassis ground (the steering column, the easiest available ground). Am I an idiot with wiring? Should I have spliced it into a ground wire on the TCU? (although don't those just run to a chassis ground anyway, and should have the same effect?)

[b3lha]

Holy shit! Sorry about that. It definitely shouldn't make sparks. Was it definitely the right pin of the right connector?

I only ask because in the photos I've seen, the TCU board looks the same as the SVX TCU board. The connectors are the same, and some SVX guys are even running Legacy transmissions in their cars. So it's bizarre to find out the wiring is different.

[James614]

I'm positive I got the lead. It's even in the exact same spot as the guy said in the SVX forum (from the pics the TCUs look identical). I think the issue may have been with my wiring job rather than the vehicle. Either incorrect voltage levels from my choice of ground, or a flaky connection, since my setup included Christmas ornament hangers and the leads from a voltmeter fandangled together.

[b3lha]

Yeah. I don't understand what happened in your case, but t was probably due to your wiring. A mate of mine had a JDM Legacy with a power mode button, it must have been wired to the TCU somehow. Most likely via that pin.

Any JDM Legacy automatic owners here able to check if there is a wire to that pin on their car?

[skid542]

Well I want to give an update.

I have everything wired in and all the switching works, mechanically that is.

Problem is the car's system actually runs at 14V instead of 12V (don't know why I didn't think of this before) so my voltage out of the LM317 is actually 7V instead of 5V.

The TCU doesn't like this much. It didn't put into power mode and basically just confused and upset the TCU i.e. sparatic shifting pattern and hard shifts at part throttle. It was after this that I checked and discovered the 7V signal instead of the anticipated 5V. But it doesn't seem to have done it any harm.

So I guess I need to put in a resistor to drop the voltage down and give that a shot. I still have hopes that the idea will work, just needs a little refinement.

If it doesn't, I might give B3LHA's idea a shot once I have access to a spare TCU incase something goes south.

Any input is always welcome and I'll keep everyone updated on the progress.

[skid542]

Well I went out and verified that I had everything hooked up right, or so I believe.

There are three pins Vout, Vin, and Vadj.

Vin gets the power at ~14V with the car running.

Vout goes to the TPS.

Vadj goes to chassis ground, via the stereo ground (that's where I spliced in).

Somehow though my Vout floats between 7V and 5V.

I haven't checked to see if there is a voltage potential between the ground I'm using and the chassis, maybe that's the next step.

At this point I'm a little confused and would appreciate any input on why I'm not getting the nice 1.25V at the Vout pin that I should.

Thanks.

[vrg3]

Vout goes to the TPS?

[skid542]

Sorry, I should have clarified, Vout goes to the TPS pin on the TCU.

[vrg3]

There's probably insufficient load on the output. Try putting a 200-ohm resistor across the V_out and V_adj pins.

[b3lha]

FWIW, there is another method that definitely works, but you need a computer in your car connected to the select monitor interface:

http://www.subaru-svx.net/forum/showthr ... post473050 (scroll back a bit if you can read assembly code)

When you kick down the gas, the power mode comes on and the TCU sets a timer. The timer counts down and when the it reaches zero the power mode turns off. The trick is to keep resetting the timer so it never reaches zero. Just a simple loop writing 0xFF to the memory location.

Messing with the TPS signal will not necessarily give you the effect you want because the TPS signal is used for lots of other stuff inside the TCU, not just the power mode.

[James614]

I found out what went south in my #4-ground trick. The coating on the voltage-tester leads I was using is unbelievably cheap; it superheats and blows off at the slightest hint of voltage, causing the impression of sparking. But further testing confirmed that the cheap bastard of a volt meter is dead (making me think I killed the terminal on the ECU), and those sparks fly off of every goddamn circuit I touch it to, without causing any significant disturbance in the current.


So I went and did tried it again the right way, with some solid copper wire and a fused switch, and I got a decent quality volt-meter. Testing showed a steady 7.15 volts from the #4 pin, which I guess is good (how much is there supposed to be?). However, grounding it out does nothing, nada, zilch.

Oh well, now that I have the supplies, I'm gonna see if I can't get a torque converter lockup switch going......

[James614]

Holy crap!!!!@@!@!@!!!!

Somebody give me the dumbass of the year award, mabye I should try driving the car before asuming something doesn't work

Grounding pin #4 works! The power light comes on with the shifter in D/3/2, and at very light throttle (such as drving in my alleyway), there's no change. But step on it just a hair harder, and you'll notice that it'll start shifting at 3600 - 4000 rpm where it used to shift under 3000, and the result is simply amazing. The trans goes from saving gas to staying right in the sweet spot of the power band. Dig in just over 1/4 throttle, and it'll take you up to just over 5 grand. The downshifts pretty much come whenever you want them, just blip the throttle and you're flying, versus the standard method of mashing the pedal and praying that the "normal" power mode will kick in (which doesn't always happen in my Touring Wagon). I didn't even know my car had this much power without the pedal mashed to the floor!

Depending on your style, you probably don't even need to wire a switch in, since it stays on the normal map at very light throttle and even after it goes onto the power map, it still stays within a reasonable range for normal driving until you dig into it.

[TheSubaruJunkie]

Wow, i think i wanna try this.

Awesome. Keep us updated. Expecially with your fuel milage. My milage right now is HORRID so I shouldnt be doing anything to make it worse, however I like the sound of having a manual power mode

[James614]

If fuel economy is in your vocabulary, then just forget everything you read I took a drive tonight and it looks like I'm gonna be right around 15mpg, vs. 18+ normally, although I didn't do the official fill-up measurement yet.

But then again, you might have the self control to simply leave the switch off, which I apparently lack Oh yeah, if you do any highway driving at all, definitely have the switch installed and turned off. Apparently overdrive doesn't come easily in power mode no matter how light you are on the gas, and the torque converter won't lock either.


I'd still like to see the TPS signal trick working, because theoretically it'll still provide better control for all-out driving, and the thought of forcing the TCU to stick with a specific shift map regardless of how you drive intrigues me. There could be one switch for "street" power mode (#4 pin grounded) and one for "race" power mode (modified WOT signal for forced redline shifts and instant kickdowns without having to blip the throttle).

[b3lha]

Holy crap!!!!@@!@!@!!!!

Somebody give me the dumbass of the year award, mabye I should try driving the car before asuming something doesn't work

Grounding pin #4 works! The power light comes on with the shifter in D/3/2, and at very light throttle (such as drving in my alleyway), there's no change. But step on it just a hair harder, and you'll notice that it'll start shifting at 3600 - 4000 rpm where it used to shift under 3000, and the result is simply amazing. The trans goes from saving gas to staying right in the sweet spot of the power band. Dig in just over 1/4 throttle, and it'll take you up to just over 5 grand. The downshifts pretty much come whenever you want them, just blip the throttle and you're flying, versus the standard method of mashing the pedal and praying that the "normal" power mode will kick in (which doesn't always happen in my Touring Wagon). I didn't even know my car had this much power without the pedal mashed to the floor!

Depending on your style, you probably don't even need to wire a switch in, since it stays on the normal map at very light throttle and even after it goes onto the power map, it still stays within a reasonable range for normal driving until you dig into it.

Woohoo!

I just knew it had to work! Your comments echo those of the SVX guys. The power mode switch releases hidden performance and makes the car much more sporty and responsive. This functionality was built in by Subaru, but they only fitted the switch on JDM models.
http://www.subaru-svx.net/forum/showthread.php?t=43498

If you drive sensibly, it doesn't make much difference to fuel economy. I keep my power mode on all the time except for sustained high-speed motorway journeys.


Try it out guys. It's a zero cost performance upgrade.

[b3lha]

Well I want to give an update.

I have everything wired in and all the switching works, mechanically that is.

Problem is the car's system actually runs at 14V instead of 12V (don't know why I didn't think of this before) so my voltage out of the LM317 is actually 7V instead of 5V.

The TCU doesn't like this much. It didn't put into power mode and basically just confused and upset the TCU i.e. sparatic shifting pattern and hard shifts at part throttle. It was after this that I checked and discovered the 7V signal instead of the anticipated 5V. But it doesn't seem to have done it any harm.

So I guess I need to put in a resistor to drop the voltage down and give that a shot. I still have hopes that the idea will work, just needs a little refinement.

If it doesn't, I might give B3LHA's idea a shot once I have access to a spare TCU incase something goes south.

Any input is always welcome and I'll keep everyone updated on the progress.

I honestly don't think messing with the TPS signal is going to do what you want.
I'm trying to reverse-engineer the TCU firmware, but it's very hard to do because there are so many unknown variables. The process by which the TCU decides when to shift gears is very complex.

[skid542]

Freakin' sweet James. I'm glad that it worked out so well.

A big thanks to B3LHA for bringing this to light. It's not often that discoveries like this are made anymore on these cars after the 'greats' have been playing with them so long.

I'm still going to try the resistor as Vikash suggested this evening. At this point it has become a principle thing , but if that doesn't work it's comforting to know there is a backup plan that works.

Keep us posted James.

[vrg3]

Neato!

Assuming the LM317 thing works, I'd be very curious to hear how the two approaches (grounding pin 4 vs. tweaking the TPS signal) compare. You may be right, b3lha, that it would just make things behave all wonky, but who knows?

[James614]

So, I measured my mileage, and it turned to be 18.4mpg, vs 19.1 for my last fill-up. And I hit 6000rpm more than a few times*. So... um..... yeah..... it takes like 75 seconds to do this without wiring a switch in, it's freakin' free, and it doesn't have a noticeable affect on fuel economy. Best. Mod. Ever.


Oh yeah, let me throw in that this was mostly city/rural driving, if you travel on the higway for long distances theres no torque converter lock and your mileage will probably suck more than normal, but unless you're one of those idiots who doesn't adhere to the national 55mph speed limit, that won't be an issue

[TheSubaruJunkie]

You have any pix by chance? I did alittle searching but couldnt find any info on where to locate the TCU. And if I did, im not sure which pin Im lookin for. Saw the scans of the pinouts so that should help alittle.

[svxistentialist]

Freakin' sweet James. I'm glad that it worked out so well.

A big thanks to B3LHA for bringing this to light. It's not often that discoveries like this are made anymore on these cars after the 'greats' have been playing with them so long.

I'm still going to try the resistor as Vikash suggested this evening. At this point it has become a principle thing , but if that doesn't work it's comforting to know there is a backup plan that works.

Keep us posted James.

Lee

b3lha and I worked on this mod for the US 4EAT because we both have Japanese SVXs, and we know the benefit of full-time Power mode.

Your US trans computer allows you to have Power mode available by simply grounding pin A4. This is the way the factory does it, so this is the way you also should do it. You can ground it permanently, or you can apply a switch. If you do long highway miles in your commute, I would advise a switch. While Power is on, the torque converter can not lock up. As it is of no benefit at steady speed, it is better to be able to switch Power mode off for long highway miles, you will probably save 2-4 mpg.

My current UK car does not allow this mod, different software, if I ground pin A4 it activates Econ, which is an economy mode, it denies the use of Power mode under any circumstances.

B3lha is working on a way to modify the software so that Power mode can be forced in UK gearboxes. If he has it finished I will be fitting it to my '92.

However, just FYI, there is already a method to force Power mode using the TPS signal pin. It was devised by SVX-Commuter, I think, if memory serves. You do have to feed the WOT signal to the correct pin. However, you have to use a one-way circuit, because you can't have WOT signal feeding the ECU, all hell will break loose.

In addition, you also have to pulse the signal to the TPS input of the TCU, otherwise like b3lha says, the software will turn off power mode.

Now in conflict with what b3lha says earlier, this circuitry was got working only when the pulsed signal was a fraction of a second, not 10 seconds. So you have to wire in a switching circuit.

I'm hoping to avoid all this if b3lha's software fix works well. Be assured though that it can be done that way and that it does work. As advised earlier, in the case of US transmission controllers, it is obvious that you should apply the simplest [factory] solution, rather than over-complicate things, when there is no need to.

Joe

[skid542]

Thanks for the post Joe. Very informative. It's really nice to see some good information about the TCU's and their functionality.

I will definately putting in a switch to pin A4. I got the LM317 working, that resistor did the trick. However, while it will foce the TCU into power mode, it's gets confused as hell at part throttle decelerations. Entering and exiting turns just didn't work as it wouldn't downshift and instead would let the engine rev back up until it felt that it was 'gear' matched and then it'd start transfering power back into the tranny.

It just didn't feel right or sound right. I'm glad I gave it a shot though and thanks Vikash for the help.

Now to go with b3lha's suggestion and your's Joe. Hopefully this will prolong the 'life' of the auto in my car before I feel the need to finish the 5sp swap .

Do keep us posted though Joe as you make progress on the UK TCU. There isn't a lot of information on the BBS about this subject and I hope this thread will become a resource to others.

[ericem]

Ok I plan on trying this, just to see and feel the difference. Then I will determine installing a switch. The car is N/A. Is this connector right on the back of the TCU?

[svxistentialist]

Yes. Go back to page 1 of this thread and click b3lha's link to the SVX Network. Nomake Wan has put up good pictures. Click on them, they enlarge. You need the black 20-pin connector from the back of the TCU. There is only one 20-pin block, so you can't mistake it if you count the connections. Pin A4 slot is empty. That is where you insert your ground wire. Needless to say do a neat job and don't leave any stray wire touching where it should not. Joe

[vrg3]

Thanks for chiming in, Joe. Are you certain that the torque converter doesn't lock up when the TCU is in power mode? The reason I ask is that, from my understanding of the torque converter lockup control mechanism, it seems that the converter is always locked up in 2nd, 3rd, and 4th gear, no matter what. I believe the hydraulic circuits in the transmission are designed so that, in those gears, there is no oil pressure supplied to the lockup clutch.