Let me take you back about a decade. In those days I was just starting out as a freelance automotive writer, and was desperate for leads that would give me stories. At that time, engine management chip rewriting was in its infancy - and I had a healthy scepticism about how much power could be made by tweaking the fuel and ignition maps in otherwise standard engines.
Some of the stories that I was sourcing were from companies just beginning in the chip writing business. In retrospect, that desperation to break into the market led to their taking some very major R&D shortcuts - not every aftermarket chip made any positive difference to the power output of the engine. These days, that comment'll surprise no one; back then, my column words to that affect were sufficient to provoke threats of legal action against the magazine that published them. And my belief that a one-size-suits-all hot chip in a naturally aspirated engine does little or nothing for power has been confirmed many times since - note, however, that such a chip is not at all the same beast that can be developed by real-time dyno work.
But one re-written chip that I experienced all those years ago did impress the hell out of me. And it was also the sort that could've been sold over the shelf, car unseen. The chip was part of the control system for an auto trans, and the one that I tested quite definitely improved - by a substantial amount - the driving performance of the transmission.
It's of course quite ironic that the chip that worked was never - AFAIK - one of those actually sold to the public!
And what did the re-written software maps in the auto trans controller actually do that impressed me so much? Nothing mystical or difficult to prove. And come to think of it, at full throttle not really any difference at all. But the way that the chipped trans behaved at part throttle, and the way that it made the car feel when being driven, were major positives.
But what did it actually do?
The trans more easily picked up a lower gear, and it held gears to higher rpm at part throttle.
Simple as that.
But think about it. Consider driving along in fourth gear (or fifth in a 5-speed auto) and reaching a smallish hill. The car slows as it starts to climb, and you push down a bit harder on the throttle. Nothing much happens - the transmission stays in top gear. You put your foot down a bit more, until - finally - the trans drops a gear and suddenly you're racing up the hill - going fast enough that you find yourself backing off. Which then causes the trans to upshift..... Aaaagh!
What some people would consider a too-eager downshift is just what you're looking for; what car company engineers saw as a fuel-wasting down-change to be avoided is just what you - as a sporty driver - want.
So a performance auto trans chip can, in the real world of on-road driving, make a very substantial difference to the driveability and performance of a car. Of course, at full throttle on a dyno there'll be zip, zero and zilch difference to the power curve. And perhaps that's what dissuaded chip companies from launching into this area - they were so busy pedalling (often false) chassis dyno graphs showing claimed power improvements from engine management re-mapping that they forgot that here was a market that would find plenty of happy customers.
Here and Now
OK - now fast-forward to 2002. Pretty well all auto transmissions are controlled electronically - so plenty are ripe for chipping. But - and I'm sorry if I have been leading you astray - this story isn't about writing new software maps for your transmission.
Nope, instead it's a helluva lot simpler than that.
It's about making an adjustment which - at least on the trans on which I performed it - had very much the same affect on driving behaviour as that chipped trans that I experienced way back then.
But before I tell you want I did, perhaps I'd better background auto transmissions a bit.
[And if you think that this is the sort of story where the journalist tells in 2070 words what could have been summarised in one standalone paragraph, you could well be right. But I don't want to bury in a few lines of text one of the simplest and most effective modifications that I've ever performed - especially considering the zero cost and the really major improvement that was made. Plus, one thing that I have learned in all these years of writing for car magazines is to just let it all flood out... and out... and out. Just like this.]
How Transmission Control Works
In conventional, old-fashioned transmissions, the control is by means of a 'hydraulic computer' (normally called the transmission's valve body) that uses oil pressure to perform its function. Oil is pressurised by an internal pump and this pressure is modulated by two main variables: (1) road speed, and (2) throttle position. In the simplest two-speed transmission, these two variables oppose each other by bearing on opposite ends of the same spool valve.
This diagram shows a schematic of this type of control system. A spring holds the shift valve in the "first gear" position, to allow the car to start off in low gear. If throttle pressure assisting the spring is high (ie the accelerator is hard down), then the shift valve will resist the rising governor pressure, which is proportional to road speed.
However, when the throttle pressure drops (ie the accelerator has been lifted) or the road speed rises sufficiently, then the governor pressure will cause the valve to move to the right. A 1-2 gear change will then occur, as the valve directs fluid to the correct planetary gear control clutch and/or band.
Obviously, if gear changes are to be completed quickly and aspects such as kickdown are required, then some additions to this simple system are required. A manual control valve (so that P-R-N-D-L can be selected) is also needed. However, all fully-hydraulic transmissions are controlled using essentially this type of valve-pressure approach.
As implied, you won't find a fully hydraulic trans in any cars of the last 15 or so years. Instead, many cars of this period have hybrid transmissions, using some elements of electronic control on transmissions that are still basically hydraulically controlled. Generally, the electronics are used for features such as kickdown and torque converter lock-up.
One example is the Jatco L4N71B, as used in the Holden VL Commodore, Nissan Skyline and some Mazda models. This transmission is a real Grandpa's axe, with the 1970s 3-speed 3N71B having had an overdrive unit added and then some electronic control juxtaposed on top. The electronics are able to override the hydraulics only in the selection of overdrive (fourth gear), kickdown and torque converter lock-up.
But these days, fully electronic transmissions are the most common. Note that electronically-controlled transmissions still use hydraulics to apply the clutches and bands, but all the valves are triggered by the electronic control unit working solenoids. Up to 14 inputs are used in some transmissions, with sometimes six internal hydraulic solenoids controlled by the electronics. In many cars, the engine and transmission management computers are integrated, allowing the manufacturer to include refinements such as retarding the ignition during gear changes, to give smoother progress.
But - and here's the crux of the background story - the fundamental logic of the transmission operation has remained much the same over the last 50 years, with road speed and throttle position inputs being the main data on which gearchange decisions are made.
Back to the Story...
Over the years, the throttle input to the transmission has mutated more into an 'engine load' input. In fully electronic transmissions, the engine management ECU outputs a signal that sums the torque produced by the engine and the additional real-time torque multiplication calculated by the ECU for the torque converter. That signal then determines how early the gearchanges occur (just like in a simple fully hydraulic trans) and in addition, how firmly the friction elements are applied.
So, even in late model transmissions, if the trans 'thinks' that the throttle is down further than it really is, then:
Does that put a few ideas in your head?
Making the Modification
The modification - or, perhaps expressed more accurately, the adjustment - was made on a 1991 Lexus LS400. On this car, engine throttle position information is conveyed to the transmission by means of a simple Bowden cable. As the throttle is opened, so the cable going to the trans has its inner core pulled on. (The next model of Lexus, with the 5-speed auto, has load information sent electronically, rather than by a cable.)
And, making things ridiculously easy on the '91 model, the transmission throttle cable is adjusted by simply undoing a couple of nuts on the throttle bracket fitting. Then, just as with an accelerator cable, the pre-tension on the cable can be adjusted.
Wind it one way and the trans thinks that the throttle is open more than it is. Wind it the other way and the trans thinks the opposite.
And - he's finally getting to the point after 1660 words - making the trans behave so much better than standard is simply a case of loosening the adjusting nuts and moving the collar so that the inner cable is tighter!
And that's all you do!
But surely that would result in a lot of negatives - like won't the engine now bump the rev limiter before changing? And won't all the shifts be harsh? And what about when flicking the Power/Economy mode switch - that must do something bad.
Well, on the Lexus the answers are no, no, and no.
The improvement in transmission behaviour is major. It changes more crisply. It better uses the (sweet) rev range of the 4-litre V8. When climbing hills or wanting better in-gear acceleration, it drops back a ratio more readily.
And the effect is easily and completely adjustable. Since it's my fiancé's car, if she decides that the modification is good - but it'd be better with a little less of it - then it's literally a five minute job to take off the throttle plate cover and make another transmission cable adjustment.
I must confess that this isn't the first modification that I've made to an auto trans. Way back when I owned the Nissan turbo six VL Holden Commodore I completely rebuilt the trans, making up my own shift kit that lifted clamping pressure and gave completely revised shift points. But that was a huge job, involving much oily trial and error of springs within the valve body.
The VL conveys the throttle position (or more accurately, engine load) by means of a vacuum line from the manifold to a vacuum modulator on the side of the trans. Adjustment of throttle valve pressure is made in that case by swapping in different length rods that connect the modulator to its trans valve. Changing the rod length should therefore alter the resulting throttle pressure signal.
And no doubt there are other methods of conveying engine load to auto transmissions - some very old transmissions used a solid rod, like an accelerator linkage - but in all cases they should be readily adjustable.
So while I have seen a few transmission experts kinda suggest that making minor changes to the throttle signal is pointless (and of course it's not going to turn a 300Nm trans into one capable of handling 500Nm!), on the Lexus the no-cost, five-minute change has made a wonderful difference to on-road driveability.
And that makes me pretty happy. Even if it took me 2081 words to get it out....