Last week we looked at how the air/fuel ratio can be simply altered, using components costing just a few dollars. In this article we look at getting rid of turbo boost fuel cuts, altering one type of frequency-based output signal, and automatically switching modifications in and out. Again it's all very simple and cheap.
Overcoming Turbo Boost Fuel Cuts
Many engine management systems cut the injectors when turbo boost levels rise too high - it's the manufacturer's way of protecting the engine if a wastegate hose falls off. However, if you're deliberately raising boost, it's a brick wall that you don't want to run into whenever you put your foot down. If the boost cut is made on the basis of a sensor voltage (eg from an airflow meter or MAP sensor) that rises above the preset level, a very cheap and simple clipping circuit can be used.
What it does is prevent the signal from the load sensor rising above a certain level. This also means, of course, that if that sensor is being used to control the amount of fuel being injected, then this fuelling amount won't keep increasing with load - after all, the clipping circuit is stopping the output going any higher! (This is just the same as nearly all modules on the market which disable overboost fuel cuts.)
The circuit consists of two diodes and a 1 kilo-ohm pot. Its sensitivity can be adjusted by rotating the pot - this should be set so that the circuit clips at as high a manifold pressure as possible. As with any of these techniques, the power over the car's electronics is very great, so extreme care should be taken that the engine doesn't suddenly run lean during the setting-up procedures.
Modifying Pulsed Signals
As indicated earlier in this series, changing pulsed signals with only very simple electronics is quite difficult. However, there is one category of output signal where changes can be successfully made, again using only a pot. It's a rather unorthodox approach but in practice it works very well. What you need to have is a situation where:
- a flow control solenoid valve is controlled by a variable duty cycle signal
- the requirement is that the duty cycle be reduced in order that the modification outcome is realised
Note that we're talking in this case about a high speed pulsing, where the valve actually stays in varying degrees of openness, depending on the signal duty cycle. That's quite different to the pulsing of an injector, where the valve is completely open or completely shut for varying lengths of time. Valves which are pulsed at high frequency to give varying degrees of openness (or shutness!) include pretty well all flow control valves other than the injectors - so solenoid valves in the power steering, auto transmission and boost control systems.
Note that this modification
reduces the effective duty cycle that the valve sees, so you need to work out if a reduced duty cycle leads to the solenoid valve condition that you want. (Ascertaining this can sometimes be quite difficult - for a systems approach to follow, read
Part 1 of this series.) The modification is then quite simple - a pot is placed in series in one of the wires running from the ECU to the solenoid. This reduces not the duty cycle of the signal going to the solenoid, but the average voltage that the solenoid actually sees. The result is that the solenoid
acts as if it has a reduced duty cycle signal operating it.
Note that you cannot use a typical pot from an electronics store for this task. Because the pot is on the output side of the ECU, it needs to dissipate lots of power, so a wire-wound pot (eg 3 watts) mounted on a big heatsink is likely to be needed. You will soon tell if the pot power rating and/or heatsinking are too low - the pot will get too hot. The resistance of the pot will depend on the solenoid coil characteristics, but I have used a 22 ohm pot with success when modifying the action of a power steering solenoid in this way. High wattage pots like this are available at specialist electronics suppliers - eg RS Components.
For a specific example of this technique in action go to: "Modifying Speed-Sensitive Power Steering".
Switching Mods In and Out
A key point in these types of modifications is that very often in normal driving you will not want the modification to be taking affect. For example, if you use last week's 'two pots' approach to modifying the mixtures, it's quite likely that you will only want this modification working when you're on high loads. After all, if the car runs perfectly in some conditions without any electronic modifications needed, there's no point in wrecking that so that it will run better elsewhere. To achieve this all that you need to do is switch the modification in and out.
Here the circuit is shown for adding extra resistance to the coolant temp sensor - which in most cars will tell the ECU that the engine is colder than it actually is, causing an enrichment of the mixture, especially on transients and at low loads. (For more on this approach see the earlier parts in this series.) If you wanted to have this mod occur only at low loads, simply fit a microswitch to the throttle (or on turbo cars a boost pressure switch plumbed to the manifold) that closes when the throttle opening exceeds a certain amount. When the switch is closed the modification is bypassed.
However, often a relay will need to be used so that multiple circuits can be switched simultaneously. In this case, a microswitch is used to trigger a DPDT relay. The circuit that connects the mixture modification is bypassed at the same time as the original circuit from the MAP sensor to the ECU is reconnected. (In this diagram both the relay contacts are shown open - normally one closes as the other opens.) The switch could be activated when the throttle position or manifold boost pressure exceeds (or falls below) a preset value.
More complex switching can be made on the basis of engine speed or the voltage output of a sensor. We've covered two techniques for making your own switches that will achieve these outcomes - "Simple DIY Rev Switch" and "Warning!".
Conclusion
If you think through the process that you're trying to modify, have available a good workshop manual and want to do some simple experimentation, it's amazing what you can achieve while spending nearly no money. That's especially important when you're talking a car that didn't cost much in the first place!
|
Successes and Failures
The techniques covered in these articles have, for the most part, been thoroughly tried out in practice. For example, on my (now long departed) high boost Daihatsu Mira turbo I used the "two pots" technique (covered last week) for altering mixtures to allow the running of bigger injectors. That is, the mixtures were corrected with the simple electronic mod. Prior to fitting the larger injectors, I used the overboost turbo fuel cut circuit described above to prevent the cut shutting the engine down. (After the larger injectors were fitted, the over-boost cut circuit wasn't needed - the signal from the MAP sensor was being altered sufficiently by the other pots that the cut didn't occur.) More recently, I have used the 'Modifying Pulsed Signals' technique covered above to alter the power steering weight in my current car, a Lexus LS400. In short, these simple techniques can work very well. However, it's only fair that I tell you of a recent experience where the approach didn't work. Again it was on my LS400, and it was using the same technique that I'd applied so successfully to making the power steering weight adjustable. That is, the use of a heavy duty pot to reduce the average voltage being seen by a flow-control solenoid valve - having the same sort of effect as reducing its duty cycle. On the Lexus, the auto trans control is fully integrated with the engine management ECU. So rather than getting an engine load input via a throttle cable or even a throttle position input signal, the Lexus control unit exchanges this information internally. This is important, because what I wanted to do was to firm up the trans shifts - and the easiest way of doing this is to tell the trans that the throttle is open more than it really is. (For an example of this approach on an older car that uses a throttle cable running to the trans, see "Tweaking the Shift".) But with this info exchanged internally, it's not possible to alter an input signal to achieve this deception. Why not? Well, of course it would be easy enough to fool the throttle position sensor input, but since this is also used as part of the electronic throttle control, engine management, traction control and stability control systems, it's not something to modify without the implications being thoroughly considered! Instead, what I chose to do was modify one of the output signals - the pulse train going to the Linear Solenoid for Line Pressure Control. In just the same way as was done with the power steering flow control solenoid, clipping the top off the square wave peaks would have the same effect as reducing duty cycle, so changing the valve opening - and with this system, that would lift line pressure. (Line pressure is normally raised when the throttle is open a long way, providing better internal transmission clamping forces and also speeding up the shifts.) I accessed the ECU, cut one of the two wires leading to this solenoid and installed in series the high power pot. By altering the pot position, the firmness with which the car responded when being shifted from Neutral to Reverse or Drive was clearly able to be varied - so line pressure was being altered by the pot as it should. I then went for a drive - all seemed well, with the shifts firmer. But after a short while I realised something had changed. The trans would no longer kickdown, and the shifts had returned back to the standard firmness! What was happening was that the kickdown was being disabled as the trans went into limp-home mode, and the shifts were returning to standard (even when the pot was tweaked such a small amount that limp-home didn't occur) because this transmission is a learning design. What it does is record the length of time each shift takes then adapts the control system to re-establish those times... So in this case, the simple interceptor technique wasn't up to the task.
|
|