Geartronics paddleshift closed-loop information

Closed loop - what is it, and why is it necessary?

A closed-loop system is any electronic or mechanical system that monitors its own operation and modifies that
operation depending upon the feedback received. In the case of paddleshift gear selection, the closed-loop
system monitors the gear position sensor to determine if and when a shift has been successfully completed.
This information is fed back to the control ECU so that the pneumatic actuator, engine cut or throttle blip can be
turned on and off as necessary to effect the fastest and most reliable gear shift.

By contrast, an 'open-loop' system does not have any feedback mechanism and simply relies on fixed time
delays for engine cut and actuator operation. At best, this is a significant compromise which leads to increased
shift times. At worst, the gearbox can suffer excessive dog wear or even breakages.

To get a better understanding of why closed loop control is not only desirable, but necessary, we must first
understand the mechanics of a gear shift using dog engagement...

Most racing gearboxes use either 4 or 6 drive dogs. Ideally, the dogs should be designed to maximise the
chance of a first time engagement, yet maintain sufficient strength to avoid breakages. To increase the
chances of the dogs engaging cleanly, the gearbox designer will engineer more 'backlash' into the dog
arrangement, thereby increasing the window of opportunity. However, this means that smaller and fewer dogs
must be used, which leads to a reduction in strength. Clearly a compromise has to be reached. In the real
world, the chances of a 'clean' engagement on each shift are in the region of 50-70%, depending upon dog
design. Put another way; on average, as many as one in every other shift will result in 'dog on dog' baulking.
When this happens, the shift force must be maintained (within reason) until the dogs align, drop into mesh, and
are able to take up the drive. If the shift force is removed before the dogs have come into mesh, a mis-shift will
occur. Now, depending upon the relative position of the dogs when they meet, and the relative speed of the
gear and dog ring, it can take a varying length of time for the window of opportunity to open up. The worst case
scenario is if the dogs meet on a leading edge at low speed. When the dogs baulk on a leading edge, we have
to wait until the full length of the dogs have passed each other before they will drop into mesh. If the engine is
only spinning at low speed and the gears and dog rings only have a low relative speed, the time delay for the
window to open up increases. Furthermore, if the shift is between 5th and 6th gear, the ratios tend to be so
close together that the relative speed becomes even less, thus compounding the problem. It's also critically
important that, on up-shifts, the engine power is kept off until the dogs have fully engaged. If the power is
turned back on when the dogs are only partially engaged, the driving torque will cause them to lock in this
position and the selector mechanism will probably not reset. Almost certainly, a subsequent shift will then fail.
It's therefore essential that the barrel position sensor is accurately monitored in order to detect when the dogs
have fully engaged. The Geartronics ECU monitors the barrel position to an accuracy of 1/3 of a degree.

So, you can now appreciate that the time it takes for the dogs to come into alignment can vary, depending upon
several factors. What you might not appreciate is how dramatic the variation can be. The results of our
datalogging show that in the ideal situation of a clean engagement, the shift time is basically as fast as the
pneumatic actuator can move the gear lever. On our Quaife gearbox this seems to be in the order of 40-45ms.
However, at low speeds in the higher gears the shift time can increase to a lengthy 200ms!

Now, if a semi-auto shift system uses open-loop operation, the timers can only be set to cater for the slowest
possible shift - in the case above, 200ms. If the timers are set to a shorter period (as is usually the case) there
will be an increase in mis-shifts due to the fact that time period has expired before the dogs have engaged.
But, if the timer is set to 200ms and the dogs happen to engage in 50ms, this means that the engine will be
turned off for 150ms longer than it needs to be! Not only is this wasting time, but it also causes very aggressive
and jerky shifts. During the excess 150ms, the vehicle would effectively be in an over-run condition and the
gears will be driving the engine on the 'back' of the dogs. When engine power is resumed, the dog driving
faces will slam into each other and the drive will be taken up very aggressively. Increased shock loadings on
the transmission can lead to premature breakages when this happens.

If the time delays of an open-loop system aren't set to a conservative value, there will be an increased risk of a
mis-shift occurring. When a mis-shift does occur, we need to follow a very careful strategy if gearbox damage
is to be avoided. By the very nature of an open-loop system, it's NOT possible to implement a safety strategy
because the control unit has no way of knowing if the shift has been successful or not, let alone be able to
determine what is happening immediately after the failed shift!

To understand the possible consequences of a mis-shift, we must examine what might happen to the gear
selector mechanism when the shift force is removed before the dogs have engaged. Depending upon the
design of the selector mechanism and the indexing 'star', the gearbox may behave in a number of different
ways. If the index star has shallow ramps and rounded peaks, the gearbox may be prone to finding 'false
neutrals'. If the star has steep ramps and pointed peaks the selector barrel can only ever come to rest in a fully
in-gear position. Depending upon the relative positions of the peaks on the star and the barrel rotation, the
gearbox may land in the previous gear or have a tendancy to pull itself into the desired gear after a dog-on-dog
situation. Once the star has travelled past the peak it will get to the point of no return and the desired gear will
be engaged. If the star is already on the downward ramp as the dogs start to engage then the selection should
be successful as the detent roller will not want to travel back up the ramp. If the dogs are very worn and have
rounded edges then there is the possibility that as the worn surfaces meet the dog ring will be forced away
from the gear and try to push the barrel in the opposite direction. You can therefore see that the design of the
star is critical for successful gear engagement. Some poorly designed gearboxes may even jump back to the
previous gear after a dog-on-dog baulking event. Whatever happens, you can be sure that it's not good news
for the gearbox!

So there you have it - the proof that closed-loop control is absolutely necessary, not only in the interests of fast
and reliable operation, but also in the interest of gearbox reliability & longevity.

All text is protected by copyright, Neil Wallace & Geartronics Ltd. No unauthorised reproduction in part or in whole. By all means provide a link to
this text, but under no circumstances whatsoever must you copy it to your website or otherwise publish it without the prior written consent of the
author.