One support question we get often is related to issues with gear detection in manual transmissions (especially FG+) where the gear detection fails and the gear indicator on the dashboard clears, cruise control keeps dropping out in 3rd/4th or CustomOS boost by gear isn't obeying the commanded boost target.

Why does it happen?

Basically, there is no gear detection sensor in any factory manual transmission (other than Reverse) and unlike the autos (BTR, 5r55, or 6hp26) where either the PCM or TCM knows the gear by the commanded clutch/brake states (along with some plausibility), the manual gearboxes are dependent 100% on a computational method.

The computational method works like this (simplified):

1) Measure Engine RPM > 2) Measure OSS (Transmission Output Shaft Speed) > 3) Calculate RPM/OSS > 4) Wait for Stable > 5) Compare to Gear Ratio Table > 6) Select Closest Gear > 7) Lock Selection > 😎 Start Timer

In some calibrations the guys at Ford made particularly tight allowances for the gear ratio's for part 5/6 of the method:

A number of factors (clutch slip, jitter in RPM or OSS) can cause the instantaneous ratio (e.g. RPM/OSS) to move out of range and the gear detection program will fail. When this happens, the "gear selected" will go to '0' (e.g. Neutral) and any number of routines in the PCM relying on gear will quit or 'go funny' due to the gear changing from say 3 to neutral, or in this case the cluster displaying an error.

How to detect it?

If you think it's happening you can use the PCMTec Datalogger to verify. The following logging channels are useful:

MID36381    Estimated Gear Selected - MTX
MID61215    Instant Transmission Ratio - MTX
MID37788    Current Gear Ratio - MTX
MID110730     Time In Gear - MTX

Along with Throttle, RPM, VSpeed, and VSS Fail Flag.

Here is a classic example.

You can see the orange trace (MID37788) getting "bumpy" at a few locations, at the same time you can see the estimated gear (MID36381) drop from 3rd to '0'. (Yellow Arrow, Red Arrow). Time in gear (MID110730) resets to zero (green arrow).

When the gear estimator can't 'snap' to a gear it will change to and reflect the instantaneous ratio (green trace, MID61215), NB: these are not exactly the same at my marker due to some delay -and- different logging rates and polling times.

At the indicated point you can see the instant ratio is "1.3431" (indigo box). If you look at the image above for the tuning parameters, you can see this is just outside the detection range (auF0337). So with the instantaneous ratio being out of range for the required plausibility time (few ms) of any gear, the gear selection program reverts '0' - Neutral or unknown gear.

NB: this is a valid selection because if OSS = Zero and RPM > 1 (clutch in or in neutral), this ratio will actually go infinite (there is actually a clip parameter to set the default value) and this routine must accommodate that and provide some method of detecting Netural.

Why?

Assuming this particular example is driving on the road or a dyno (with a decent amount of drivetrain inertia) the rapid stabs of the throttle are casing a decent torque increase (and not so much rpm increase) which is likely slipping the clutch just enough for the above check to fail momentarily. Depending on the car, its age, and power targets, it might be due for a new clutch. 😇

How to fix it?

Once you have identified the cause and found the offending gear, usually opening up the ratio limits is sufficient (as mentioned, our opinion is they are a little tight from the factory for 3rd and/or 4th). In this particular example, 3.6 on the upper limit (auF0337) of the 3rd gear ratio was sufficient and the problem of gear detection abated.

Happy Tuning.