How to repair G200 (Roll - lateral) and G251 (Pitch - longitudinal) sensors

These are the sensors for the ESP system which detect acceleration / braking or g-force in a longitudinal direction (pitch) or sideways g-force on cornering (yaw). Both are single axis acceleration sensors and both the same part (1J0 907 651 A) but just mounted differently - one 90° to the other. The lateral (yaw - G200) sensor is mounted under the steering column and the longitudinal (pitch - G251) is mounted behind the glove compartment on RHD UK Mk1 TT vehicles.

See here for a guide by shaneg for removal: Lateral-G200-Removal. The removal of the other sensor behind the glove box is similar and shown here: Longitudinal-G251 Removal .

Note that this repair guide doesn't cover the later combined sensor G419.

There are a couple of guides in German here: G200-251-Link1 and here: G200-251-Link2 which are good for the pictures at least if you can't read German. All credit to them for delving into the units, finding the fault and posting the information.

Fortunately Dani (A3DFU) speaks fluent German and has kindly provided a couple of handy translations into English - much better than using Google transation!

View attachment G200-251-Link1-English.doc
View attachment G200-251-Link2-English.doc

Notes on diagnosis and repair.

Diagnosis:

With an ESP light showing and Vag-Com reporting a G251 longitudinal sensor with an implausible signal I tried zeroing the sensor using Vag-Com. Here's an excellent guide by Tony Rigby: G200-251_Read_and_Reset_Guide

N.B. When using the above recalibration procedure G200 will "zero" to 0m/s but G251 will "zero" to 24m/s and report a successful reset within perameters if all is working.

Unfortunately the sensor was reading a constant too high a value and could not be reset so needed replacing or repairing. I needed to first ascertain which of the two positions (steering column and glove box) the sensor was located. There seemed to be scope for confusion.

The lateral sensor G200 was reading near zero and in order to double check the good one is not about to be removed - if you reach up and twist the good sensor on its bracket a little you can see the reading change slightly. This is reassuring and a good check. The faulty sensor won't respond. If you unplug the faulty sensor the reading will change but to another fixed offset - also confirming you've got the correct one.

Repair:

Having read the translated guides I thought I'd add some comments of my own with some pictures from repairing A3DFU's longitudinal sensor.

Firstly you need to gouge out the silicone rubber potting compound. This is best removed with your thumb nail or a plastic implement rather than a metal screwdriver to avoid damaging components or scratching through PCB traces. Some IPA will help loosen the bond as the material is rubbed away with a wooden cocktail stick or similar and help to clean the PCB. Dig away round the edges of the PCB too to help undermine it using something sharp.



You then need to de-solder the three connector pins. This is best done with vacuum desoldering equipment but you can use de-soldering braid. Whichever way, make sure the pins are completely loose in the holes by checking they can be moved sideways with your fingernail.



You can then prise out the PCB being careful not to damage anything. Having removed the PCB you will probably find the chip is still embedded in the potting with its legs showing. Carefully cut around it and prise it out too




With the chip removed I did notice that the pins han no solder on them. The PCB had little solder. The pins on the chip looked like stainless steel in appearance although I later confirmed they were nikel/tin plated copper.





It did seem likely that the cause of failure was too little solder combined with vibration and muvement with thermal expansion and contraction over the years that had caused fatigue of the tiny amounts of solder holding the chip to the PCB. Poor solder flow due to lead contamination or poor flux could also be an issue The early days of lead free solder production were plagued with this sort of thing.

Here's the chip cleaned up.



Before re-soldering the chip it's a good idea to clean the pins using some abrasion to form a key and break through any surface contamination and oxide. A glass fibre pen is being used here although careful use of a small piece of wet and dry paper would do.



Make sure the dot on the face of the cuip (pin 1) is nearest the semicircular PCB cut out and carefully solder back into place but using a more generous amount than in original manufacture so it's mechanically stronger.





Inspect the other compoents with an eyeglass for poor soldering but in this case it they were all fine. I suspect the only problem was with the chip.



Before replacing the PCB, excavate more of the silicome potting from around where the chip was in order to allow new potting to flow around the chip and PCB and not have problems with fitting. Then replace the PCB firmly on it's mounting pegs and re-solder the connector pins.



Here I'm pouring in a suitable two part mix silicone rubber potting compound designed for this purpose. If you have access to vacuum equipment you can also remove all the trapped air so when atmosphere is allowed back in the silicone will shrink down and have no voids.



If you don't have the correct potting material and equipment you could instead use silicone bathroom sealant or similar but this is not the best due to it being an acetic ure requiring moisure in the air so may take some time to go completely solid. In one of the German guides someone used hot melt adhesive. This will probably work too but will not bond as well to the existing potting.



The reason the potting is there is not so much to prevent moisure from entering but to form one solid inertial mass and prevent unwanted vibration in the sensor. I don't think its there to hide the simplicity of a £200 part either!

When set completely the sensor can be returned back to the car and the reset procedure applied.