Will the ABS work without this switch?
Where is it located? Underdash?
Thanks
Evan

 

this is just hearsay, but since no one else has yet replied:

the ABS computer will go into fault and not operate if the accelerometer is disconnected. you can trick the ABS by placing a resistor of the correct value in place, and the computer thinks the car is always going straight.

this was an illegal trick that some auto-xers used, saying the car responded better to trailbraking.

i have not verified any of this myself.

-michael

 

Hello Evan,

According to Gordon Killebrews school text, "The lateral acceleration switch is located under the front of the console behind the A/C control head."

Some additional comments, "The lateral acceleration switch contains two mercury switches. When .5 G or greater is denoted the switch goes closed teslling the ECBM the car is in a lateral movement. In this case ABS cannot be used. The car would not be stable."

MY 90 and above has a different design, voltage varies with g forces.

The system should not work with a bad sensor. I believe this signal is part of the BIT check in all years.

dlmeyers 90 coupe zf6 3 speed shocks

 

True with a note. At initial test the pre-90 Vette ABS controller will cycle the pump and test all all 4 wheel sensor inputs, which also activates all 3 relays and tests them indirectly. If any part fails the system test the controller sees a fault condition and disables ABS. At startup the lateral accelerometer should read 0 ohms, so if you remove the acceleromoter and connected the wires it would appear good to the controller. The system will work, however....

Interesting idea, but incorrect. The ABS controller continually monitors the 4 wheel speed sensors for input voltage. At 15 mph the front sensors should each read approx. 0.65 millivolts and the rear sensors should each read 1.1 millivolts AC. This voltage level varies with vehicle speed, but when traveling in a straight line the fronts should always be about the same as each other with the rears each approx. 0.45 volts more than the fronts.

If one of the wheel sensors shows an abnormal value compared to its partner on the other side of the axle, the controller will check the resistance of the lateral accelerometer switch to see if the car is in a turn. If so, then the voltage difference is to be expected since the outer wheel of both axles will be rotating faster than the inner wheels causing the voltage discrepencies. (this is where the resister to fool the switch idea came from.) While this could fool the controller during turns, the controller also checks volatge readings when driving straight and expects to see 0 ohms when the readings are equal side to side. That will not be the case with the resistor soldered in and it will see a fault when driving straight. (Might work for oval track racing :)

You can test the lateral acceleration switch yourself. Remove the right console carpet and locate the switch behind/under the AC/blower controls (it will have a orange and orange/black wire).
Test by holding it in your hand in the same position as it was mounted in the car and measuring the resistance with an ohmmeter. It should read zero ohms. Then turn it 90 degrees to vertical and check for an open (or infinite reading). If it fails this test replace it. Check the junk yard for one. Nobody usually buys them there because they seldom go bad. You should be able to get one for about $20. If you can't find a good one at a local yard, I have a spare I'll sell you.

 

thanks for the education.

i know someone who claims to have replaced the sensor with a resistor, and it did not disable the system. he also claimed an improvement in auto-x. are you saying this could not be true, and that doing so would necessarily throw a fault code?

thanks.

-michael

 

Well, I'm not a GM ABS engineer, so I don't claim to know everything about it. I just know what I read in the GM documentation.

According to GM the lateral acceleration switch (LAS) has 2 mercury switches in series inside it which will open based upon the forces of angular acceleration which act on it in a turn. "When the vehicle is cornering faster than a given curve speed, one of the two mercury switches opens, sending a signal to the control module. " (GM Manual page 5E-3)

Also according to the GM Manual page 5E-4; the ABS system CONTINUOUSLY monitors the wheel speed sensors, the lateral acceleration switch and other parts of the ABS system. The test for LAS component failure is as I already posted, 0 ohms in normal position and an open or infinite reading when turned vertical. (page 8A-44-10 procedure E). With a resistor in place of the LAS there would not be a reading of 0 ohms and the during its CONTINUOUS monioring cycle the controller should see a fault at the LAS. This would cause it to set the ANTILOCK fault light and disable the system until the car was restarted.

The 86-89 ABS system does not set a code. It is an all or nothing approach. Either everything is ok or the controller sees a fault condition which causes it to disable ABS and set the ANTILOCK light.

If it was possible that the ABS controller did not need to see a continuous 0 ohm reading from the LAS when traveling in a straight line, the resistor would be able to trick the controller into believing the car was traveling in a less drastic turn than it actually was by showing a fixed value near 0 ohms during the turn. Eevn if this was possible, when the wheel sensor voltages didn't match due to differing travel radius of the outer wheels with respect to the inner wheels the controller would be looking for a value of OPEN at the LAS. Now the trick is exposed because if at any time during the turn the brakes are not applied and the voltages don't match between axle pairs, the controller must assume there is a fault in one of the wheel sensors and that will cause it to set the ANTILOCK light and disable ABS.

Assuming that the tolerances allowed in every system were enough in this case to allow a resistor to provide a "close enough" range for the controller to be tricked, I have to ask you to consider this...
1) Which resistor will be "close enough" to both 0 and infinite ohms at the required times if the controller continuously monitors the LAS?
2) The LAS is mounted near the center of the car. This results in more manufacturing costs that if it was mounted in the ABS compartment with the controller assembly. It must be important to mount it where it is.
3) If a cheap resistor soldered between two wires would acomplish a better result than an expensive dual mercury switch based accelerometer and the additional wiring needed to mount it away from the controller and near the center of the car, why wouldn't GM engineers have figured this out?

The pages mentioned are all from the 1988 GM Factory Service Manual, the 1988 GM Electrical Diagnosis Service Manual Supplement, and the Kent-Moore ABS test unit for 1986-1989 Corvette ABS.

 

you make good points, but this one i will try to poke holes through.

GM wants to stop the car from spinning out at all costs. the contention of the auto-x crowd is that the lateral sensor causes the ABS to release a lot of rear brake pressure in extreme maneuvers. other points to support this theory would be the gravitation towards more front-biased tuning in the spring/bar rates, going to smaller front tires, lower rear roll center from '89-on, etc.

it's about liability, and GM knowing who mostly drives the cars. (i think we're definitely the minority here on the forum.)

but i do realize now that my thinking on the matter was simplistic. i do see how the 4 wheel speeds being different could be confusing, and so the lateral sensor would help the computer figure out what's happening. however, i have driven/raced cars (not vettes) without a lateral sensor that subjectively felt superior in overall braking and ABS function.

maybe back then the computers weren't good enough without the lateral sensor?

-michael