Rover Holic

The Airflow Meter (AFM) in my fuel-injected Triumph TR7 failed due to condensation and corrosion three weeks ago. This note is meant to help others avoid the problems I had in debugging the problem, and possibly even keep them from experiencing it in the first place.

The AFM is a Bosch part of L-Jetronic fuel-injection. A similar part is used in the Triumph TR8s that are fuel-injected.

Electrical testing of the AFM is easy if you know the schematic and the values of the resistors. Since this information does not seem to be publicly available, I have included a table of values measured from my own TR7 AFM. Also included, for comparison, are values from a Canadian Rover SD1 (Rover V8), measured by Mark Elbers. To decode the resistor labels you will need to refer to the schematic (size: 41.8KB). The resistors are positioned on the schematic relative to their position on the actual circuit board. Just orient the AFM so that you are looking down on the circuit cavity (plastic cover removed) with the wiring harness connector facing away from you and then compare to the schematic.

Note: do not use these resistor values as gospel, you may expect your measurements to vary by up to 30% from those shown here.

The following describes the circuit schematic mentioned above.

Note: The pin-out numbering I use is as shown in the schematic, i.e. When looking at the wiring harness connector with the plastic circuit cavity cover facing up, the pins are numbered 1 to 7 from left to right. (On Mark's Rover SD1 AFM, the pins are numbered on the connector, left to right, as follows - 39, 36, 6, 9, 8, 7, 27. That is, Mark's pin 39 corresponds to my Pin 1.)

The switch between pins 1 and 2 is one of the controls for the fuel pump. It closes whenever there is even a small amount of air flowing through the AFM and moving the flap.

The thermistor, between pins 3 and 7, allows the electronic control unit (ECU) to compute a correction to the fuel need based on the temperature of the air being combined with the fuel. It shares pin 3 with the potentiometer, which is connected to ground within the ECU.

The potentiometer, shown with its wiper on pin 6, is a track of resistive material shunted by eight laser-trimmed resistors. The wiper is connected to the flap within the airflow. The shunt resistors are trimmed to give the right input-output relationship as explained by Tero (k124775@ee.tut.fi):

>The air-flow flap angle is logarithmic to air-flow, voltage
>is exponential to flap angle, and thus voltage is linear to air-flow.

The resistor values given are the parallel combination of the track material and the shunt resistors.

Pin 4 is connected to the battery voltage through the ECU when operating. Pin 5 allows remote sensing of the voltage on the top of the potentiometer by the ECU, which is critical for high airflow applications because of the logarithmic nature of the potentiometer (its small values and thus low voltage changes near the top).

The AFM can be tested while installed in the car. In the TR7, you can remove the air cleaning filter and the small plate under it for difficult but useful access to the flap that is deflected to measure the airflow.

To test the AFM, check that the right resistances are seen between all pairs of terminals. This is my recommended set of tests:

1. pin 1 shorted to pin 2 if and only if the flap is away from the
no-air position.
2. pin 3 to pin 4, 273 ohms within 30%.
3. pin 5 to pin 4, 100 ohms within 30%.
4. pin 6 to pin 3: less than 300 ohms with no flap movement, changing
smoothly and mostly increasing as the flap position moves towards all
the way open.
5. pin 3 to the casting: over 1 megohm.
6. pin 3 to pin 7: temperature dependent, see your manual for details.
Under pleasant temperatures, it is about 2 kohms, decreasing as it gets
hotter. This is unlikely to be bad unless it measures a short (less
than 10 ohms) or an open (greater than 1 megohm).

If it passes all these tests, and the flap motion doesn't feel erratic due to binding or a bad bearing, then the AFM is good.

Another interesting test of the AFM is that, if the car runs better with it disconnected (at least on the TR7), then it is bad. This test uses the Bosch L-Jetronic "limp-home" mode. Not all vehicles (i.e. TR8, Rover SD1) that use a system based on the Bosch L-Jetronic system have this feature.

In the absence of any reports of problems with the resistor track, I would recommend not trying to clean it. Certainly, the Radio Shacks of the world are full of chemicals ("tuner cleaner") for ancient televisions with mechanically switched tuning elements that had corrosion problems, and these might work well. But there is some chance that there is a lubricant here that you don't want to displace. And the track has redundancy (two contacts). And I'm sure Bourns/Bosch testing included simulations of hundreds of thousands of miles worth of wear on this track. I would say, Leave it alone.

A solution to the condensation problem

In my car, condensation occurred within the electronics area of the AFM, resulting in three problems. There was dirty water on the resistors, giving me a reading of 20 kohms from pin 3 (or 4, 5, or 6) and ground (the aluminum casting). There was corrosion that blackened the surface and eliminated the conduction of the trace from the top of the resistive track to the 2.4 ohm resistor. There was similar, nearly invisible corrosion on the trace from pin 4 to the 100 ohm resistor, right where the corrosive condensate was trapped by surface tension under the lead from connector pin 5. The AFM would have run pretty well with the first two problems.

The condensation problem could be avoided by drilling a drain hole in the bottom of the AFM where there is a "well" in the casting. In my car, this well was brimming with condensed water. The hole is hard to locate well, so measure carefully before drilling. It should be near the lowest point inside the electronics cavity when the AFM is installed (slightly angled) inside your car. The hole must be big enough that water surface tension will not keep the condensation from draining out; I recommend a 3/16" (5 mm) hole. On my first attempt, I drilled through the wall into a cavity that is accessible to the metered air, and had to seal the error with Duxseal.

I would recommend the drain hole improvement for anyone who lives in a condensation-prone environment. Here in Santa Rosa, California, the average summer low is 50 and high is 85 degrees F, so there are lots of mornings with dew or fog or condensation. Nonetheless, I had no problem for the first 16 years of the car's life.

If the AFM has failed due to corrosion, it can be repaired. Clean the surface of the resistor network with alcohol or other circuit cleaner. Solder small wires (e.g. #28 AWG wires for wire-wrap applications, stripped of insulation) across those conductors that opened due to corrosion. I was able to work under a microscope, and use special silver-bearing (4%) solder. But I don't think either the special solder or the microscope is required.