As cars become more complex, so do their problems. I remember the days when just replacing spark plugs could solve a problem.

Although we do have an ally in the onboard diagnostic (OBD) programming built into the computers that come with today’s cars, these systems rarely give us an exact answer.

But they do help. They can jump-start diagnostics in a certain system of the car and help with test plans. Twenty-five years ago, we had to pull out factory wiring diagrams and a digital voltmeter and test each wire one by one. We still do that, but the OBD system helps narrow our search.

As early as the 1970s, car manufacturers were installing onboard computers to manage fuel control and ignition, but few of them had diagnostic capabilities. Those that did were extremely limited, and the scanners were expensive. Only the factory and a few dealerships had such testing capabilities.

We recently had a customer bring in a 1987 VW Vanagon Syncro that has a computer inside but no onboard fault diagnostics. So we had to work our way out to fix the problem.

The customer said the idle was going above 1,000 rpms and would occasionally fluctuate between 2,000 and 3,000 rpms. He wanted to get the base idle under 1,000 rpms to comply with California emissions standards. Normally, the idle on this car should be 850-950 rpms.

This car uses an automatic idle control motor controlled by a separate idle control computer. The computer gets information from the throttle switch, engine rpm sensor, temperature sensor, power-steering pump pressure switch, oxygen sensor and mass air flow sensor.

For the idle computer and the idle motor to work correctly, all the components and the basic setting of the engine must be in specification.

Diagnosing the problem

We first checked the ignition timing; it was OK. Then we checked for vacuum leaks and discovered that the crankcase ventilation chamber was cracked, causing an air leak. If this idle system gets too much air passing through the intake, the idle computer will keep adjusting the idle.

We sealed the leak and tested again, only to find that the throttle body gasket was leaking vacuum. We removed the throttle body to replace the gasket, but also inspected and cleaned the throttle body. With the throttle body off, we checked the idle switch; it tested OK. The throttle plate looked a little worn, but not enough to cause a problem.

After putting the throttle body back on, we tested the system again. The idle was still high.

We next tested the idle control motor, idle control computer, mass air flow sensor and temperature switch. All tested fine.

During testing, we noticed that another shop had disconnected the power-steering pressure switch. Because this switch gives a signal to the idle control computer, we repaired the wire harness and reconnected it. There was still no change.

After looking at our data, we decided to replace the throttle body and finally managed to adjust the idle down to 900-950 rpms.

Breakout solution

We test-drove the car a few more times before calling the customer to tell him it was ready for pick up. While driving it later that afternoon, the customer called to say that the base idle was OK, but after a longer drive, the idle was fluctuating between 2,000 and 3,000 rpms. I told him to bring the car back and I would test it again.

We did not realize the car needed to be driven for at least 15 minutes before the idle started to fluctuate; the average test-drive is usually five to 10 minutes. It had been so difficult to figure out the base idle issue that I just missed the 2,000-3,000 rpm fluctuation.

We all took a step back and looked at the data. We knew that all the basics were OK but still checked them again.

We were convinced that the idle fluctuation problem was caused by temperature.

That’s when I pulled out a tool I had not used for at least 15 years – a computer-controlled breakout box. We disconnected the car’s computer from the main wire harness, inserted a communications adapter and connected a scanner to review all the live data coming from the computer. We basically turned an old-style computer with no fault-finding into an onboard diagnostic car.

We had one person behind the wheel while another looked at the live data on the scanner during several 15-minute test-drives. On the first two test-drives, nothing happened, even with 20 minutes of driving. Fifteen minutes into the third test-drive, the idle remained above 1,200 rpm for just five seconds and there was a quick drop out of the engine temperature sensor.

We returned to the shop and connected a digital voltmeter to the temperature sensor. After driving for 15 more minutes, the temperature sensor dropped out intermittently. We installed a new engine temperature sensor, drove the car again and immediately saw a better response time of the base idle. The idle was finally OK.

This was a difficult and time-consuming repair because the car had several failures contributing to the idle problems. It made me appreciate the modern onboard diagnostic systems. If this car had an OBD system, it would have been setting codes long ago.

Matt Pataky owns Sunnyvale Foreign Car Service, 15 Pioneer Way, Mountain View. For more information, call 960-6988, email or visit