One of the main strengths of the internet is the ability to dispense information to anyone interested and having the means with which to access it. One who has experience in a field, and is able and willing to dispense it, has this information.
I was a Mazda mechanic from September 1989 to April 1995 and a Toyota mechanic from May 1995 to May 1996. While a mechanic, I was also exposed to some Hyundais, Fords (Mazda Navajos and pickups), Suzukis, and Hondas. In the latter part of my short-lived career, I kept some notes. Now with my own web site, I can make my notes available to anyone able and willing to read them.
I warn anyone reading them, however, that my notes are not a substitute for the expertise of a knowledgeable mechanic. I make them available only to share my experience in the automotive service industry.
I hope you find them helpful.
Drive Belts
Timing Belts
Oil Leak Detection For Non-Dynamic Seals
Other Oil Leak Notes
Cooling System Leak Detection
Sealing
Oxygen Sensor Removal
Alternators
Preventive Maintenance of Alternators
Wheel Balancing
Brake Caliper Guide Pins
Refrigerant Leak Detection
Rusted Threads
Using a Torch
Valve clearances change. As time passes, they usually increase, but they can decrease also to the point where they will produce a dead cylinder and eventually burn out the valve.
In the early stages of a valve lash not having sufficient clearance, the cylinder affected will be dead the first few seconds after starting. The symptom is like a choke not opening to allow sufficient air to pass, but the choke is open and works fine. This will progress into a dead cylinder at regular idle speeds. The engine seems to run fine at higher rpms. If allowed to progress, the valve will burn out.
Depending upon the vehicle, valve clearances should be checked about every 50,000 miles. (6 Nov 1998)
A/C, alternator, air pump, and power steering belt tensions are typically checked and adjusted when the belts are cold. Belts that have run for more than five minutes are considered to be used belts and adjusted as such.
1991 Hyundai Excel: Their alternator belt squeals at below-freezing temperatures due to insufficient belt tension. Their belt has to be tighter than one would perceive to be normal.
1993 Toyota Camry: A belt too tight, such as a power steering belt on a four cylinder Camry, will slip when wet, exhibiting a power steering inoperativeness.
If a belt makes noise, put only enough lithium white grease on the contacting surface to remove the noise. Putting more can compromise the gripping power and life of the belt.
Timing belts stretch as they get old. They do most of their stretching during the first several thousand miles.
On every cold startup, as the engine reaches operating temperature, the timing belt tightens; therefore, tension is checked and adjusted with a cold engine.
1989 and older Mazda B2200s: PCV valves were known to allow too much oil to pass through them, which prematurely fouled spark plugs.
1988 Toyota Cressida: One of these released blue smoke from the tailpipe while driving or idling. A spark plug was removed and, seeing through a bore-scope, I saw oil coming from the intake valve. It looked almost as if the oil was coming from the head gasket but that wasn't possible because only coolant could be adjacent to the combustion chamber at the head gasket area. In this case, leaking valve seals can be determined as the cause of the blue smoke.
1991 Honda Accord: One of these 2.2L engine sedans came in complaining of a no start. When I got it, the valve cover and distributor were already off. The camshaft had split in two at approximately the #3 journal. The other journals were severely scored indicating an oil starvation. I removed the oil pan and only about 1/3 of a quart of oil was present. The main leaks were at the valve cover gasket, the distributor o-ring, and maybe the front camshaft seal. I did not see any black main bearing caps1 in the short block2. The cylinder walls looked smooth and were not scored in any way, so I assumed the short block was okay. The new cylinder head that I ordered had everything in it: camshaft, valves and springs, even areas that needed to be sealed were sealed. When I put the car together, the engine ran smoothly and quietly, but as soon as it got to operating temperature, blue smoke came out of the tailpipe. The hotter the engine got, the more blue smoke spewed out – especially at wide-open throttles. Reluctantly, I released the car to the customer. Apparently, the blue smoke was temporary because the car did not exhibit any when customer came back at a later date with an unrelated problem. There must have been excessive oil residue in the catalytic converter.
To determine the exact location of an oil leak, clean the suspected area, pressurize the system to 5 psi, apply a soap and water solution, then note where the bubbles appear.
1985 Chevrolet Sprint: The front crankshaft and camshaft seals were known to have been leaking on this car. With a fine mist of oil, the leak was enough to reach the rear window of the car when driven at freeway speeds. Oil was also present on the front of the engine. Among all areas, oil was on the lower two-thirds of the timing belt cover and alternator belt. I assumed that the oil covering the rear of the car came from the front crankshaft seal, and that the oil that was scattered all over the front of the engine, possibly by the alternator belt, came from the front camshaft seal. After replacing the front crankshaft seal and timing belt, which was surprisingly free from oil contamination, and driving the car at freeway speeds, I confirmed that the cause of the oil leak, that covered the rear of the car, came from the front crankshaft seal.
When a cooling system leaks only a little, it will leak while the engine is cold and warming up. When the engine is hot, everything expands to fill in gaps and there won't necessarily be any coolant leaks visible. Coolant leaving traces when it dries up, can be cleaned with water.
When the engine is cold, leaks can be detected by pressurizing the system to 15 psi (in the case of the 1993 and later RX7s, pressurizing to 20 psi) and looking for coolant traces. If coolant is suspected to go into the combustion chamber, remove the spark plugs, disable the fuel flow, and crank the engine. Watch for coolant to exit the spark plug holes.
An engine that runs hot frequently will cause the water pump to leak, necessitating replacement, and may cause other cooling system leaks where hoses are clamped down. Other results will be coolant hose expansion (swelling), radiator cap rubber seal expansion, coolant reservoir cap meltdown, and radiator filler neck damage (when overheated).
If there is an aluminum looking crud that collects inside and at the top of the radiator, it may have come from the water pump.
It is good practice to replace the thermostat after an overheating incident. It was suggested, also, that cooling system sensors and switches be replaced.
To prevent any chances of even a new gasket leaking, apply silicone to contact surfaces. Silicone can be applied to backup sealing o-rings too. Putting silicone on head gaskets, however, is not recommended.
To know that your vehicle is burning fuel as intended, splice into the oxygen (O2) sensor(s) and observe voltage fluctuation. Voltage should fluctuate between 0.2 - 0.8 volts. Separation between high and low readings should be at least 0.3v, and fluctuation should be approximately eight to ten times every ten seconds. For carburetor engines, the mixture control solenoid dwell should be in specifications, but can be susceptible to variables such as temparature and/or altitude.
If the vehicle has been idling for a length of time, the O2 sensor voltage can become steady. Raise the engine rpm to 2500 to decontaminate. If voltage fluctuation does not resume shortly, attempt to trick it by purposely richening and leaning the fuel mixture. O2 sensors in need of replacing are typically unresponsive and/or their signal is not strong enough to reach the ECU. Manufacturers recommend replacing oxygen sensors as part of preventive maintenance.
1992 Mazda MX-3: A V6, manual transmission, MX-3 failed the carbon monoxide test at idle due to the neutral switch failing to provide ground to the electronic control unit (ECU) when in neutral. Because of this, the idle is boosted to approximately 900 rpm, from 720, and the O2 sensor voltage is up and steady at 0.9 volts. If the clutch pedal is depressed, the clutch switch will provide a path to ground to the ECU, bringing down the rpm to 720 and allowing the O2 sensor to fluctuate normally. The clutch and neutral switches share the same terminal at the ECU so ground can be supplied to this terminal if either or both switches are closed (the clutch pedal being depressed or the transmission in neutral).
When an oxygen (O2) sensor fails, chances are that it is rusted on. Run the engine until it is at operating temperature, and then run the engine at 2,500 rpm for two minutes. Spray with water for four seconds, the base of the O2 sensor where the threads are in contact with the exhaust manifold. Wait two seconds, and then remove the sensor. The theory is that the O2 sensor will contract faster than the exhaust manifold, freeing up the threads.
As a battery discharges, electrolyte level will come down and when it is recharged, it goes back up. Hence, electrolyte level can vary with the battery’s state of charge. When the battery is discharged, accurate testing cannot be acquired. It should be slow charged for more than fifteen hours at a temperature above freezing, preferably at room temperature. After charging and passing a load test, wait two hours and load test it again. If it fails the load test, the battery is bad, if it passes, the battery is good.
Given that the battery is good, when a car is idling and all unnecessary electrical loads are off, the voltage at the battery should be approximately 14.0 volts. Allow the car to be at operating temperature and the car to idle approximately five minutes. Delco Remys have shown to be okay at first but when allowed to warm up, the voltage output could go down to approximately 12.0 volts.
Commutator brushes should be replaced every 200,000 miles depending on whether most of those miles are highway or city miles. If those miles are more city driving, then brushes should be replaced in less than 200,000 miles.
An example with my 1985 Chevrolet Sprint with more than 360,000 miles, on 28 March 99, the alternator light came on, and there was evidence that the battery was not charging due to the dimness of the headlights and the lower frequency of the turn signal blinkers. When the alternator cooled, the alternator light went off normally, and the alternator charged normally. When the alternator warmed up, the alternator light flashed intermittently, and then immediately remained on. Upon disassembly, one of the commutator brush contact surfaces was grooved, meaning the brush in contact ate into its contact surface area, which is part of the rotor assembly.
1993 Mazda 626: A four cylinder 626 exhibited rear tire feathering, even though total toe was precisely correct. The thrust angle, however, was off to the left.
1991 Toyota 4Runner: An alignment machine indicated that one V6 4Runner's caster and camber will allow the vehicle to run straight without pulling, but it pulled to the right. The caster and camber had to be adjusted to show that the vehicle will pull left before the vehicle actually went straight.
1986 Toyota MR2: Under normal circumstances, tie rod ends can be adjusted without affecting the vehicle’s tendency to pull. The more one MR2’s steering wheel was tilted to the right, the more it pulled to the right. The steering wheel is now straight, and the drifting/pulling is no longer a problem. Let the record reflect, however, that this MR2's inner tie rod ends are bent a little.
1992 Mazda MX-3: If toe-in is too much, whether on drive or driven wheels, and you rub your hands on the treads counter-clockwise, you will feel rough edges. These rough edges may appear only on the outside edges of the tires. This one vehicle showed signs of underinflation. The middle half of the tread was normal.
When a tire is mounted and balanced, especially with minimum amounts of weight, it should not have to be balanced again unless steering wheel shimmy or vehicle vibration start to intensify over a period of time. A technician has told me, however, that the big tires of 4Runners and trucks go out of balance over a period of time. Unfortunately, I cannot confirm his observation.
An out of round tire and/or rim may be balanceable but may produce a brake pulsation, premature wheel bearing wear, and premature tire wear in the long run. It may not necessarily cause the car to shake or the steering wheel to shimmy.
Do not use “Antiseize” to lubricate guide pins despite its claim that it is a lubricant and can be used for anything in the brake system. “Antiseize” has been noted to turn powdery in extremely high temperatures, thus seizing surfaces that it is suppose to be lubricating.
Wheel bearing grease seems to work fine but is not as slippery under normal operating temperatures. Lithium grease seems to work fine and remains slippery under normal operating temperatures. Supposedly, the best is a special lubricant with PTFE from Volvo.
1988 Mazda RX-7: A customer complained that her brake pedal went down to the floor. I was able to confirm this by putting the car in drive, stepping lightly on the brakes, only enough to stop the car, then releasing the pedal only enough to allow the car to creep forward. I repeated this light up and down motion of the pedal while the car was in “Drive”. The pedal's height got lower on every down stroke. Once the pedal reached the bottom of the floor, I allowed it to rise all the way, and then when I brought the pedal down, it resumed its normal height. The brake fluid was at normal height. I concluded then that the seals within the master cylinder were leaking.
The system should be pressurized to operating pressures and cold when testing. Soak the suspected leak location with soap and water. Bubbles can be very tiny and take a minute to show up. After a couple of minutes, it can look like foam.
When adjusting rod ends that have some rust, first spray the threads with penetrating oil and let sit. Use a wire brush to remove loose rust, clean off the area by spraying more penetrating oil, then attempt to loosen with a strong combination wrench. If the nut won't budge, heat the nut with a torch until red hot, and then loosen (see "Using a Torch").
Fasteners too rusted to remove with sockets or wrenches alone may easily be melted off, or loosened, with the heat of the torch.
Make sure all valves of the torch control stick are closed. Fully open both the acetyline and oxygen valves of the tanks. On the torch control stick, open the acetyline a small amount, light the torch, and continue to open the acetyline until the carbon from the tip of the flame begins to disappear. Open the upstream oxygen valve all the way, then slowly open the downstream oxygen valve until the flame turns blue and approximately 2 - 3" in length. If the torch starts to “whistle”, the oxygen is opened too much. With welding glasses on, melt off the nut or bolt on the side of the part being replaced. As soon as the fastener is melted off, push out the rest of the stud with a punch and hammer. Avoid melting parts that are not going to be replaced.
NOTE: If the bolt is on an aluminum block, the following will not work. If a bolt breaks because it was rusted and there is something to grab on to, heat up the stud until it is red-hot then twist it off with vice-grips. This is done by opening the acetyline a small amount, lighting the torch, opening the first upstream oxygen valve completely, then opening the downstream oxygen valve until the flame is blue and approximately 2" in length. At this time, the flame is only hot enough to heat the stud red hot, not hot enough to melt it.
1. A main journal bearing cap that is black indicates that the
bearing had “spun”. If the cap is removed, there should be severe scoring of
the journal and the bearing’s radius should appear to have shrunk.
2. A
short block is the part of the engine below the cylinder
head.