Starter Motor Bench Testing and Repair

3 March 2011

In our group we had a Mitsubishi starter motor (Model Number M3T 49381) to pull apart and test various components to see if it was all working correctly. (All multimeter reading were measured using a 200ohms scale.)
The booklets we were given had a basic summary of all the tests we were about to perform, the possible results and why this would be so that we were actually understanding what we were doing.

To start, we did a no load test on the starter motor to make sure that the solenoid was operating. The result of this test was 11.5v and 35.2A. We then followed a strict procedure to dismantle the motor, marking sections so they could be realigned when put back together.

Armature:

First, the Visual Inspection, we checked the armature for any possible signs of overheating, burning, physical damage or poling. There were no real major visual signs of distress, only minor scratches on the plates which still means that the armature was serviceable. We then performed a ground circuit test, placing one of the meter leads on the armature core and the other lead on between each of the commutator segments. This should read infinity as there shoul be no circuit. As this passed, it shows that the armature is serviceable. We then had to test the continuity of the armature by placing one probe on the commutator and moving the other around the commutator (not breaking contact) to see how much resistance is within the circuit. The result of this was 0 ohms which is within recommended specifications and is therefore still serviceable. Then we had to measure the diameter of the commutator and the mica undercut depth. The diameter of the commutator that we were working with, was 20.05mm which is below manufacturers specifications of minimum diameter (26.8mm - 31mm). Therefore in this regard it failed and was not serviceable, however, it was explained that the specifications were perhaps not for the particular model that we were working with. Also the mica undercut was suggested to be 0.7mm - 1mm, it was said that the measurement of this was impractical and to make an estimation of depth. The armature was then to be placed "V" blocks to perform a dial test to check for run out. The armature had to be turned 360 degrees to get an acurate reading. The specifications for this were 0mm - 0.2mm. Our test came back with 0.05mm and therefore is serviceable.
We then used an alternative method of testing the armature using a 48 volt test light, to test continuity and ground. For the continuity test, it is performed the same way but on the machine, this test applies greater pressure to the windings and should be used where avaliable, the light on the test glowed, which means that the continuity of the circuit was good and still serviceable. For the ground test, it is also performed the same way, but in this instance the light will stay off to show theere is no circuit, this means it is still serviceable.
Also, to test for short circuits in the armature, it is placed on the "V" of the growler and switched on. A hacksaw blade is held loosely on top of the armature, the blade did not vibrate which means that there are no short circuits.


Field Coil and Pole shoes:

First, the visual inspection, where we checked for and visible signs of overheating, burning, physical damage or poling. Again, no major visual faults, only minor such as covering of wires tearing slightly, and coil wires becoming loose and frayed. This means that it was still serviceable.
We then checked the fied coils for continuity, but first we had to check the internal resistance of the meter itself which turned out to be 0.4 ohms. We were given 2 options on the worksheet 'If Grounded' (specifications - Test N/A) or 'Not Grounded' (specifications - 0-0.02 ohms). The results of my test turned out with just the internal resistance of the meter, so there for the test was N/A, meaning it was grounded and serviceable. To then check if the field coils were grounded we had to place a positive probe on the field wire or brush, then the common probe on the body of the starter. Again, given the same 2 options 'If Grounded' (specifications 0-0.02 ohms) and 'Not Grounded' (specifications Infinity), and being that our meter read infinity means that the component is still serviceable.
The length of the brushes had to be measured with the vernier calliper and specified to be longer than 5mm. Also, we had to check for cracks or other damage of the brushes. There was no visible damage to the brushes and results showed the 2 brushes were 5mm each which passed. The brushes also needed to be checked for short circuits between insulated brushes. This is achieved by placing one probe on the brush and the other on the metal retaining plate. The result of this test showed infinity which means that the holder is insulated and it passes to remain serviceable.

Solenoid Magnetic Switch:

We first had to identify the terminals. The battery terminal of the solenoid is obvious as it is isolated from the rest, the starter motor supply is also obvious as it lead from the solenoid into the starter, therefore leaving the remaining terminal to be the ignition/starter switch supply. We then checked that the pull in windings of the solenoid were working by connecting a 9v power supply to the ignition terminal and the starter motor supply terminal. This is only applied for 5 seconds to prevent heat damage to the windings. The maufacturers specifications of the current draw of these windings should be 8-12 amps, the pull in windings of ours however had a current draw of 23A which is too much current to be activating the windings. When they are activated however, the plunger of the solenoid should be pulled in which it did. ALso in the solenoid are hold in windings which are activated after the plunger is pulled in to hold it in place. This time, the 9v power supply is connected to the ignition terminal and the solenoid body. The plunger must be pushed in by hand and released to see if it is held in place. The current draw of this should be 5-8 amps. The result was 7A which passes.

Pinion Gear and Overrunning or One Way Clutch:

This needs to be inspected for damage and smooth movement along the shaft. The bushes must be checked for wear and also the bush clearance in the appropriate end housong. Each were in good condition and serviceable.

The starter motor then had to be reassembled, housing realligned to marks made. However, when we were putting the component back together, the frayed wiring of the field coil terminals became increasingly worse and one of the terminals snapped off completely. It was suggested that we continue to put it together without that terminal and complete a no load test to see if it would still work.
The final no load test was unable to be performed and the starter motor was unable to perform.


REPORT:

The initial no load test carried out gave results showing the starter was within manufacturers specifications. However, once taken apart it became apparent that the brush wires were frayed and therefore the brush came off completely causing the circuit to be incomplete. When final no load test was carried out after re-assembly of starter, the results were unavaliable as the windings within the solenoid were not able to generate anything. I would recommend changing/replacing brush holders.