VERMILION, Alta. – As a professional educator, success is measured in teaching a student to not only learn the pre-ordained material but also opening them up to new ideas and possibilities never before considered.

As a result, classes are asked to experiment with different solutions to many common problems. During a recent fourth year heavy equipment technician class, a new use for an oscilloscope was discovered that could prove helpful for many of the shop techs working in the field today.

Most manufacturers suggest a normal mean time between failures for a starter is between 17,000 and 20,000 cranks.

This is now easy to measure on most modern electronic diesel engines – it shows up on the diagnostic data reader as a key on/off cycle – however, this only lets you know roughly when you might expect a failure.

There is a gap in the range of 3,000 attempted starts, which could represent several years in true operating terms.

So the key was to find a way for a technician to accurately predict an imminent starter problem before it turned into a costly emergency repair. To do this students first had to determine how efficiently the direct current (DC) pulse indicates the condition of the brushes as well as the field and armature windings in any given starter.

A starter takes 12 volts of DC and produces rotational torque. The effect of this is a pulsed DC stream.

As a starter begins to wear down, its brushes no longer contact the commutator on the armature or the armature windings can open – the latter often happens at the solder that connects the windings to the commutator – and it melts due to long cranking periods or starter overload.

Either condition will create an opening in the current flow.

Using a 619-cubic-inch displacement (CID) agricultural engine with a common 40MT starter and a well-charged 1200 CCA battery, the key was cranked in a warm shop using standard 15W-40 oil.

Three test results drew an average of 590 amps at an approximate cranking speed of 260rpm as measured with a photo-tach. This amperage draw is acceptable for a diesel engine starter of 619 CID, the normal rating being slightly less than one amp per CI. The revolutions per minute were well above the minimum of 200rpm considered standard under normal starting conditions. Remember, it’s not how long you crank a diesel, it’s how fast.

While the test was conducted, an oscilloscope was used to look at the pattern of the DC pulse. The line was almost flat with little to no discernable ripple.

After disassembly, it was proven this starter was in good shape.

There were no opens, shorts or grounds in any of the fields or on the starter armature. The brushes were in serviceable shape, made good contact with the armature and the brush springs were still firm not having been softened by overheating. Students then cut one of the armature windings with a hacksaw (it really hurt to watch this process by the way.)

It should be noted it is not possible to detect an open of this type with a test light or standard ohmmeter as there are two wires feeding into each commutator bar.

Once re-assembled, the starter was again connected to the same engine. The battery was recharged and the complete assembly was ready for the next go-round.

When the same test was repeated – this time using the now wounded starter – the results were very different. The amps measured increased slightly to 690. While this didn’t represent a statistically significant change, revolution speed was dramatically impacted.

The engine’s cranking speed was now measured at only 170rpm, a tremendous decrease. The results on the oscilloscope pattern didn’t lie. The damage collapsed the magnetic field and led to a voltage spike. A new wave emerged on the screen indicating the lone open wire. If undetected the commutator bar before the open bar would burn and fail.

Given most oscilloscopes have a $10,000 price tag they probably aren’t O/O-type tools. For fleets however, think about the cost of getting towed to a repair shop in the middle of nowhere, kissing your volume parts discount goodbye, paying the list plus 25 per cent out-of-towners markup price for a new starter, and then tack on the labor charges calculated using a very unfriendly clock.

If you catch a handful of units before they go bad, $10,000 may in fact be a relatively small price to pay.

– Don Henry teaches at Lakeland College in Vermilion, Alta. His e-mail is Don.Henry@lakelandc.ab.ca.

Add_Italic