One of many first jobs I had, after I first obtained out of faculty, was for a corporation that designed and manufactured screens for giant vans, the type utilized in mining operations. This firm was a small entity with round 25 staff and a few engineers. The primary product was a monitor that sat on the sprint of those vans and watched over issues like oil strain, coolant temperature, and degree, hydraulic strain, and so forth. Variations of this monitor had 4, 5, or 6 indicator lights that lit if the monitored level went out of spec. An alarm additionally sounded, and the truck was shut down by a relay connection.
Do you could have a memorable expertise fixing an engineering drawback at work or in your spare time? Inform us your Story
One other engineer and I made a decision it was time to carry this analog monitor into the microprocessor period. The concept was to observe the identical features, however solely have one indicator mild with an LCD exhibiting the problem. Together with the alarming perform, we might additionally add extra info on the LCD, like temperatures and strain readings. It wasn’t a really advanced design. At the moment, micros didn’t usually have watchdog circuits, so we added one of many few exterior watchdogs out there on the time. Our concern was that some transient would throw the micro off target, and we needed the watchdog to reset the monitor in that case. The 24-V enter voltage and all sensor inputs had some degree of transient suppression (however, after a number of many years, I’ve forgotten what the circuit consisted of).
We accomplished a design, and it labored very effectively on the bench. Subsequent, we hit it with varied transients that we might generate. Not accessing any transient check gear, we needed to invent some strategies to check this. Worse, we had no specs or common info on what sort of transients these vans can expertise, however we ourselves have been happy that it was prepared for a beta check.
After testing, we despatched the monitor to a neighborhood mining firm to have it put in on a working truck. We additionally despatched a harness system with leads lengthy sufficient to get to the sensors situated across the truck. The corporate known as us after they obtained the monitor mounted on the sprint and all of the sensors wired to the harness, so a go to was scheduled to check the monitor on a operating truck.
I must cease at this level to explain the truck. It was a 175-ton dump truck. There are larger vans now, nevertheless it was very massive for the time. Image tires 10 ft excessive and a 1600 HP diesel/electrical generator system powering electrical motors turning every wheel. The motive force’s cab was about 18 ft off the bottom and was reached utilizing an connected ladder. The motive force and the 2 of us climbed this ladder to start the check.
So as to add to the strain, there have been a dozen or so managers and staff on the bottom watching the checks. The mining firm managers gave the go-ahead to start. The motive force began the truck (give up a roar)—the monitor fired up, and the LCD started exhibiting the standing of the monitored factors… nice!
After a couple of seconds, the truck shut down… not nice. We checked out one another—a couple of seconds later, the truck roared alive once more—monitor working—a pair extra seconds, the truck shuts down—a couple of seconds later, the truck restarts, and so forth., and so forth., and so forth.
After a half dozen of those cycles, we instructed the driving force to close the truck down. We couldn’t tie up the million-dollar truck any longer, so we couldn’t do any extra investigation. We packed up our gear and left with our heads down.
Again on the store, we talked via what went on. We concluded that the monitor’s micro was disrupted by an unknown transient. The watchdog then found the code operating amok and tripped the shutdown relay. The watchdog then rebooted the micro, resetting the relay, which allowed the truck to restart itself.
One of many main design points was that some sensors required tens of ft of wire and have been unshielded single leads (most sensors used chassis floor). These single wires (or ought to I name them antennas) might have been shut to varied relays and electrical actuators on the truck, or worse but, close to the cabling used for the generator-to-motor system. Additionally, the watchdog, which did uncover the problem, didn’t fulfill its perform—it allowed the truck to restart.
That is the place “Tales from the Dice” articles inform us how they fastened the problem by including a bigger resistor, fixing a nasty solder joint, or transforming a reversed diode. On this story, there is no such thing as a joyful ending. The boss didn’t need to proceed with the venture, and I’m positive the client was not impressed. The venture was cancelled. So why did I write this up?
I believed it was a very good instance of what can occur on engineering tasks—generally they fail (shifting from the lab to the sphere typically exposes design points), and generally you don’t get an opportunity to repair the design. Younger engineers ought to perceive this and never be disenchanted when it does. Don’t let it get you down. Bear in mind, we be taught loads by failure.
Shortly after this venture, we obtained the chance to design a full, micro-based dashboard for a big articulated truck. One of many issues we designed was a fiber-optic cable data-transfer system to the again portion of the truck. This minimized the size of sensor wires, offering antennas for the transient. On this design, the system labored flawlessly.
Damian Bonicatto is a consulting engineer with many years of expertise in embedded {hardware}, firmware, and system design. He holds over 30 patents.
Phoenix Bonicatto is a contract author.
Associated Content material
- When a hoop isn’t actually a hoop
- Within the days of previous, when engineers have been daring
- Software program sings the cold-weather blues
- Going towards the grain mud
The publish Watchdog versus the truck appeared first on EDN.
