Engineering Investigations
Comment
WKetel
Haldor is certainly correct in his remarks about 5S, which we started referring ...
Rupert Goodwins
When I was embarking on my (sadly abortive) career as an EE, one of the first ...
The case of the not-so-melodious motor
Chuck Hill
2/17/2011 3:43 AM EST
I once worked with a test system for certifying and sorting 8-inch floppy disks.
It is a source of many interesting stories (including this Engineering Investigation) because of its unique combination of electronics, mechanics, and raw power. The main component was a sorting shuttle that zipped up and down an 8 ft. track. The motor could move this five-pound shuttle the length of the track in about 2 seconds.
When our senior analog engineer fired up the prototype for the first time and commanded the shuttle to move, it reached its destination and began to "sing." Examining the position control error voltage reveled a 3KHz oscillation that just happened to resonate nicely with the motor housing.
The result was an ear-piercing tone that put anyone exposed to it into a migraine in a few minutes. Add to that, the shuttle moved just enough from this to appear blurry. Your eyes would constantly try and bring it in focus but just could not quite manage. It would have sent anyone with even mild epileptic tendencies into seizure.
If you just rested your finger on the motor housing, the oscillation would stop, so our analog and mechanical engineers spent a week looking for a resonance in the motor and track design. But nothing could bring this demon under control. I was eventually asked to have a look at it. My first impression was this was an unstable control loop--not a mechanical problem. So I asked the senior engineer for his design notes.
What I found was that the bench prototype shuttle and track were very different mechanically from the model we were now working on. So I got out a spring scale and measured the mass and friction for the shuttle. It was nearly 2X different from the values in the notes. The mechanical pole (mass x friction) is the dominant pole in this kind of position control loop. So I doubled the value of the capacitor in the lead / lag compensation and the oscillation stopped. This all took me about 2 hours to sort out.
I hated to upstage our senior engineer (being a fresh out 20-something), but he made a fundamental mistake (literally). He forgot to re-examine his fundamentals. He was so sure of the integrity of his original work, it just never occurred to him there could be a problem there. The one thing he did do right was document his original work in an engineering notebook. This diligence saved probably several days having to re-do those calculations for myself. This is something I never see young engineers do today as a regular practice. And it quite often costs a lot of time retracing things when debugging in the lab.
Now that I am one of the ‘old guys,’ I try never to let some young engineer show me up this way, by remembering this lesson. But it doesn’t always work. Never underestimate the creativity of young minds.
Author Chuck Hill has 30 years engineering design experience.
|
|
|
|
|
zeeglen
2/17/2011 8:33 PM EST
That senior engineer, back when a young engineer, had probably been advised by a mentor to "write it down". Good for him.
Starting at an early long-ago job the first item we were issued was an engineering notebook. Hard cover, quad-ruled paper, and property of the company. These books got dog-eared and bulged with Polaroid scope photos, but everything we did was on record. Useful sometimes for patents (we were told) but mainly used to go back a few months during one of those "Why the heck did I design it that way?" moments as you related.
These days it is far more convenient to record design specs and lab test sessions as PC files with a word processor along with instrumentation plots, partly because it is so much easier to click on "find" rather than flip through pages, can be emailed to anyone that needs them, can be given a table of contents, and those darn Polaroid photographs don't bulk up the pages.
But I have to wonder if a word processor document has the same legal impact in a patent courtroom as a paper lab notebook.
Maybe the "young engineers" of today rely on their computers to retain all of their working data. Come to think of it, the last few jobs I've had did not require the notebook - everything was on computers.
Sign in to Reply
TimothyRyan
2/18/2011 8:43 AM EST
For a while, we had a "Legal Beagle" for a CEO. On one of his dog&pony tours of the labs, he was horrified to see some of the Engineers with lab books. He wanted all of them destroyed (which no one did). To his mind, they made it way too easy for someone to sue us for product liability. And this guy was always calling us an Engineering Driven Company. Yow!
Sign in to Reply
zeeglen
2/18/2011 9:40 PM EST
At one place (no longer in business) a head honcho was an ex-pro football player. Excellent choice to run a hi-tech corporation :-( . He did not bother us about lab notebooks, he had no clue what a lab notebook was. His main concern was that all lab cables must be placed in the overhead cable trays to prevent the lab from looking "messy".
Sign in to Reply
Haldor
2/24/2011 1:55 PM EST
We are currently implementing Lean Sigma (the quality program dejure).
The most visible part of Lean Sigma is 5S (basically clean up and organize your workforce). The 5S precepts are valid for a production facility, but result in serious problems in a development group. A big part of 5S is to discard anything you aren't currently using regularly. This means that we are constantly having to order parts that got 5S'ed and test setups that represent huge investments of engineering time are routinely discarded (only to be laborously recreated later when MOL is required).
As you can imagine I am not terribly impressed with Lean Sigma as applied to development engineers. It has given us a few bit of humor though. At one department meeting I suggested we try implementing JIT processes (Kanbans) to our engineering output.
I would have a Kanban of firmware sitting on my desk. When someone need some firmware they would come grab one out of my Kanban. The empty slot in turn would trigger me to write another piece of firmware. One of the drafters thought that was a great idea and he wanted to do the same thing with engineering drawings.
This idea was not warmly received.
Sign in to Reply
Haldor
2/24/2011 2:00 PM EST
Typo correction
"clean up your workforce" should have read "clean up your work place". Sounds like a bit of a Freudian slip.
Sign in to Reply
bblack
2/19/2011 10:08 AM EST
Sometimes we experienced folks can loose sight of the basics and think we have it right and then make a choice made from our own arrogance. Sometimes the learning experience from that mistake can lead to a fractured leg.... A hard lesson from neglecting good practices!!
Sign in to Reply
kinnar
2/20/2011 2:29 PM EST
Documenting the work is not attracting the young minds, may be some new audio visual technique might bring back the interest of archiving the original concept.
Sign in to Reply
Haldor
2/24/2011 1:44 PM EST
In the past we routinely used email to document change requests and decision processes. In the last few years corporate has forced us to stop doing this. First our individual email server usage was greatly reduced to 200 MB per person and all local copies of email messages are prohibited. In addition the email system now automatically deletes the contents of our inbox and outbox that are more than 90 days old.
The story is that this is to protect us during legal discovery requirements. If a manager commands that email files that are possibly subject to legal action be deleted this can result in criminal charges, but if emails are routinely destroyed then we can claim we were just following policy.
The problem is that the process of decision making (why did we chose to solve a problem a particular way) is often lost when the email message chain gets deleted.
This was not a positive change.
Sign in to Reply
Rupert Goodwins
2/25/2011 2:57 PM EST
When I was embarking on my (sadly abortive) career as an EE, one of the first things I was taught as a student engineer at a big UK defence contractor was "Always document what you're doing in your notebook. We might need it for patent filing, and you might need it when you go back to an old project." Wise words. Not so useful when one's handwriting looks like a spider on acid. However, it was a habit I kept when I turned into a programmer and scrupulously over-commented code, and it saved my backside more than once.
Sign in to Reply
WKetel
2/26/2011 8:50 AM EST
Haldor is certainly correct in his remarks about 5S, which we started referring to as "5F". But when we got a new (idiot) manager, neatness trumped all other considerations. Yes, they would now throw out parts and assemblies just to make things look neat and uncrowded for our Japanese visitors, who were also evidently MBAs, like the new manager. Of course, this was the same manager who announced to a division meeting that they would be reducing engineering staff, and if it turned out that they needed more engineers, then they would just hire them, "since all engineers are the same, and completely interchangeable". The only good that came from that was that we all knew exactly how he felt about our talents, and just how valued we all were. I won't name the manager or company, but it is close to Evergreen and Ten Mile, in Southfield.
Sign in to Reply