Nerd Stories

Discussion in 'Free Thoughts' started by Ivan Seeking, Jul 29, 2016.

  1. Ivan Seeking Registered Senior Member

    Messages:
    957
    This came up the other day and it occurred to me that I've never had a chance to tell anyone the entire story. And it's a good story! So here you go.

    A job that significantly advanced my career was done for a large manufacturing company while I was still working as a technical support specialist, many years ago. They made vibratory conveyor systems that consisted of long vibrating tables. The motion of the table creates a preference for the direction of motion of the product being conveyed. There was a system [one specific model] that no one had ever been able to get under control. High impulse loads sent it into fits. The motion amplitude variations would swing wildly out of control until the safeties shut it down.

    I took the program and schematics and started a review. Before long I had found the problem. The analog inputs that read the position sensors of the conveyor had a sample rate that was too slow. The frequency of motion of the table was about equal to the sample rate of their controller. So there was no way to know the actual amplitude of the motion. No problem! I designed a little circuit to trap the maximum position of the table and then read that with the controller. They also had a problem with the PID control [an equation used to control motion]. I had been given a long list of failure modes caused by resonances that that had to be anticipated, that would trigger an escape routine. These routines would allow us to escape the failure mode and regain control of the motion of the table without it shutting down first. So I wrote a very complicated PID routine by hand that accounted for all of these alleged failure modes.

    Finally the day came for the big test. I started the system and showed that we had basic control. Then the customer literally jumped on the table to simulate a high impulse load during production. This was considered to be the ultimate test. The table lost a bit of amplitude for a moment and then went right back to nominal values. Everyone was extremely impressed!!! This was the first time this system had ever passed the load test after ten years of people fighting it. Problem solved and I was a rock star for a time.

    This is the kicker to the story. As I had more time to review what was happening in the control systems, I found something too funny to believe. The key to keeping the table under control was to do nothing at all. In order to keep the system stable, the output from the controller was never changing. The PID was just sitting there on one value. In other words, they could have just used a potentiometer and set the motion, and left it alone. The product load never had any real effect beyond minor transients that sent the controller into fits. All of the years of efforts to control the system had led everyone down the wrong path. The only good control was no control at all.
     
    Last edited: Jul 29, 2016
    Confused2 likes this.
  2. Guest Guest Advertisement



    to hide all adverts.
  3. cluelusshusbund + Public Dilemma + Valued Senior Member

    Messages:
    7,985
    Grate story... i can emagine you'r joy in solvin the issue

    Please Register or Log in to view the hidden image!



    At a job i worked at... an electronic part that was very expensive... in high volume use... an coudnt be replaced wit much cheeper parts because a necessary adjustment was unstable when usin the cheeper parts.!!!
    This went on for about a year... an then i checked out the cheeper parts an discovered that the necessary equipment adjustment was unstable because the cheeper parts required that a diferent area on the potentiometer be used for the proper adjustment... an that particular area on the pot was a bit dirty an wasnt makin good contact... so the cheeper parts wasnt defective... the pot on the test equipment was defective... so i was a "rock star" wit the company... but not so much wit some other guys who had kept tellin the parts department that the cheep parts woudnt work... lol.!!!

    Anuther time a new branch sent us some equipment that was intermittent an they coudnt find the prollem... an it was a prollem that i had solved years earlier... which was a transistor wit a perfectly good lookin solder joint was actually bad an just needed to be resoldered (It was prolly caused by corrosion on the transistor legs when they was first used?)... an the branch company was tickled pink

    Please Register or Log in to view the hidden image!



    It was a fun job wit lots of little wins over the 19 years... but after 4 years of me an wife ownin our own bidness -- havin both jobs was to much so i quit... an then bein self employed an semi retired for 20 years was almost as good as bein fully retired

    Please Register or Log in to view the hidden image!

     
    Ivan Seeking likes this.
  4. Guest Guest Advertisement



    to hide all adverts.
  5. spidergoat pubic diorama Valued Senior Member

    Messages:
    54,036
    Wow, that's amazing! Love to hear stuff like that.

    When I was about 17, I had a temp job at a company that made biomedical products from cattle. Fetal bovine serum and the like. My job was to fill metal racks with test tubes that would later be sterilized. I worked at a table with some long term employees, all older ladies. They had a method for filling the racks, they would gather a bunch of tubes together in their hands, and work them around until they fell into the holes part way, then turn the whole rack upside down and tamp them all down completely against the table. It made quite a racket, and they often broke the tubes. After using their method for several days, I got bored and did a test. I timed how long it took them to fill a rack. Then I tried inserting one test tube at a time completely down in the hole. Amazingly, I cut about 10 seconds off their time consistently. So I starting doing it that way. I got fired. Either because my process looked slower to the supervisor, or because I had rebelled against the veterans, he wouldn't explain.
     
  6. Guest Guest Advertisement



    to hide all adverts.

Share This Page