Oh, I have an idea. Not so sure there is always a black and white line, some things are multifaceted and it may depend on how forum members are likely to react to some particular aspect. For example, you have mentioned some cultural differences in engineering in some parts of the world as compared to here. Didn't seem to raise political hackles. It was more of an engineering story or that is how I took it.
I should have read that description of Poe's law more carefully. I made a sarcastic comment that had nothing to do with engineering and went over everyone's head. let's just drop it I apologize for the distraction.
Yah, wonderful.
In the last month I have "explained" to those newly minted engineers;
1. An involute ring and spur gear cannot be pushed together with a spring, it causes the teeth to bottom out in the root. With the encoder after the ring gear, it is not possible to close the feedback control loop and be stable.
2. Increasing the number of micro steps does NOT lower the torque produced by a stepper motor, just the incremental torque of a single micro step.
3. What a Bode plot is. Believe it or not, controls engineers with masters degrees have no idea what a bode plot is. Nor do they have a clue as to what phase margin is.
4. They've no clue that the torque against either a stepper or 3 phase brushless motor changes the frequency response of a physical system. As a result, optimal tuning is dependent on the reaction force against the physical system.
5. Devices which move magnets closer together or farther apart will see non linear forces. Trying to use linear algorithms for motion control will NOT work.
6. It is not possible to fix bad hardware by tweaking software!!!
7. If you do not have any hardware experience, how the h##l are you going to learn how to control complex hardware with a computer???
Now, what were you saying about the no brainier part of replacing actual experienced engineers with wet behind the ear video gamers????
...You run a company so I know you understand, that rant was not about you Ed..
...the list presented is a subset..my presentation on what needs teaching is already over 150 slides, it gets larger by the day..
Jn
Last edited:
isnt it the schools object today to get more income... student fees, grants, subsidies etc and Not a high level of education?
So, that students leave uneducated but get A's so the school looks better and make more money. Its all a shell game now and about money.
Sorry... my own rant.
I dont think analog is being taught any more for a long time now.
-Richard
So, that students leave uneducated but get A's so the school looks better and make more money. Its all a shell game now and about money.
Sorry... my own rant.
I dont think analog is being taught any more for a long time now.
-Richard
Last edited:
I know your sarcasm.... and I got it. It wasnt nice, as usual, nor funny. So, I just ignored it and spun it my own way. And answered literally.
-Richard
I know but it you give them a transfer function they know how to use Hilbert tranforms to determine controllability. It is apparently assumed they will learn it from someone like you on the job, or they will learn it somewhere along with way during their PhD and dissertation activities.
Last edited:
#3 one surprises me. I couldn't have completed my undergrad without coming across a Bode plot, and my degree has Computer in the name.
#6 does not. I have to routinely address this with experienced people and management. It's usually the people responsible for the bad hardware asking
.
This is not a complete solution. BL modulation will impact the pickup coil. The B will change when the magnetic field is impacted by fluctuating i through the VC. L is not linear between Xmax and Xmin. So it would work for linearization of the FR, but not for reduction of some major distortion inducing factors.
I guess, I must live a different life than most. It has been fairly exciting at times. Some of it my own doing. Skip the 3 years in USArmy -- where I made 5 grades/promotions in first 18 months and was again in charge of men.
I was assigned to Paul Rupert on a new project he was in charge of. PR was from NASA and was director there on several early manned shuttle missions and was the highest paid manager at LLNL.
We had a rush project that was funded for one year and time was flying by.... every 15 minutes one day he was yelling my name down the hall to me. Each time he wanted to know when the custom designed equipment would arrive as their installation would take months. He need about 75 of them. Only 35 were completed.
My contract drafts man had a commercial pilots license. he flew to work and home every day bringing a couple of other contract employees with him. He suggested, if I wanted to, he would fly me to the factory and we bring as many units as we could back. This would put the techs to work installing and get us on time.
Hey, what could go wrong? Plane crash or ? LLNL sued for milllions? I gave him the go-ahead and I would go with him. he rented a plane, we got it fueled the next morning and were off flying down the middle of California. We radio ahead and the company met us at the airport we landed at with a full load of completed modules each bubble wrapped. in the back of the plane they went and we were off back to LLNL.
Paul was so happy we got his units and could proceed with the project. .. He never asked me how I got them. The pilot (draftsman) said, how are we going to reimburse him for the plane and fuel. He thought, silly man, that i would pay for it just for my peace of mind.
I told him to send the bill to LLNL to pay. He did and about 2 weeks later I am called into Paul's office.... Did you fly a plane down there without prior approval and authorization? Yes. blah blah blah. I said what are you going to do, fire me. Go ahead. I did it for the cause.
The project was important to LLNL or they wouldnt have put Paul Rupert on it. They didnt fire me. In fact, I got the largest raise that year ever.
Sometimes, you have to know when you can bend the rules. But, it better turn out OK if you do.
-Richard
I was assigned to Paul Rupert on a new project he was in charge of. PR was from NASA and was director there on several early manned shuttle missions and was the highest paid manager at LLNL.
We had a rush project that was funded for one year and time was flying by.... every 15 minutes one day he was yelling my name down the hall to me. Each time he wanted to know when the custom designed equipment would arrive as their installation would take months. He need about 75 of them. Only 35 were completed.
My contract drafts man had a commercial pilots license.
he flew to work and home every day bringing a couple of other contract employees with him. He suggested, if I wanted to, he would fly me to the factory and we bring as many units as we could back. This would put the techs to work installing and get us on time. Hey, what could go wrong? Plane crash or ? LLNL sued for milllions? I gave him the go-ahead and I would go with him. he rented a plane, we got it fueled the next morning and were off flying down the middle of California. We radio ahead and the company met us at the airport we landed at with a full load of completed modules each bubble wrapped. in the back of the plane they went and we were off back to LLNL.
Paul was so happy we got his units and could proceed with the project. .. He never asked me how I got them. The pilot (draftsman) said, how are we going to reimburse him for the plane and fuel. He thought, silly man, that i would pay for it just for my peace of mind.
I told him to send the bill to LLNL to pay. He did and about 2 weeks later I am called into Paul's office.... Did you fly a plane down there without prior approval and authorization? Yes. blah blah blah. I said what are you going to do, fire me. Go ahead. I did it for the cause.
The project was important to LLNL or they wouldnt have put Paul Rupert on it. They didnt fire me. In fact, I got the largest raise that year ever.
Sometimes, you have to know when you can bend the rules. But, it better turn out OK if you do.
-Richard
Last edited:
May-be the best cook in the world ? (For the ones that love chiles).
While french food is made from chords, to make a musical parallel, the Thai food is made of harpegies. All perfumes and tastes coming one after the other, continuing to play in your mouth long time after you finished your mouthful.
And these melodies are so subtil !!!
Yes. Most companies survive by people bending the rules, because when you follow the rules too closely nothing useful gets done. This may be because the rules are meant to control and systematize what you were doing last year, not what you need to do this year.RNMarsh said:
The phrase "it better turn out OK" triggered a memory. Many years ago I was asked to lead a small team putting in a new feature into complex software (used to control power station plant). After some thought I told management that this was a high risk project. By this I meant
1. it might not work, we might be unable to find a way to do it - we might have to abandon the project
2. it might cause unacceptable side effects in related software - we might have to abandon the project
3. it might need far more effort than planned - we might have to abandon the project
I was given the go-ahead, but after some months I had to report that it wasn't working, and reminded management that I had told them that it was high risk. They were surprised, because when I said high risk they thought I meant
4. it might need just a bit more thought and effort than other projects
It didn't occur to them that high risk might mean failure.
Bending rules is one thing. Renting a private plane without approvals on a govt project these days would be hard to hide. I know BMW have chartered helicopters to collect parts from stranded lorries to keep production lines running mind. But that is a simple risk/reward calculation as beany mcbeancouter will know exactly how much it costs for production to stop for n hours.
Unintended consequences are the bane of software development. Software products accrete features and fixes, and the whole creaking mass begins to diverge from thevoriginal design documents. Developers are uhder pressure for deliverables, so going back to fix bugs is an unwelcome chore. We have dev, test, qa, and uat environments, and still fixes go into production and have unexpected side effects, because although they deployed in uat they didn't do exhaustive regression testing.
I think part of the problem is that the s/w has well defined data structures and interfaces, bug fixes start to look more and more like hacks that bypass those interfaces (and of coursecare not documented).
I think part of the problem is that the s/w has well defined data structures and interfaces, bug fixes start to look more and more like hacks that bypass those interfaces (and of coursecare not documented).
- Status
- Not open for further replies.