Lets do a bit of 'chest beating' for a change.

[Peter Nachtwey]

http://www.controleng.com/single-ar...c-controls-for-runway-testing/e17b8b3882.html
Someone else on this forum was involved in replacing the drives that move the machine up and down the rails. I have driven that machine although it is no big deal since it can only roll on the tracks.
The hydraulic motion controllers are controlling the force on the runway in an effort to quickly see how a runway will wear out and how to make better runways.
This project was very complex but it worked the way I thought it would until we discovered that the whole contraption flexes. Then we had to get a laser measuring device to detect the flex of the fram.
I can explain more if anybody is interested.

http://machinedesign.com/hydraulics/electrohydraulics-drives-engine-research
The department of energy was trying to make their high pressure fuel injection system work for two years. What they didn't realize is that the hydraulic designer had screwed up the design terribly. I should have been consulted on the original design. However, I was paid for 15 hours to simulate the system. There were about 30 simultaneous linear and non-linear differential equations in the simulation. I was able to tell how the system would run and predict the gains so the system did work. What was good is that the Department of Energy used Mathcad. Since they paid for the simulation I gave them the worksheets so they can do their one predictions, modelings, and estimates in the future.

I have many more. Most of the controllers we sell are installed without any problems. There is always the 2% that are mostly due to poor mechanical and hydraulic design. Of those about 1 in 50 is just a difficult project that requires lots of real engineering where the customer or original designer is over their heads.

I am a math, physics and control geek ok? I can program PLCs but it isn't what I like to do. I prefer look at the bigger picture getting into the physics of the system and algorithms required to control them.

 

[Ronnie Sullivan]

One of mine that although not overly technical - gave me immense pleasure for several reasons and it all came about by a casual comment.

A director at a galvanizing dipping company was complaining about the huge monthly gas bills and the maintenance bill for the 'dipping pot'.
This dipping pot is massive and filled with molten zinc. (thousands of tons) They galvanize huge steel girders (10 at a time) and car sub frames among other things.

The control panel was 50 years old and always breaking down. There were no drawings and it was a nightmare to fix because time was a factor as the zinc was cooling. (if it cools too low and the zinc sets - it takes days to gently heat it back to working temperature)

I said to him that I could probably save him money on both gas and maintenance with a new panel. He absolutely jumped on it 'get it done' no price was asked for - just get it done.

In retrospect I can now see why he said that. They were losing customers because of the pot constantly being down - and the bills were crippling them too.

It's not often you get a job where money is no object so for starters we had a top of the range PLC and a large hi-res HMI. The only fly in the ointment was - i'd told him I could save him gas..........without really thinking how. Luckily, the overriding factor was reliability with no breakdowns.

I spent weeks with the new panel programming lots and lots of 'new features' into it and bench testing until I thought it was perfect.
It had four burner controllers and the only thing I couldn't test was 'flame detection' until I had a brainwave. I connected an ultraviolet flame sensor and when it was time to light the gas I held a lit cigarette lighter in front of the sensor.
The controller saw the flame and continued to running position and the millisecond I took the lighter away, it locked out.
Come the big day, I was nervous. We had to take the old panel out, remove the wiring carefully marking it all and then put the new one in its place. All this while the pot was cooling. Not a time to find you have made a mistake or forgotten something.

And not to add pressure or anything but the bosses were all milling about asking questions which mainly consisted of 'how long?'

3 hours in and it was time for the first ever run of the new panel (a one button press instead of multi settings and buttons as before) We now had an audience of the entire factory. The forced air fans started correctly and it went through it's purging routine (which btw takes absolutely ages when you are nervous)

It cycled through beautifully and all the burners lit and started modulating. Oh Joy!

To save gas I consulted a gas engineer. He came up with lots of great ideas. The main one being that during none working hours The pot can automatically be allowed to run below working temperature and brought back to temperature as the shift began.

Before, it ran at working temperature 24/7 whether it was being used or not.

The savings on the gas alone paid for the panel within four months.

 

[sigmadelta]

I had a client with a soft-PLC (Siemens WinAC RTX) and a custom-written HMI in C++ language. The machine that the client bought never ran properly, and the manufacturer went bankrupt. It turns out I was the 11th-hour attempt to get it going. The only reason the project landed on me is because someone knew someone who knew someone... (3 more times)... who knew me.

I had no source code for the C++ HMI, only the .exe file. Press start on the HMI, nothing goes. I had no way to find out which PLC addresses were being used by the HMI, until I found a way to decompile the C++ program into assembly language, but some parts were decompiled into actual "understandable" code. Thankfully, the understandable parts contained PLC addresses! So I was able to figure out that unless a certain PLC bit was set, most of the HMI functions would not work. I made this bit to be always on in the PLC, and the HMI was working 100% normal!

 

[deanfran]

Not nearly as cool as some of these stories, but we are kind of proud of it. We use Blow Fill Seal machines for liquid filling. The equipment we have is early 1990s era vintage. One of the machines had several moulds 4,6,8,10,12 and 16 up, and the control electronics to match. In the original control design, there were two circuit boards, board 1 controlled filling of container 1 - 8, and board 2 handled 9 - 16. We had never filled more than 8 containers. We got the opportunity to manufacture a product that required the 10 up mould. The first time the lead operator tried to fire up the machine, lo and behold the second board was completely non functional. The OEM didn't have replacements anymore, but was happy to upgrade the electronics to the tune of about $100,000.The board was a Brown Boveri, and ABB no longer supported it either. Sitting in a meeting with the bosses, the lead operator said, "We could build our own control for containers 9 and 10 with a PLC, for a lot less than $100,000" Of course the bosses said, "Great can you have this running this afternoon?" One small problem, no one had any idea of even where to begin, so they all looked at me. Thus began my journey into industrial control. 3 weeks later, the machine had a CLICK PLC, a C-More Micro HMI, the little system was qualified, and the job done for less than $500. We later upgraded it to handle 12 up filling, and it has been filling containers 16 hours a day, 5 and 6 days a week for about 3 years.

 

[RussB]

A bit OT...

I found a baby Robin beneath a bush near our office door and after looking into the bush spotted the nest with a baby Cow-bird in it about twice the size of the baby Robin. I tossed the baby Cow-bird out and put the baby Robin back in the nest.
I don't know what happened to the cow-bird but the Robin grew to maturity.

 

[mort81]

Home brew code

Two of mine that stand out are HMI related. The first being a large Fix32 to iFIX conversion. Anyone that has done these type of projects knows that the converter tools from GE (or at least at that time) are only going to get you so far. I custom wrote some VBA within iFIX to basically take the old "report" files from fix32 and fix (no pun intended) the issues the converter does not. I also added a feature to have the new VBA of the converted screens do an automatic de-compile to check for errors. The one exe for the converter has issues and I had several open cases with GE, but they never gave solutions and would not share the source for the exe. The second is a self created .net program that will take two wonderware Intouch directories and compare them for differences. This is like the MDT software. It actually takes the files and converts them to binary and creates hash codes out of the files to check for true differences. By doing this it is why you don't have to be worried about false time stamps and etc. Worked perfect for keeping taps on live systems that projects are going on to make sure you are developing on the most current screens and scripts. Obviously now with app server this has become somewhat obsolete, but still may run into 9.5 or earlier from time to time.

The reason i'm proud of them, is i learned (or self tought as it were) vba through cutting my teeth with iFIX in my early days, and these two programs really streched me. But credit due where it should be, there is no way either one of them could have been done without google!

 

[Jeev]

I was working on a project on site in New Zealand, and found myself with a very rare coupe of Sundays off. The tiny motel I stayed at was owned by a lovely elder couple who didn't know a great deal of tech, and the lack of reliable internet access was driving me nuts.

On my first Sunday I traced their LAN problems back to a dead 24-port switch in the rack. I drove to Auckland and bought a new switch, replaced it and BAM LAN to the rooms. On my second Sunday I reconfigured their restaurant wireless modem/router, went back in to Auckland, bought some access points and fitted the place out with a decent wireless network that actually reached the rooms down the end.

Their connection itself was surprisingly fast, the owners paid for all the hardware in cash, and I now get my choice of 4 beers on tap for free when I stay there. This was simple and more rewarding than any of the big domestic or international projects I worked on last year

 

[mk42]

My on-topic proud story is of my first machine. I can only imagine the grievous programming errors I made back then, and the basic lessons I learned along the way. None of this is earth shattering to anyone who had done it before, but for someone who is essentially teaching himself on the fly, it was all new.

When I was starting out after college, I was in a sort of rotational program. I had gotten some PLC training, but I really hadn't done much useful, and was more or less a marketing intern at the time. They planned to rotate me into some other groups later, like engineering.

My boss called me into his office one day, said "hey, you're planning to head to your parents' for Easter next month, right?" I had been thinking about it, but they were two states away, and it was a heck of a drive. "How about you stop at our other factory on the way back? You can rent a car for the whole thing and expense the gas, because this is too many miles for you to expense on your own car."

"Uh, sure thing, boss. Free gas sounds great to me, what's up?" Apparently one of the maintenance guys/operators was trying to automate an electrical testing machine as a pet project, and the engineering dept didn't think it was worth putting any "real" resources to help out. I called the guy, and he seemed like he had a good understanding of the process. They wanted to re-purpose the HW (PLC, relays, a couple pushbuttons) from an old scrapped tester for a new set of tests. He had gotten started on some code, but quickly realized he was over his head. I don't have that much more experience than him, but whatever, I'll try to help. I block out the week after Easter to try to get this thing going. Downside: I'll more or less be the only person in the building with any PLC training, and I don't know anyone to call who knows more than me. As I said, my manager was marketing.

Knowing I'm kinda set up for failure, but wanting to give it a shot anyway, I arrive on site and we figure out what needs to be done. Step 1, prove that the PLC can actually perform the different tests. Step 2, write a sequence to run those tests in series, and report any failures.

It took about a day and a half to prove out the the different tests. We had to turn on THIS relay to power up a board in one mode, then turn on another relay to activate another mode. Then run power through the board for awhile, and make sure it didn't overheat.

It took another day to actually automate those tests. There were quite a few steps that we were mentally performing that we didn't know about. Its amazing how easy it is for something like "wait 5 seconds" to be vitally important to the process, and you have no idea when you do it by hand, because it takes 10 sec anyway.

Next I started writing the sequence. It ended up being something like 70 steps across 10 sets of tests. I learned very quickly that it was harder than I expected to write the correct transition logic. I kept looking for conditions that were already true when the step started, or that would accidentally trigger if the step failed.

It took some late nights (I think Thursday night they pretty much had to kick me out, and I only barely made it to wendys for dinner before it closed) but I got it done, even with no backup. Amazing feeling. It gave me a lot of confidence rolling into my engineering rotation, and things took off from there.

Lessons I learned here:
1) Using integers for the steps in a sequence, with another integer for "next step", is a fantastic idea. I figured out that on my own, and am a bit proud of it.
1b) I didn't learn until waaaay too late that I need to leave lots of room for new steps in the middle. You would have thought that after renumbering the whole thing for the 4th time I would have started leaving gaps, but no.
2) Just because the other guy understands the process doesn't mean that he can explain it to me well. Getting things written down and clarified is clutch.
3) Getting operator buy-in is key. If the operation of this machine were my idea and I had to explain it to him, I don't think there is any way he would have used the junk machine we ended up with. Instead it was his idea, and he loved it.
4) I CAN DO THIS!

 

[mk42]

My off topic proud story is from when I was 17 in Boy Scouts. The whole district had a big campout with a bunch of troops, where we had a junkyard wars type of event on Saturday.

I had marching band on Friday night, so I didn't get to the camp until late morning Saturday. I arrive and find out the task for the day is BUILDING CATAPULTS! I was assigned to a team that was all very young. The oldest kid (who had been leading) had probably only been in scouts for a couple years (14 at the oldest), and he wasn't doing a bad job, but definitely felt over his head. They had a good design going, so I didn't really try to take over. I just gave a little bit of advice: we needed to tighten our knots (to waste less energy), we could add a counterweight, things like that. Then I dug in and started helping while the 14 year old more or less kept supervising.

We were the last group to launch. When they started at the other end of the row, the dads were all standing waaay back, because they didn't know how what to expect. They collectively took a couple steps up as each catapult in the line made a short throw, or misfired, or fell apart. You could see by the time they got to us, they weren't expecting much. We loaded up and pulled the rope, and our catapult shot OVER THEIR HEADS, at least double the distance of the next closest team.

-----------------

That was awesome, but it isn't the part of the story I'm most proud of. Years later, I heard someone giving a speech on leadership. He told a story of how, back when he was a new scout, how he went on this fun campout where they were building catapults. Almost everyone on the team was young like him. The Dads helped them move the logs, but they really didn't know what they were doing. Then, halfway through, an older scout showed up, pulled the team together, gave them confidence, and lead them to an overwhelming victory!

It's really humbling to realize that someone you didn't even know (and who didn't know you) has been looking up to you for years. I also learned that leadership is much more about feeling lead than it is the leader.

 

[dmroeder]

My crowning achievement was when I put a BSOD bitmap on a PanelView, made it the only thing on a screen (full screen), then when the user deleted a product out of a database from the HMI, it would switch to the BSOD screen and look like the PanelView crashed. I was pretty proud of that, but my bar is low.

 

[ASF]

My crowning achievement was when I put a BSOD bitmap on a PanelView, made it the only thing on a screen (full screen), then when the user deleted a product out of a database from the HMI, it would switch to the BSOD screen and look like the PanelView crashed. I was pretty proud of that, but my bar is low.

I have to ask...why?

 

[Jeev]

I have to ask...why?

Have you met operators?

 

[dmroeder]

I have to ask...why?

I'll answer your question with a question... why not? Square.

Actually, I'm just that immature. Or my sense of humor is that lame.

 

[Doug-P]

After upgrading to PLC-5s for our conveyor control the Exor interface panels had to be upgraded too - they weren't available anymore. Did some research on HMIs and settled on National Instrument's Lookout. Spent some time on the learning curve but eventually got going with color graphics, multiplexed system layout on screen vs. printed graphic, mouse, added new capabilities and notifications, help screens, etc.

I learned a lot, it was pretty and useful and, the operators really liked the new ease of use - the originals were, shall we say, clunky.

A low-tech success was retrofitting the mechanical cart sensors for the in-floor towline with a separate A-D PLC dedicated to sensing cart/no cart for traffic control. The mechanical approach could fail rather dramatically!

 

[ASF]

I'll answer your question with a question... why not? Square.

Actually, I'm just that immature. Or my sense of humor is that lame.

Fair call. I mean I would probably have found it hilarious as well, but I just wondered if it was in response to an equally amusing "incident"