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

This forum is full of problems that people are having and possible solutions by the guru's.
But the guru's never really get a chance to 'brag' or expound their skills at programming or - system designing.
So don't be shy chaps; lets hear about a few of your proudest, best jobs.

Jobs where maybe you were really up against it - or even felt a little out of your depth (we all do from time to time)
and not necessarily the really big projects - sometimes a small panel doing a great job can give you (and usually only you) great satisfaction.

I'm sure you can look back on a few jobs and think 'boy I'm really proud of that one'
Time to tell the world how you 'overcame' and produced a fantastic system.
[BRAGGING COMPLETELY ALLOWED] Put your modesty away and let us share your glory
because in the real world we are seldom congratulated.

I'll add mine later.

One thing I did for a customer was modify a SLC5/05 with a iFix HMI that the original programmer noticed the iFix was only capable of displaying one alarm - so if he found an alarm condition he set that alarm and then stopped looking for any more alarms. The customer had one tank of the line (16 dip tanks) removed for repair and the iFix showed Tank 15 Low Level alarm.

However it never looked for any other, and there was an Under-temperature condition in one of the tanks the operator should have been notified of, and they ended up rejecting hundreds of circuit boards.

I added a routine in the 5/05 to check for all alarm conditions, then scrolled through each and every alarm, changing after 5 seconds, then repeating. It is set up to allow all 43 alarm conditions to be displayed in one repeating loop. The customer was ecstatic and the original programmer told me it was impossible and wanted to know how I did it. I had the customer agree to not release my routine to anyone.

A customer was purchasing some robotic water deburr cells but didn't have any onsite robot (Fanuc) support but had a lot of CNC programmers. So I used the A-B PLC/HMI to display and edit G-code and then translated it into commands that I fed to a program I wrote for the robot that used the commands to make the moves. Took a while to debug but worked really good and secured orders for the next 7-8 machines, the last one having two robots.

Good idea for a thread!

I built a SCADA project for a Brewhouse using Ignition. It was my first project doing anything related to brewing, and first project using Ignition. I added all sorts of bell's and whistles to it. I had to create a custom recipe system, on top new visuals, objects and ways to interact. It also used 3 monitors which made it even more challenging. I could base 50% of it or so on our company standard. But Brewhouse's need more process detail and interaction than we typically dealt with. Being the first iteration of it, I'm pretty damn proud of it. Towards the end of it, I realized it wouldn't take much more to create a true batching system similar to WonderWare InBatch.

We actually had 2 brewhouse projects going on at the same time by somewhat separate project teams. Customer "B" required Wonderware System Platform, project team "B" worked on this. My project, project "A" was originally Wonderware InTouch. I laughed at my boss for selling InTouch on this system because there was no way it could handle the recipe management side of things. Not to mention the painstaking process it would take to build the extra's a Brewhouse would need. While project "B" was about 20% larger in scope than my project "A", I built the SCADA system pretty much solo, in about ~800 hours. Project "B" had a team of engineers working on the SCADA, and had to basically start over when the customer asked Wonderware Development to get involved to assist. They easily had 1500+ hours into it and who knows how much money the customer got taken on for WonderWare consulting fees.

Needless to say, my solution, hands down was better than project team "B". By miles.

oh GOD... where do I start, I have gotten lucky a few times...

I added UPS's to a couple of large presses, every time the power would glitch it would cost about 20k, the PLC's would lose the place in the cycle/recipe and open after the power would come back on, living in the south power spikes ot fast outages happen a lot, so this ended up saving the company about 200k per year

Another one was a energy audit, at this plant the gas bill was about 75k per month and electrical was about 150k per month, by tuning the system and a few modifications we were about to save about 30k per month after all was said and done, took about 2 months to do but well worth the time and effort

Great thread!

I design a three flying cutoff for a pipe mills based on controllogix, two of them are hydraulics and the other one is using an ACS800 drive working as a servo drive. I include PV700 plus as hmi interface in all cases.
The bad part is that they work so fine that they never call me back.

Had a new customer with 2 obsolete displays (1 was already broken) that managed about 50 pumps via a AB SLC500. They were vital to the operation the customer was running.
Asked them why they didn't change the displays to something newer, and was told that the company that had supplied the solution had went out of business. In their world it was an impossible problem to solve, especially since their old electrician had told them that there was nothing that could be done except change all hardware including the PLC.

I had never touched any Rockwell hardware before, and hadn't done any ladder since my schooldays some +10 years before.
Bought a SLC500 from Ebay, learned Logix500 and found AdvancedHMI. 2 weeks later the old displays were gone and the PLC program was rewritten from scratch (the original was what you call spaghetti programming). Also added handling of alarms that the old displays didn't have.

I was a king in the eyes of the customer

Very first clx project. Also first project using rio. 1150 I/O points, and 11 PF70 drives on ethernet. I think I spent two hours reading manuals for every hour programming and wiring. Stripped the 16 Omron C20's running the machine and installed all the new hardware (me and two electrians) , while relocating the entire line in 11 very long days. Usable production in 5 days after power up. Full production took another 6 days. According to my wife I was mumbling UDT in my sleep for two weeks after the job was completed.
Oh, and we were pulling 60 words in from a SLC500 for distribution into tags for processing by Wonderware. This was done by a programmer in England via skype. Talk about a steep learning curve.

AND NONE OF IT WOULD HAVE BEEN POSSIBLE WITHOUT HELP FROM THE GURUS ON THIS SITE.

Special thanks tn Ken Roach among others.

Similar to Soggy Canuck, I had to learn RS Logix 5000 programming, write the program from scratch and build an HMI, all in just over a month. Then myself and one other electrician replaced the control system on a machine with 5 remote I/O racks over 9 days. We had production back on day 10, with some bugs. It took a few more days and a visit from an AB tech to get them out. When I installed Studio 5000 I had no clue how to make a tag or anything. Thanks to Youtube and this site, I was able to get it going.

My job in the metal shop was to clean out the sludge tanks, cut up and sort the scrap into appropriate barrels, and use a TIG welder to repair the cosmetic notches in flange plates when the plasma cutter came on at the wrong time during a transit between parts. I was sixteen and loved the noise and flame and power of the metal shop.

I pestered the operator to show me how G-code was loaded on diskette into the Burny 5 CNC controller, and looked up which codes commanded the torch to ignite, and figured out that they were getting incorrectly injected into the program when it was created on a digitizer up in the engineering office.

I figured out how to automatically parse the program to find those torch command and how to edit it so that the errors wouldn't happen anymore. I typed up the process and submitted it to the lead draftsman on my last day at the end of the summer.

I came back the next Monday, and peeked into the drafting office. The lead draftsman wasn't there, but another guy told me he had read my technote and said "that little **** thinks he's so smart". The technote had been torn in half and was in the trashcan.

I went downstairs and found a three-foot wide piece of quarter-inch stainless that had been ruined by a stray torch command. I cleaned up the slag on the edges and took it over to the corner of the shop and spent two hours with an air engraver, painstakingly replicating my technote in steel.

I left that on the lead draftsman's desk, picked up my final check from the payroll office, and headed off to college.

Ken Roach said:

My job in the metal shop was to clean out the sludge tanks, cut up and sort the scrap into appropriate barrels, and use a TIG welder to repair the cosmetic notches in flange plates when the plasma cutter came on at the wrong time during a transit between parts. I was sixteen and loved the noise and flame and power of the metal shop.

I pestered the operator to show me how G-code was loaded on diskette into the Burny 5 CNC controller, and looked up which codes commanded the torch to ignite, and figured out that they were getting incorrectly injected into the program when it was created on a digitizer up in the engineering office.

I figured out how to automatically parse the program to find those torch command and how to edit it so that the errors wouldn't happen anymore. I typed up the process and submitted it to the lead draftsman on my last day at the end of the summer.

I came back the next Monday, and peeked into the drafting office. The lead draftsman wasn't there, but another guy told me he had read my technote and said "that little **** thinks he's so smart". The technote had been torn in half and was in the trashcan.

I went downstairs and found a three-foot wide piece of quarter-inch stainless that had been ruined by a stray torch command. I cleaned up the slag on the edges and took it over to the corner of the shop and spent two hours with an air engraver, painstakingly replicating my technote in steel.

I left that on the lead draftsman's desk, picked up my final check from the payroll office, and headed off to college.

Click to expand...

I LOVE THIS! Very well done.

I dont think i can contend with the Maters on here, but my favorite was when i found a sticky HMI keys from a Quickpanel causing havoc in a program. fixed it with some oneshots, and made a repeat customer afterwards

Ken Roach said:

....

Click to expand...

That was great!

I have a client who has VantagePoint spit out a handful of reports every morning at 5am. I gave their IT guy a call and had him automatically email the reports to a list of people at 6am. Everyone happy.

Then, they want a couple of new reports. I set them up to generate, and give the IT guy a call to add them to the list. Unfortunately, the IT guy I'd dealt with last time has left. No worries, I'm sure the remaining IT guy can sort it out. Or so I thought. 6 weeks later, the client is getting increasingly agitated at me because he still doesn't have his reports, the IT guy is completely stumped because apparently the guy who left "was a script whiz kid, and none of us know much scripting". Even after ringing the previous IT guy and finding the script, he's still unable to understand it enough to modify it and make it work.

Eventually, I give up and spend an afternoon with google learning how to use a powershell script to search a directory for new files, attach them to an email, and send them at a scheduled time each day. Once I have it working, I make a call to the client. In an immensely satisfying stroke of luck, the IT guy just happens to be right next to him when he takes the call. I tell him I've taught myself how to do it, and to check his email - he checks his email, confirms it has all come through, and turns to the IT guy and tells him that I've worked it out. In the background, all I hear is "he's what? The clever little sh*t!"

I felt good about that one for the rest of the week

Programmed a PLC-5 for a machine that assembled key locks for doors. Locks are loaded onto a pallet (with RFID tag on it) in pairs. Then they go around the machine and get all the different parts put in. Then get unloaded in a station with 22 lanes where the locks are sorted as they are unloaded. The challenging part was when they did a work order that has dozens of locks that have what they call a Construction key and a Master key, like a large building or apartment complex. With those they had to be unloaded in the correct lane grouped by building code and unit codes. Had to build a bubble sort routine (in PLC-5 ladder) to sort them.

After work a few years back I stopped by a used machinery dealer to say hi. While walking around his warehouse he pointed out a machine he lost $10,000 on & was going to be melted down for scrap steel the following day. He paid $5,000 for it and spent another $5,000 rebuilding it, had a customer that wanted to buy it for $30,000 - but it had to be running. His guy worked on it for 3 months and he finally resigned himself to scrap it out. I said I wasn't doing anything & that I could look at it.

It took a full 2 hours tracing unmarked red wires (the schematic was inside) to find a SPDT micro-switch located in the motorized Bridgeport type table was wired wrong. Moved 3 wires and he was $30,000 richer and ecstatic.