Tips on trolley project.

The plant where I work is putting a plating line in. This line will have an overhead hoist. There are 25 prox switches to locate trolley at tank locations. Trolley will travel until it counts prox to location. The hoist will then either pick up or lower a barrel into tank. Then go to next step. Looking for simple tips on doing this in program using compactlogix 1769-l33er. Thanks

RussellRR said:

There are 25 prox switches to locate trolley at tank locations. Trolley will travel until it counts prox to location. The hoist will then either pick up or lower a barrel into tank. Then go to next step.

Click to expand...

Why would you count prox switches -- if each tank has its own prox?
Maybe I'm not understanding you.

I would make it follow something along the lines of:

Trolley passes the first prox and enters the line.

This builds some sort of fifo stack to track the trolley (if the first prox is made there is a part on the trolley)

For each tank:

Trolley hits prox before tank. Stack is shifted forward.

This is where it gets a bit tricky. At this point if it is possible for the part to jam, you will need a timer to verify that the part made it to each position or else stop the line and give an alarm.

Part hits prox above tank.

Trolley drops into tank for allotted timer and bit is set to give state of part.

Timer times out.

Trolley raises.

Trolley moves to next prox. Once trolley hits prox another jam timer.

Stack is shifted to recognize part left prox

Rinse and repeat.

These are just basic ideas I came up with as I typed to help get you started. You also want to get your UDT setup for each trolley/part with status bits to indicate if a part is present. Hope this helps.

I don't understand why you want to count prox switches either.

I have some experience with plating lines, 25-32 stations, with 6-8 hoists in a system. I can't claim I programmed much of it, but I did reprogram the trolley/hoist movement logic so I've got some sense of it.

Depending on how accurate the hoist placement needs to be, you might be able to get away with prox switch detection, however you'll never reliably detect your prox, and stop the trolley/hoist in a a position that maintains the prox, unless you create some larger sensing area. The logic is pretty simple, I'm messing with ST so here you go:


Our trolley/hoists were controlled by VFDs and had High Speed and Low Speed settings, accuracy was important as we approached, but so was speed to ensure throughput. Once we got 'close' to the station we slowed down until the target was reached. We had laser measurement sensors to monitor position in relation to the station position. It also allowed us to determine direction. We also had prox switches that served as a confirmation the target position was reached. If we lost the prox there was logic to jog the hoist slowly until position was reached again. Simply to correct for an overshoot/undershoot situation. We had to be very accurate as the basket was lowered onto a shaft so it could be rotated while at the station.

I would add some type of position feedback if possible, but again if your target tolerance isn't that important you probably can get away with it.

With a single prox at each station, the controller will know exactly which station the trolley is at. I too am confused on why you'd want to count.

Can you put the sensors on the trolley instead of at each station ? If they trolley were equipped with 5 sensors, you could have an array of trigger flags at each location to turn on sensors as a binary code.

00001 = Station 1
00010 = Station 2
00011 = Station 3
...
11111 = Station 31

Paully's5.0 said:

Depending on how accurate the hoist placement needs to be,

Our trollye/hoists were controlled by VFDs and had High Speed and Low Speed settings, accuracy was important as we approached, but so was speed to ensure throughput. Once we got 'close' to the station we slowed down until the target was reached. We had laser measurement sensors to monitor position in relation to the station position. It also allowed us to determine direction. We also had prox switches that served as a confirmation the target position was reached. If we lost the prox there was logic to jog the hoist slowly until position was reached again. Simply to correct for an overshoot/undershoot situation. We had to be very accurate as the basket was lowered onto a shaft so it could be rotated while at the station.

I would add some type of position feedback if possible, but again if your target tolerance isn't that important you probably can get away with it.

Click to expand...

I would consider a shot-pin -- after the slo-mo move to position.

kwade said:

I would consider a shot-pin -- after the slo-mo move to position.

Click to expand...

This was 10 years ago. I'm not even sure if it's in operation! Much of that company got moved to Asia, a shame too I was really proud of myself (6 months out of college). It was an impressive system, a 3rd party contractor came in and made a dynamic scheduling system to maximize throughput. My job was to re-write the logic so the hoists didn't crash into each other given the improved dynamics...took awhile

That contractor, probably the smartest guy I've met in the automation world.

Paully's5.0 said:

This was 10 years ago. Much of that company got moved to Asia,

Click to expand...

Been there, done that!
It is amazing how corporate Instant Gratification means more than future earnings and reputation.

I would not use prox switches for positioning way to much trouble
Use a laser measuring system, then write the code to have a teach mode
Manually move the carriage to each position and then have the PLC store that value
Much simpler and cleaner with far less maintaince to deal with.
This gives you the ability to make minor adjustments without having to climb over the machine
You will need to program at least 2 subprograms
These will be used at every station
1. Drop off the product
2. Pick up the product
The positions will be offset from the cradle position
It’s better to do the offset by position rather than travel time.
Others subprograms may be helpful depending on the application.
Then you will need a sequencer to automatically move the produce from position to position
In the correct order and time.
You should keep track of each positions product present or not to avoid collisions
Other programs may be necessary depending on the needs.

Another simple way of absolute positioning is a barcode or rfid reader on the trolley...

Well people there are a few different options to look at. There is another line here that operates close to the way this one will. The hoist/trolley has the prox mounted on it and the flags are on the rail with another prox mounted beside it for over travel. That one just counts the flags. This one has the flag mounted on hoist/trolley with 25 prox switches. There are no parts to detect just location of trolley to drop/lift barrels at correct plating station. The proxs identify stations. This will start at station 1, pick up barrel, travel to station 9 drop barrel, wait 5min, pick up travel next station 10 drop wait 2 min so forth then back to station 3 and 2 for rinse and then 1 to unload. The system does not go in order one after another. It may need to go from station 10 to 20 then back to 15. Hope this helps you to help me. Thank You all for the positive feed back

RussellRR said:

Well people there are a few different options to look at. There is another line here that operates close to the way this one will. The hoist/trolley has the prox mounted on it and the flags are on the rail with another prox mounted beside it for over travel. That one just counts the flags.

Click to expand...

Does 'That one' hoist/trolley ever 'go to the wrong station?'
----Maybe miss a prox flag, or double-counts a single prox flag.
On a power-cycle does it have to waste time locating one end or the other (HOMING)?

At home all the new lifts are made following this logic, but with bi-stable magnetic switches. Works ok, but at power-up needs initialization of course.

For cranes, especially for long distances outside, where laser won't do, as far as I know usage of gray-code linear encoders is preferred.

Yes sometimes it does miss a flag or loose position. When this happens we take it back to station 1 and there is a reset button and goes back to last step completed.

possible to post a pictures the trolley / tank