Tracking Problems with AB Micrologix

10dulkar,

You probably will want the resolution to stay high. Instead of changing your encoder resolution, simply make the length of the BSR's greater, until you have a space in PLC memory equal to the number of encoder pulses between the sensor and the dispenser, up to 4096. If you can set your encoder to give 1 pulse = 1" of conveyor travel, you will be able to achieve excellent control. For example if it is 4096, then set the BSR Length = 4096, and set File B3 Properties to have 256 words, which means that the next BSR will need to skip over to another B File. If you run out of memory in one Binary file, then add as many Binary type Files as you need to get enough space for the BSRs. For esample in one program for a shipping conveyor routing boxes to different branch lines, I used Files B9 up to B24 as space for the Bit Shift Registers. Each BSR used TWO Binary Files, each with maximum size of 256 words.

Dandrade,


In his first post, 10dulkar says:

"There are 2 can sizes small & medium. And the plc knows which can size is being fed."

I suppose if the PLC knows which cans are small medium, then this information can be used to determine which BSR tracks which can. Otherwise, a routine is needed to determine the size of each can. You are correct in that the number of "1" pulses can be used to measure the width of the cans, because each encoder pulse will represent a specific distance traveled by the conveyor. The number of bits in the Bit Shift Registers depends on how many encoder pulses occur between the Sensor and the Dispenser.

10dulkar, You said that you only have to worry about triggering the dispenser. Doesn't the dispenser have to stay on longer for the Medium sized cans? I think that somehow you have to differentiate between a Small Fill and a Medium Fill.

Sorry guys I have been MIA as I have been out of town at a client site. You'd think they'd have some form of high-speed internet access in every corner of the world by now...

Anyways unfortunately the conveyor isnt here yet so I havent been able to try this yet. But to answer Lancie1's question the dispense is one shot and programmed at the dispenser. Hence I dont have to worry about keeping the dispenser on for different times...

I have written programs to simulate both suggestions above. Just waiting for parts to show up.

Will keep you guys posted.

Thanks again for your help.

I had this post a looooooooong time back and you all gave me some good inputs. Some of you were interested in the outcome and hence this post. Sorry for the delay as we kept running into one brick wall after the other.

Anyways, I tried a mixture of Alaric and Lancie1's approaches and they worked fine TILL I ran into another unforeseen problem - Dirt around the encoder. Basically the environment isnt very clean and the encoder wheel is riding on the conveyor and so is everything else in the environment. So I had to go to square one and ask the designer to put the sensor closer to the pumps and another sensor looking across the 6 cans - since we cant move the encoder. And encoder is pretty much everything here.

Crux of the matter is, after plenty of brackets I am just using handoff from one sensor to the other to detect my cans and dispense. It works well so far! (phew!)

Thanks a lot for everyone's input!!!