Timing problem

Hello everyone, I have been catching up on all of the post, lots of information here. I work in a poultry processing plant aka the chicken plant!! We are presently running all production lines on AB products, including our ice blowing system to deliver ice to the lines. These programs were straight foward and simple, just time consuming. To move on with our goal of standardising the plant, we have started reworking our Dapec machines. This is a piece of equipment that is aprox 10' wide by 55' long, runs in an oval with a shackle every 10 inches, basicly a person hangs chicken on one side, it cuts the chicken into up to 8 piece, then bags at 2 chickens per bag.
One of the modules on the machine is a churche's module for cutting, Churche's chicken. It basicly times itself to the shackles on the main line and allows the production personal to advance or retard its timing to adjust the cut on the chicken. The encoders are german built, as is the control board that operates it. The price for the control board is just over 5k, encoders 2200 each. With the addition of the sanitation department putting this panel under water every night for 8 hours, even with positive air pressure on the panel, we loose the main control board at a rate of every 3 to 4 months. The simple answer would be to move the control panel for the machine off the floor and into the main mcc room, but the cooporate people do not like this idea, imagine using an mcc room to hold control panels!!
Now to the meat of this post. I am having trouble timing the smaller conveyor on the churche's module to the main line. Through gear reduction both encoders will be rotating at 40 rpm. We purchased AB encoders that matched exactly what was on the machine, took a bit of doing but we found some, the original plan was to just replace the encoders with a product that was 1/4 the price and a lot easier to get. Then the plant manager asked why don't we get rid of the control board that operates them as well. So me in my naivitiy thought this is going to be straight foward and simple. I have built a scale system in our shop, and have managed to get the slave to follow the master in speed, but timing and beign able to adjust the timing through the logic rather than a physical means is what is giving me the trouble. I am not asking for someone to write the logic, just point me in the right direction.

The hardware we purchased for this venture includes:

(1) micro 1500 lsp, with awa base (the machines control circuit is 120vac.
(1) 1769-hsc
(1) 1769-if4xof2
(2) 845h-sjdz24cry2c encoders at 2500 ppm

The inverters are 3hp flex4 Ab, We use the same inverter on all applications 3hp and below, just set the inverter to the motor, it is simpler for the maintenance department that way.

I know without absolute encoders, with a power down it will loose its position, the present system takes about 6 inches of travel to resinc itself in time, so this limitation is not an issue.
The information I do not have in front of me, is how far the lines move in relation to a turn on the encoders, that information is at work, I am at home. I do know that the encoders are geared to move at exactly the same rpm.

Thank you for your time given in reading this post, and any help along the way.

I think I'll give KFC a miss tonight!

as a vegetarian I found i couldn't read through the post, kept thinking poor chicken...

Sir, I do not have a solution for your timing issue but you may extend the life of your boards with Clear Shot.

CLEAR SHOT™

CLEAR RUBBERIZED COATING by www.mantek.com



• Protects, seals and insulates against moisture, salt, road chemicals, dust and dirt

• High dielectric strength when cured*

• May be used on terminals and connectors

• Impervious to water — even if parts are submerged

Read the entire label before using this product.

Ha Ha thanks for the replies, when not cutting Churches it is cutting kfc. We presently use lectra shield, all chemicals we use in the plant have to be U.S.D.A. approved, and also approved by the vet in the plant, so anything worth using is not usable. Still working on it, presently trying to use the rate through one pid to control the speed, and alternately use the count, through a second pid, every quarter turn to control the output to sync the encoders in time. AB suggested buying a drive made to do this on it's own, so I suggested him coming and explaining that to my superiors, which he promtly declined. Will update when we make it work, or blow up the shop in frustration which ever comes first, thank you again everyone.

Your "shackle" line will probably be your master, so the only thing that you are worried about with it would be line rate.

The "churche's" line needs to be setup as a slave to the shackle line. To do this, start by matching the speed of the churche's line to the speed of the shackle line as close as you can.

Then take the encoder counts of the churche's line and subtract that from the encoder counts from shackle line. Apply this error to a PID to either speed up of slow down the churche's line (the PID correction would only be a small part of the speed control of the churche's line).

You can add an offset to the lines by adding an offset to your math. So far, this should be pretty obvious.

The "gotchya's" appear in how you setup the mechanics of your system.

The typical way of setting up the gearing (we would use an Edon unit for this) is to have the encoder rotate once per job spacing. This way the encoder automatically resets each time a new cycle starts. Is yours setup like this? When you get to the end of a job space what happens to your count? And how far apart are your shackles spaced? This is really important.

Another thing to consider is the the location of the Edon units (basically a gearbox that is attached to the conveyor that rotates its output shaft once per job space). They need to be located as close as possible to where you will be doing your work. This removes any error caused by chain stretch.

We've seen other problems occur because of line loading, but I suspect that your chickens aren't going to weigh anything near what our cars weigh.


These guys might be able to help:

http://www.mfg-sol.com/p/autotech_uticor/pos_trans_dec/line4_autotech-uticor_pos-trans-dec_m1890-programmable-position.htm


BTW, we used to do all of this using WWII military surplus selsyns and the innards to the meters in gasoline pumps (Veeder-Root).

No, seriously.

http://www.maritime.org/fleetsub/elect/chap10.htm

Thank you for the direction, problem solved. Long story short, due to time constraints we purchased a product from this company, http://www.contrexinc.com Local vendor, off the shelf item. We finished the project this afternoon, including the injection part of the line. So we didn't get the entire line down to one devise, but we came close. We were able to remove 3 seimens and one omron plc and run the entire line from one panel instead of 4, so we are getting closer to the goal of less controls. Still cannot move the control panels off the floor, but one day!! We did move the panel to the top of the line out of water and gave no controls on the door. This keeps the sanitation department from standing 20 feet from the machine and hitting the start button with high pressure water. Well they still do but not on the panel. We installed the hmi inside for the maintenance department to view trouble shooting screens. Have a great day everyone its miller time here.

I feel your water problem grief with Cleanup been fixing their damage for a decade now must love torture.