Jimmie_Ohio
Member
Well, I just got back from my meeting with my Allen-Bradley salesman and tech support guy. I have a new project to do that is interesting… Here is the problem and the restrictions plus I few things I already know. Please feel free to post replies based on any opinions you may have. And thanks in advance to all who do.
PROBLEM:
To control a conveyor system, which also consists of slide gates and other simple digital inputs and outputs as detailed below:
Seven (7) conveyors (less than 10HP motors)
Two (2) rotary feeders (less than 2HP motors)
Sixteen (16) dump valves (solenoid to open – solenoid to close)
Three (3) slide gates (solenoid to open – solenoid to close)
PAST SOLUTION:
Historically, I have controlled similar systems with the following hardware:
1) SLC500 Processor – typically a 5/03, 5/04 or 5/05 depending on the individual plant’s communication preference. Our systems are “stand alone”, but interface to a plant “DCS” or another PLC that I really have no involvement with. This interface is always done “by others”. At most, I map some signals to them.
2) SLC Local Chassis – usually 5 to 13 slots, depending on number of I/O
3) 1746-IA16 – 120VAC inputs for pushbuttons, switches, and all field input devices (speed switches, positional limits (prox’s), etc.)
4) 1746-OA16 – 120VAC outputs for pilot lights and solenoids (I fuse the solenoids on the terminal strip since there are usually 2 to 20.
5) 1746-OX8 – isolated output cards for motor starters. The starters are ALWAYS in a motor control center furnished by the plant. These have 120VAC control. I grab the 120 volt “hot” from the starter and feed it back to the starter coil whenever I wan the motor energized. The starter auxiliary contacts all go back to the PLC via inputs (item #3 above). I usually hardwire the local controls for forward, reverse, and jog for maintenance outside of the PLC signals.
NOTE: NO ANALOG INPUTS OR OUTPUTS
6) Operator Interface – sometimes we supply a PanelView of some type. Many times, the final customer’s process people build screens that include all plant processes, including ours, that display on the plant’s DCS or Main PLC status screens. They pick up my I/O for display.
Systems I have done in the past work great and are simple to design, implement, wire, test, and start-up. Never any problems.
Now for the “wrinkle”…
The customer wishes to go to DeviceNet. The salespeople have done a huge sell job on these guys and I need to figure out the cost of this versus my “traditional” way, which we have already budgeted.
Here are the facts or ideas I already have:
1) These conveyors are laid out in such a way that I will use two separate DeviceNet networks. This will allow me to stay within “distance” restrictions. Plus, the entire system is really two small systems that work together, so this is almost a “no-brainer”. I will use two DeviceNet scanners in my local rack.
2) The conveyors will now be VFD’s with “DeviceNet cards” in them. The customer is buying these.
MILLION DOLLAR QUESTION #1 THAT I NEED TO ASK THE CUSTOMER: What are the part numbers of the drives and DeviceNet cards in the drives?
A potential problem to this is that the conveyors need to be “jogged”, reversed, or run locally at the motor for maintenance. There is no way an electrician at the motor control center can use a keypad on the drive to do what the guy at the motor needs. He needs “line of sight”.
I have thought to drop a small 4 or 8 input block using ArmorPoint, ArmorBlock or ArmorBlock Maxum, whatever fits best to get these local inputs back to the PLC.
3) The rotary feeders are simple, single direction motors (maybe 2HP at most). I would control these locally, for maintenance, the same way as the drives described above.
4) The dump valves and slide gates have 2 inputs each (for position) and 2 outputs each for solenoid control.
I have thought to put an ArmorBlock Maxum block at each of these 21 places, each as a node. Remember, these will be spread out over both networks. No network will have 30 nodes used.
OTHER RANDOM THOUGHTS / FACTS:
1) The conveyors are totally enclosed. If work is done on them, the maintenance people follow strict lockout / tagout procedures. To get hurt, they would need to disassemble huge cover assemblies while the conveyor was running. Also, these conveyors can be two hundred feet long. Providing limit switches to indicate that a part of the conveyor was removed would be ridiculous. Therefore, “Emergency Stops” are not a concern, since exposure to danger is not imminent.
Please do not comment or address any replies to the issue regarding “safety” or you will get a friendly reply from me about how this is not the issue on this post.
2) The customer’s scope will be:
a) Supply of drives with DeviceNet interfaces (whatever they end up being)
b) Running thick trunk line around plant
c) Supplying all taps and thin drops to my “blocks” (whatever they end up being)
d) Supply of all DeviceNet power supplies
3) I will need to be sure the vendors of the dump valves and slide gates supply 24VDC valves. A-B has told me I could use 120VAC ArmorPoint modules, if I wish. I’m not sure about this… I plan on picking the vendor’s brain on if they have done equipment that interface to DeviceNet before. Ideally they can package these to plug in nicely, with not much work by me.
4) I could have my panel vendor make all local pushbutton stations bigger to house “chassis based” I/O to accommodate “hand-off-auto”, “local-remote”, “forward-reverse”, and “jog” functions at the motors in one box for each. No more than 8 inputs per motor. I’m leaning a little toward this.
5) I could have my panel vendor make a local enclosure near each dump valve and slide gate to house “chassis based” I/O to accommodate positional limits (prox’s) and solenoids in one box for each. No more than 2 inputs and 2 outputs for each. I’m leaning a little toward this since I probably need to wire these devices in to some kind of junction box to split signals to “on machine” blocks like ArmorBlock Maxum anyway.
6) The plant is dirty (power plant), but nothing exceeding NEMA 4 or even 4X (IP67) would be required. Certainly moisture is not an issue.
7) I already know that I need RS Networx for DeviceNet and the appropriate adapter. Since I will probably include these in our quote and give these to the end user after startup, I will find out which they prefer. This is not an issue.
8) Documentation – I will probably need to draw a “network layout” drawing for this. I was going to start with our existing “plan view” which is fully scaled and come up with symbology to show what type of DeviceNet item is where and refer to a detail on another drawing. What do others do to show “DeviceNet”? Posting a sample .dwg or .pdf would be greatly appreciated!
9) Allegiance to Allen-Bradley – This project will be all A-B. I have no latitude on this. Please don’t suggest the products of others, unless it is a wiring device specifically for DeviceNet that is unique compared to A-B and will save time or money on the installation.
10) I was going to provide a PanelView 1000 color terminal to provide operator interface. A-B has already suggested that I not make this a separate DeviceNet node. Instead consider a DH+ model that I can tie to the processor directly. Pros or cons?
IN CONCLUSION:
What would you guys do as far as “on machine” I/O for an application like this?
What should I beware of or otherwise be cautious of?
Again any feedback is greatly appreciated.
PROBLEM:
To control a conveyor system, which also consists of slide gates and other simple digital inputs and outputs as detailed below:
Seven (7) conveyors (less than 10HP motors)
Two (2) rotary feeders (less than 2HP motors)
Sixteen (16) dump valves (solenoid to open – solenoid to close)
Three (3) slide gates (solenoid to open – solenoid to close)
PAST SOLUTION:
Historically, I have controlled similar systems with the following hardware:
1) SLC500 Processor – typically a 5/03, 5/04 or 5/05 depending on the individual plant’s communication preference. Our systems are “stand alone”, but interface to a plant “DCS” or another PLC that I really have no involvement with. This interface is always done “by others”. At most, I map some signals to them.
2) SLC Local Chassis – usually 5 to 13 slots, depending on number of I/O
3) 1746-IA16 – 120VAC inputs for pushbuttons, switches, and all field input devices (speed switches, positional limits (prox’s), etc.)
4) 1746-OA16 – 120VAC outputs for pilot lights and solenoids (I fuse the solenoids on the terminal strip since there are usually 2 to 20.
5) 1746-OX8 – isolated output cards for motor starters. The starters are ALWAYS in a motor control center furnished by the plant. These have 120VAC control. I grab the 120 volt “hot” from the starter and feed it back to the starter coil whenever I wan the motor energized. The starter auxiliary contacts all go back to the PLC via inputs (item #3 above). I usually hardwire the local controls for forward, reverse, and jog for maintenance outside of the PLC signals.
NOTE: NO ANALOG INPUTS OR OUTPUTS
6) Operator Interface – sometimes we supply a PanelView of some type. Many times, the final customer’s process people build screens that include all plant processes, including ours, that display on the plant’s DCS or Main PLC status screens. They pick up my I/O for display.
Systems I have done in the past work great and are simple to design, implement, wire, test, and start-up. Never any problems.
Now for the “wrinkle”…
The customer wishes to go to DeviceNet. The salespeople have done a huge sell job on these guys and I need to figure out the cost of this versus my “traditional” way, which we have already budgeted.
Here are the facts or ideas I already have:
1) These conveyors are laid out in such a way that I will use two separate DeviceNet networks. This will allow me to stay within “distance” restrictions. Plus, the entire system is really two small systems that work together, so this is almost a “no-brainer”. I will use two DeviceNet scanners in my local rack.
2) The conveyors will now be VFD’s with “DeviceNet cards” in them. The customer is buying these.
MILLION DOLLAR QUESTION #1 THAT I NEED TO ASK THE CUSTOMER: What are the part numbers of the drives and DeviceNet cards in the drives?
A potential problem to this is that the conveyors need to be “jogged”, reversed, or run locally at the motor for maintenance. There is no way an electrician at the motor control center can use a keypad on the drive to do what the guy at the motor needs. He needs “line of sight”.
I have thought to drop a small 4 or 8 input block using ArmorPoint, ArmorBlock or ArmorBlock Maxum, whatever fits best to get these local inputs back to the PLC.
3) The rotary feeders are simple, single direction motors (maybe 2HP at most). I would control these locally, for maintenance, the same way as the drives described above.
4) The dump valves and slide gates have 2 inputs each (for position) and 2 outputs each for solenoid control.
I have thought to put an ArmorBlock Maxum block at each of these 21 places, each as a node. Remember, these will be spread out over both networks. No network will have 30 nodes used.
OTHER RANDOM THOUGHTS / FACTS:
1) The conveyors are totally enclosed. If work is done on them, the maintenance people follow strict lockout / tagout procedures. To get hurt, they would need to disassemble huge cover assemblies while the conveyor was running. Also, these conveyors can be two hundred feet long. Providing limit switches to indicate that a part of the conveyor was removed would be ridiculous. Therefore, “Emergency Stops” are not a concern, since exposure to danger is not imminent.
Please do not comment or address any replies to the issue regarding “safety” or you will get a friendly reply from me about how this is not the issue on this post.
2) The customer’s scope will be:
a) Supply of drives with DeviceNet interfaces (whatever they end up being)
b) Running thick trunk line around plant
c) Supplying all taps and thin drops to my “blocks” (whatever they end up being)
d) Supply of all DeviceNet power supplies
3) I will need to be sure the vendors of the dump valves and slide gates supply 24VDC valves. A-B has told me I could use 120VAC ArmorPoint modules, if I wish. I’m not sure about this… I plan on picking the vendor’s brain on if they have done equipment that interface to DeviceNet before. Ideally they can package these to plug in nicely, with not much work by me.
4) I could have my panel vendor make all local pushbutton stations bigger to house “chassis based” I/O to accommodate “hand-off-auto”, “local-remote”, “forward-reverse”, and “jog” functions at the motors in one box for each. No more than 8 inputs per motor. I’m leaning a little toward this.
5) I could have my panel vendor make a local enclosure near each dump valve and slide gate to house “chassis based” I/O to accommodate positional limits (prox’s) and solenoids in one box for each. No more than 2 inputs and 2 outputs for each. I’m leaning a little toward this since I probably need to wire these devices in to some kind of junction box to split signals to “on machine” blocks like ArmorBlock Maxum anyway.
6) The plant is dirty (power plant), but nothing exceeding NEMA 4 or even 4X (IP67) would be required. Certainly moisture is not an issue.
7) I already know that I need RS Networx for DeviceNet and the appropriate adapter. Since I will probably include these in our quote and give these to the end user after startup, I will find out which they prefer. This is not an issue.
8) Documentation – I will probably need to draw a “network layout” drawing for this. I was going to start with our existing “plan view” which is fully scaled and come up with symbology to show what type of DeviceNet item is where and refer to a detail on another drawing. What do others do to show “DeviceNet”? Posting a sample .dwg or .pdf would be greatly appreciated!
9) Allegiance to Allen-Bradley – This project will be all A-B. I have no latitude on this. Please don’t suggest the products of others, unless it is a wiring device specifically for DeviceNet that is unique compared to A-B and will save time or money on the installation.
10) I was going to provide a PanelView 1000 color terminal to provide operator interface. A-B has already suggested that I not make this a separate DeviceNet node. Instead consider a DH+ model that I can tie to the processor directly. Pros or cons?
IN CONCLUSION:
What would you guys do as far as “on machine” I/O for an application like this?
What should I beware of or otherwise be cautious of?
Again any feedback is greatly appreciated.