Need guidance on spec'ing a PLC

Volts006

Member
Join Date
Jun 2008
Location
Georgia
Posts
3
I am an electrical engineer working for an A-E which designs all sorts of facilities. I’ve had all kinds of experience in lighting systems, fire alarm systems, power distribution, and telecom systems, etc.. My controls experience is some what limited especially in dealing with PLC’s.



I have a project where I have to compare the height of two hangar doors and stop a controls operation if one has moved farther than the other. I have proposed to do this with a rotational encoder on the motor shafts calibrated to measure the heights and encode it into a voltage or current. The existing controls are mechanical relay type. Therefore I have to send the encoded information into a PLC. The PLC must be programmed to compare the two elevations and then send out a stop signal to the existing mechanical relay controls.



Since I don’t deal with PLC’s much, how would something like this typically be spec’ed out.



My assumptions are 1) design a ladder diagram showing all existing relay logic and where the output of the PLC will be input into the existing relay logic to accomplish the desired operations, 2) Also, write a sequence of operations indicating how the entire system needs to be operate together, 3) Get input from a PLC manufacturer or PLC rep. as to a recommendation of a PLC, and 4) Use this recommendation as the guide to write bid specifications.



This seems to be all that needs to be designed on put on drawings as I perceive it.



As far as the PLC issues such as a) Check loss of state on power outage, b) check for the need of a signal delay, etc.; I assume that the specifications above would require that the supplier/assembler/control panel manufacturer would design/check this based on the functions that the system have to operate.



Man that was a mouthful. Any tips, ideas, guidance?
 
Volts006 said:
I have a project where I have to compare the height of two hangar doors and stop a controls operation if one has moved farther than the other. I have proposed to do this with a rotational encoder on the motor shafts calibrated to measure the heights and encode it into a voltage or current. The existing controls are mechanical relay type. Therefore I have to send the encoded information into a PLC. The PLC must be programmed to compare the two elevations and then send out a stop signal to the existing mechanical relay controls.

Questions :

1. How much error is allowable?
2. Does any thing need to happen if one of the drives inbetween the motor and a hangar door fails?
3. Is there a better place to mount the encoder, like the shaft that actually raises the door? And is there a failure possibility, like a cable snapping that can make the encoder reading false?
4. How many revolutions from bottom to top?
5. Speed?
6. Distance?
7. Voltage availible?





 
If feasable I would use a string pot instead of encoders. Check these out.

http://www.celesco.com/promo/spseries.htm

Two primary reasons for this. #1 is that they are absolute, you don't need a homing or reference point. #2 You can use an analog signal like a 0-10V or 4-20mA signal, which is much easier and cheaper to get into a PLC.

From what you've described darn near any PLC with 2 analog inputs would do. Do you have any other PLC's where you work and what kind are they. Personally I would do it with an AB micrologix.
 
Replys in red below.

milldrone said:
Questions :

1. How much error is allowable? if more than 14" unaliged, end all movement, and bypass around comparison and reset either door which is out of line.
2. Does any thing need to happen if one of the drives inbetween the motor and a hangar door fails? yes, neither door should move, i'll take of this with a normally open relay interlocked to the opposite door
3. Is there a better place to mount the encoder, like the shaft that actually raises the door? Good question-still working on that, but I'm assuming on the shafts. And is there a failure possibility, like a cable snapping that can make the encoder reading false? Good question-not sure need to evaluate further.
4. How many revolutions from bottom to top? Don't know hoping to put this on the contractor with a performance spec.
5. Speed? Don't know hoping to put this on the contractor with a performance spec. and have the contractor calibrate upon installation.
6. Distance? approximatlly 35-40'
7. Voltage availible? any type available
 
I've looked into the stringpot. But there is really no way to mount it, without the possibility of getting damaged.

I would prefer the stringpot, but can't mount it safely in this application.
 
Volts006 said:
I would prefer the stringpot, but can't mount it safely in this application.

Allscot beat me to the string pot question, but it doesn't mater anyway.

Here is an example of multi turn absolute encoder.

http://www.beiied.com/PDFs2/HMT25_Absolute_Encoder.pdf

This model has a maximum of 4096 turns before it wraps arround. If your shaft turns more than this when opening the door then this won't work.
 
Ther are plenty of absolute encoders that run on Device Net. I am personally using Peperl + Fuchs and Irwin Sick on cranes at the moment. They are programmable and very accurate but getting pricey for a simple job.

I would use an Omron CP1L with pulse encoders and a 'Z' phase for error checking. Would be quite inexpensive also. Have a setup on gantries like this being commissioned at the moment. Is working very well.

Would also look at mounting the encoder directly on the end of the shaft will a flexible coupling. Have tried gears and had problems - have had no problems with a direct drive though.
 
Volts,

You really should write your specs around a certain brand and model of PLC. You need to go ahead and choose that yourself, based on what is already exisitng and in use at this facility. It is VERY difficult for the facility engineers and maintenance people to learn to troubleshoot and maintain a bunch of DIFFERENT brands of PLCs. They can do it of course, but why put them through this bother? The cost difference in PLCs will be small compared to the ongoing job of maintaining the PLC and the program.

Each time you introduce a different brand and model of PLC to the facility, someone will have to buy software for that PLC and learn to get into the program, in case there is a processor fault or other problems, or if a change needs to be made. It is generally better to limit the number of types of PLCs to allow for more efficient operation and maintenance.

If your described hangar doors have a 14" tolerance, then you probably do not really need to use encoders. Two proximity switches mounted near each shaft, and looking at metal targets attached (welded or bolted) to the shaft will serve as a counter to count the revolutions of each shaft. This should be accurate enough to get each within 14". You will just have to write your program with two internal Counters, one for each shaft. Figure out how many revolutions (R) of the shaft equals 14", then add a rung with a Comparison statement: If Absolute Value of (Counter 1-Counter 2) >= R, then energize STOP output. If you go this route, then use the slowest shaft, probably not the motor shaft, but the shaft that opens the door, which will be much slower. If you use the motor shaft, you will need to use a PLC with a built-in High Speed Counter and Inputs for the high speed counter.

As Allscott said, nearly any PLC could do the above. Using the proximity switches, you would not even need any analog inputs, just 2 digitial inputs and a digitial output for the Stop signal. Any timers can also be internal (inside the PLC program).

This job sounds simple enough, that it would make a good project to be solved right here online. With a little more information, I would be honored to write the program for you and post it here, as we have done many times in the past. You may find that there are many other functions that could be done by the PLC, so make sure to include some spare Inputs and Outputs for future modifications.

The Allen-Bradley Micrologix series would certainly be a good choice, IF you already use those in this facility.
 
Last edited:
...if more than 14" unaligned, end all movement, and bypass around comparison, and reset either door which is out of line.
Let's review your specs one more time: That is 1 FOOT and 2 INCHES tolerance, not 1.4" or 0.14", correct?

Many who are suggesting encoders and motion controllers must have read it as 1.4".

If your current system is working using mechanical relays, it is a safe bet that a couple of proximity switches will be accurate enough, and much simpler and cheaper all around. Careful selection of the switches will insure that they switch on/off fast enough for the application, and have the proper target distance so that mounting is possible.
 
Last edited:
Lancie1 said:
Let's review your specs one more time: That is 1 FOOT and 2 INCHES tolerance, not 1.4" or 0.14", correct?

That was my thought when I read that... You could almost just randomly stop and be within 14". All you need to maintain that level of control is a prox switch. You sure don't need anything fancy.
 
Volts,

Trying to read between the lines...

I’m getting the picture that these doors are wide vertical panels (2-3 panels maybe?) related to each other so when they are retracted, (open), the panels would be horizontally "stacked"??

AND...What you really want to do is to synchronize the drives on either side of the door opening so the panels don’t get ****ed and bind??

AND... The door panels will be counterweighted??

AND... The door opening is like 100ft wide??

If the above is correct:

1. Why aren’t you spec-ing a bi-fold door? A bi-fold would require much less overhead clearance (wasted space that would need to be heated etc).

2. The tried and true method of moving vertical sliding doors would be to use 1 motor control to operate both motors. You keep each end of the door in sync via a driveshaft running from one side to the other. For positioning, if you are not going to use limit switches, then you should spec a multi-turn absolute encoder ($$$). This is so that in the event of a power cycle, when the controller wakes up, it knows where the doors are.

These doors are not very fast, (about 1ft per second). Not really an application for a motion controller. You would only need a PLC if you are using an encoder for positioning. If using limits, a few relays will do.

The reason why these doors cant be synchronized with an encoder on each end, is that the final drive would typically be a traction unit. All traction drives slip. However the counter-weighting and relating mechanism would be sprocket driven. Mechanically coupling the counterweight sprockets would give you positive error free synchronization.

Regards, Mike

 
1. Why aren’t you spec-ing a bi-fold door?
The existing controls are mechanical relay type.
Mike, it is an existing system. The Owner is probably not interested in replacing the hangar doors. I assume that the relays are not doing the job of keeping the doors in alignment.


It sounds like an airplane hangar, with two large doors on separate motors, but they need to open and close more or less at the same time. There probably is a clutch between motor and door shaft, but door shaft should not slip, at least not enough to prevent detecting a 14" difference.


More information about these doors would help a lot about knowing what could be done to keep them aligned.
 
Last edited:
I missed the "existing door" part.

I'm at a loss as to why to separate doors would need to be kept in sync. So based on our past experiences with rather large doors, I made the presumption that there's a miss understanding and the problem really is the wide panels are not traveling level.

IF this is the case, I've seen many a door operator installed by the lowest bidder, where the two ends are not mechanically coupled and over time the drives begin to slip and the doors begin to bind. If this is the case, Volts is going to expierance at best mixed results with suggestions posted so far.

Volts admitted to limited experience in this area. I'd hate answer his question directly and send him on his merry way, when I suspect he should be traveling east rather then south.

I would ask that Volts give us a more complete description of the doors, and tell us why he needs the synchronization.

IF I"M WRONG:

Then I would suggest checking into this encoder:
http://www.beiied.com/PDFs2/HMT25_Absolute_Encoder.pdf

and for the PLC I would suggest an AutomationDirect DL-06 PLC with 2 D0-DCM comm modules to talk to the encoders.

Im not sure if the BEI SSI output would be compatable with the D0-DCM, (a quick call to AD tech support would answer this), but if not, there are other multi-turn absolute encoders with Modbus RTU outputs that would be.

 

Similar Topics

So, my job is pretty much giving me a PLC lab to train with. I just finished up some basic analog signaling and processing using a process meter...
Replies
14
Views
8,966
Hey guys, I would like to ask some advice. I am working on the LogixPro bottle line simulator and I am having trouble figuring out the logic in a...
Replies
0
Views
1,261
I am working on an application in which I will need to compare two PNP (or NPN) pulses, each of which coming from light sensors at two separate...
Replies
9
Views
2,575
Hello everybody. I am into technical marketing of various automation products except PLC's. Recently I have been given the responsibility of after...
Replies
3
Views
3,581
CheezyMane13
C
I am drawing a blank on this one. I have two PCs running intouch 9.5 sp 1 with 1747-pktx cards (winXPprosp1) 17 slc5/04s on DH+ 56kbps I...
Replies
18
Views
7,037
Back
Top Bottom