Block Measuring

escoplcguy

Member
Join Date
Jun 2010
Location
New Jersey
Posts
191
Hello everyone,
I have beem looking into block measuring and i believe i have found one that could work, but i cant seem to get in contact with my RA sales rep to answer my questions...

i have been looing at the 45MLA measurement light arrays and i wanted to see if anyone here has had any experence with this or something similar?

my questions are, if a block has a slanted top will the sensors see this and be able to measure this?? also our product is fradle and crushable, if a block goes through and has dammages areas will it sense this as well???



for more info......block is being measured for height between 5" to 40" max. some blocks are straigt across others have a slant meaning it starts at 11" and falls to 10" (see sketch) would like to know when taking measurment height of block on both corners to determan if block is a (example) 10" or 10/11".....also like i said some have the corners broken off or holes in the sides, these blocks need to be rejected.

slanted top.jpg
 
I have not used the Allen Bradley measuring curtains you referred to, but I have used the Valu-Scan light curtains from STI and they were outstanding.

You can get them with analog output so they're easy to interface to nearly any PLC. They also provide a serial data stream and can do things like count holes in a web, measure the total distance blocked, highest point blocked, lowest point and a few other choices IIRC.

If your "warped" blocks travel through the beams with the sloped edge in the right direction, your PLC should be able to see that slope, detect broken corners and such, but it's going to depend on how fast you scan the input, how fast the blocks are moving, etc.

The VS6500 has a doublescan mode which can effectively double the resolution when used with a thin material like in web handling, but I think with your thick blocks, you'd need to turn off that feature, and be limited to the standard resolution (beam spacing) which I think was as low as 10mm, but we used 0.5" beam spacing for our web hole/loop height detection app.

It has been 10 years since I used the ValuScan, and since that time Omron has bought STI, so apparently the product has good longevity, but I have no idea what sort of improvements have been made over the last decade.

EDIT: After checking into the RA product, and configuring one for 1200mm/47.2" height (cat# 45MLA-AT1200P10 ), and analog controller, and adjustable brackets, the retail price was over $2000...I betcha STI will come in a wee bit less expensive too.
 
Last edited:
OkiePC the blocks are not moving very fast as they will be pushed through the light curtain by an operator (for now) on a gravity conveyor so the info from the sensor to the plc wouldnt be much a problem, and great to hear about the slant being detected, only thing is the dammaged blocks now, clipped corners could be possible im thinking if i set a tolerance, but what about the lets just call them "dents" in the side walls, you wont beable to see through them, when the curtains do you have a control for tolerance over each and every beam, or is it a totalized analog input.
 
escoplcguy said:
...only thing is the dammaged blocks now, clipped corners could be possible im thinking if i set a tolerance, but what about the lets just call them "dents" in the side walls, you wont beable to see through them, when the curtains do you have a control for tolerance over each and every beam, or is it a totalized analog input.

If you want to see chipped corners and dents in 2 or more adjacent sides, you may want to go with a full blown vision system or laser scanner...on the other hand, if there's an operator pushing the blocks through...surely that person can be expected to use his/her eyes?

Ten years ago, the STI with analog option board could be configured to send an output when a minimum height has been reached (so you can ignore the bottom edge if you need) and that output could correspong to total beams blocked, lowest beam blocked, or highest beam blocked. You would use the last option.

It had a serial connection so you could run a PC with hyperterminal (or equivalent) and perform some of the configuration as well as see a stream of data representing the state of each beam. You could also use a PLC to send simple commands to that port and receive digital information to do with as you please, but that might be more trouble than it's worth.

So as a block was pushed through at 50% of full scale (assuming 0-10vdc analog) your signal should quickly jump up to 5 vdc, and quickly drop off after the block passed.

If the slant is less than 1 beam spacing (10mm) you won't see it at all, but if it was greater than that, you should see a shift in the analog signal when that slant crossed the next beam. In practice, I doubt you would be able to tell the difference between a block that was perfectly flat, but on the threshold of one of the beams or a block with a 10mm slope. If the slope was 2 or three beams wide (1-1/2" or greater) then you could probably reliable detect that.
 
Last edited:
thanks for the info, as far as the vision system, how does this work?? is it like a camera that can detect defects?? how can i get more detailed info (if i just type in vision system to google i get 1000000 responces) and how does it report to a plc??
 
If you want to consider vision, get ahold of reps for Cognex, Banner, Keyence...(I am forgetting others...)

Have them come to your plant and give you a demo. They will be glad to look at the application and make proposals and save you the time spent researching a very application specific and vast field of study...
 
Keyence has exceptional resources on it's site. While most will complain about having to register to use, it does offer more resources than most of it's competitors.
http://www.visionsystem.com/topics/cv/resource_center.php
View attachment cv_academy8_ka.pdfSample of 1.
I believe Keyence also does demos.

Pressco was the system used at a facility I once worked at.
http://www.pressco.com/index.php It was rather impressive considering it ran on a 486dx with NT 4.0. 32000 bottles per hour and 3 hi-res pics per bottle. 1 little booger on a lens and you had water bottles and hot preforms everywhere if you weren't fast enough to shut it down.
thanks for the info, as far as the vision system, how does this work?? is it like a camera that can detect defects??
In simple terms.
The software looks at the pixels. The pixel colors or greyscale are compared by sectors or neighbors. The face of the concrete could show up as lighter pixels and the voids could be shadowed thus darker pixels. A sector with more than x dark pixels would be a fail. The angle of light would be one way to determine how deep of a void you get before you get a shadow. The angle or parellels could be detected by a profile you setup.
 
ok guys, i just got word on our specs and tolerance. its basically insane but hopefully SOMEONE out there has an answer....so measurment is important but not as important as the tolerance. which is +/- 1/8 of an inch. so with the 10mm beam spacing i dont think a light array will do, unless you know of a way....so basically forget the divits cracks and detecting breaks, they right now just want to know if it is in spec. meaning, if it is a 20 inch block and we run it down the line, it needs to measure above 19 7/8 and below 20 1/8. if it met this requirement a green light would go on otherwise a red light would go on and reject block. also please note that our blocks range in size between 5" - 36".......my mind is fried right now please some one help me.
 
Can you count on the bottom of the block remaining at a fixed position?

If so, use a laser distance sensor mounted above the block with analog output. You should be able to get the resolution you need from that...

Use the operators for now for defect rejection, then bring in a vision guru for help with automating that if necessary.
 
ok new tolerance when i walked in this morning....possible solution....

Banner EZ-ARRAY with a beam spacing of 5mm would be emitter reciever pair

tolerance specs are now +/- 1/4 of an inch. 1/4in = 6.35mm with a beam spacing of 5mm i should be good right??? just for tolerance detection at least. if i wanted block measuring i would need to put a second light array from top to bottom? or as okiepc suggested laser distance sensor?
 
...
specs are now +/- 1/4 of an inch. 1/4in = 6.35mm with a beam spacing of 5mm i should be good right???

In an ideal world you will be cutting it close.

I would opt for the laser distance sensor...
It should be less money and simpler to set up...
It should also far exceed your (already changing) spec, and far exceeding specs is almost always a good thing...

gotta run...i'll look for an example link later.
 
1/4in = 6.35mm with a beam spacing of 5mm i should be good right???

Best case is if one beam is almost precisely on the correct height. Then an under would uncover a lower beam and over would block an upper beam. The actual tolerence would be 5mm up and down.

Worst case if the ideal point is split between two beams then you will either have an up and down tolerence of 2.5 mm or 7.5 mm depending on what you choose.
 

Similar Topics

Please see attached file. I need this program in Function Block form but I am totally lost on this. Any help would be appreciated. Thanks!
Replies
8
Views
260
I am having a step7 v5.4 program where the blocks are encrypted and locked. And the manufacturer is stopped the support. Is there any ways to...
Replies
2
Views
132
We have a new machine that uses an armor block to transfer input signals for conveyor position that continuously alarms out on no position change...
Replies
0
Views
74
Hello all, I have a 1732ES-IB16 Series A block that I have set the IP address on. I am unable to disable bootp using any method. I have tried...
Replies
9
Views
267
Hello, I am stuck on trying to get an ArmorBlock, 1732E-16CFGM12QCR to respond to EIP outputs. My PLC is Panasonic FP0H-C32ET. I Registered the...
Replies
1
Views
106
Back
Top Bottom