Staple management

marzan

marzan

Senior Member
Hello everyone. I am planning out a new project. I want to make a circuit that will check for 2 staples on a book as it is traveling along a chain. We have issues on a couple of older stitching (stapling) machines at work when people working on the back don`t see when a staple is missing. Newer machines have something to check for them. I think the newer machines use a hall effect sensor to check for staples, does that sound feasible? reading up on the sensors, you can put a magnet behind the sensor which makes the staple change the magnetic field so the sensor can see it. Is that correct? would that be with most sensors do you think? Or would I have to buy a biased hall effect sensor?
The plan is to use some sort of eye to detect the book and then to count pulses till the book passes and then to reset for the next book. Of course a Picaxe would control it all.

Marz
 
L

laserhawk64

Senior Member
Depending on how sturdy the book is, and how it moves through the machinery, I'd actually pull a switch out of a computer mouse for that. They're like lever-switches, but without the lever -- the actuator is a teeny little bump with just enough travel so that it's perfect for this. If you can keep the books moving at a decently slow pace and in line such that they don't move sideways more than (literally, unfortunately) about 1mm either way (left or right) then you're all set. A simple IR photo-interrupter would work for the book-detect. Depending on the thickness of the book I'd use a standard interrupter module (available on fleaBay for quite cheap) for thin books or magazines. For anything thicker, I'd probably rig something up with a CD drive IR diode (use an old old CD-ROM drive diode, not something from a burner -- otherwise you'll fry the detector) and a photodiode detector.I'd probably wire the book detect input to the PICAXE's in-only pin, out of habit. That would trigger to start watching for staples, and if it doesn't see two staples by the time the book-detect input indicates the book's moved on -- stop the presses!EDIT: if I *had* to use a lever switch, or if the books weren't aligned well enough to get away with the mouse switch... solder a piece of quite-stiff wire to the end of the lever, so that it makes a 'T' shape -- in effect, the lever is wider now. Position the lever switch vertically, so that a book coming down the line bumps the switch in. Wire up your pull-ups (or pull-downs if that's your thing) and Bob's your uncle :)
 
L

laserhawk64

Senior Member
Why not? I imagine they're moved around by something with at least a *little* torque... I can't see why it wouldn't work.
 
marzan

marzan

Senior Member
@texasclodhopper you are right. mechanical means wont work. The machine at top speed can do 3 book a second. As well as that the staples don`t always form correctly. That`s why something to check for metal passing by is used in newer machines
 
R

rossko57

Senior Member
Selecting the sensitivity of the sensor might be fun, I wouldn't know how to go about that without trial and error. Most industrial sensors are self-contained, no external magnet needed. The idea of an optical break-beam to detect a book passing and trigger the testing process sounds right. Whether you want one or more staple-sensors depends on the orientation. Are you able to mount them in the 'bed', could be tricky arranging overhead fixing. I doubt you'll be able to tune to detect badly formed staples.

What will actually happen when fault is detected, at three books a second?
 
marzan

marzan

Senior Member
I was planning on triggering a MOSFET so they can put a red ligh, buzzer or both if it detects a missing staple
 
techElder

techElder

Well-known member
Are you saying that you would STOP a "3 book a second" production line to remove a book without the correct amount of staples? I hope not.

At those speeds, you have to "bump" a suspect book out of the regular line or somehow mark it for later recovery. Marking the errant books could be as difficult as detecting the staples mechanically.

Generally speaking, what you want to do is "detect" the irregular book, start counting books and then when that book is a certain count beyond the detection station open a "gate" to let only that book move out of the line. Remember that there could be 100 errant books in a row (due to a bad stapler situation), so you have to make rules about acceptance standards and when to shut down the line, etc.

You could easily outrun the capabilities of the PICAXE.
 
Buzby

Buzby

Senior Member
Texasclodhopper said:
.... You could easily outrun the capabilities of the PICAXE.
Click to expand...
I don't think so.

3 books per second is 330ms per book, that's plenty time to do the calculations.

The 1024 byte scratchpad in an X2 could map over a thousand books, nearly 6 minutes worth.

The real challenge is what to do with the detection.

Easiest is to mark with a paint spray, or a felt tip on a little servo arm.

Physically ejecting the bad book is more of a mechanical challenge than a software problem.
 
hippy

hippy

Technical Support
Staff member
I suspect there will be some software challenges. Even though there's a lot of time as each book passes I imagine there will be a far shorter time between books.

That problem isn't insurmountable though as it seems only necessary to detect staples and detect the book has passed through; at that instant you would have a count of how many staples were in the last book, can reset the staple count ready for the next, and then have the entire time up to when the next book has passed to handle the one before it.

Staple counting can probably be handled by interrupt or in the background using on-chip hardware, but if handling a book with missing staple did tie up processing each book adversely it could pass a signal to another PICAXE which can handle the marking, ejection, routing or whatever of that book and quickly enough that it had not moved much further up the line.

A PICAXE does seem a suitable solution. I would agree that what happens after a missing staple detection is the potential challenge. Slamming the brakes on the line would work if feasible and acceptable but there could well be better solutions. Hard to say more without knowing details of the production line and process.
 
marzan

marzan

Senior Member
Woah! not gonna be that complicated. I Hope to just automate a process that is checked manually by packers at the delivery end of the machine. Let me explain. A stitcher is called that because it forms a staple out of coiled up straight wire in big rolls like MIG wire. It is formed milliseconds before it is punched through the book. In older machines the stitcher heads tend to play up either jamming the wire feed or many other issues which stop the staple being made. After being "stitched", they are trimmed in a 3 knife trimmer which is 10 feet from the stitcher. then they go into a stacker which is preset to count bundles of completed books and eject them in their piles. Consequently there can be 100 books produced before the packer sees what is coming out the end, so even if they see them instantly and stop the machine there could be over 100 wasted books. All I need to do is check the staples are there, and if 3 in a row are not. Then a siren and/or light goes off and someone on the machine hits the E-Stop. Most wire issues mean you have to stop the machine to fix the problem, whereas thickness issues are not always issues that would make you stop. As Texasclodhopper assumed there is a diverter/Ejector. That handles book thickness issues, but without going into too much detail it is a timed mechanical setup. Hope that makes sense. It`s been a long day .
Marz.
 
premelec

premelec

Senior Member
I'd go for a mini-paintball gun that marks the bad copies or a kicker to knock them off the line... :) If your inductive sensor(s) gets a few wrong you can pick the ones that are ok off the kicked pile - just set the staple sense unit a little on the nervous side...
 
Buzby

Buzby

Senior Member
Hi Marz,

I am 100% certain we can do it with a PICAXE, but before we get there we need to detect a staple.

The first task is to find a device than do this, so concentrate on this aspect first.

Without a reliable staple detector no amount of processing power will solve your problem.

Cheers,

Buzby
 
F

flyingnunrt

Senior Member
Are the staples magnetic? Some stainless steels are not.
You may have to make your own miniature metal detector RF loop thingy to signal when a staple has passed by.
 
Pongo

Pongo

Senior Member
flyingnunrt said:
You may have to make your own miniature metal detector RF loop thingy to signal when a staple has passed by.
Click to expand...
I agree. A staple is a very small amount of steel.

Is there some point in the book's trajectory where the staples are not close to other metal? How close is other metal to the staples at that point?
 
marzan

marzan

Senior Member
I am thinking that a sensor that works from a longer distance might detect a thin staple at half the distance. I have a spot picked out where it could be mounted against the frame. I need to mount it close to the book but in a way that it wont get damaged if there is a jam up near the sensor. Buzby that`s what I thought. get a sensor, mount it, and see if it detects before doing anything else. Pongo the mounting point I have picked out should be OK. Also the chain has plastic fingers on it that raises the book 20mm above the metal chain, so that should give me plenty of space underneath. I will take pictures when I go back to work on Monday.

Marz.
 
Buzby

Buzby

Senior Member
Hi Marz,

This sensor is the most critical part of the system, so it needs to be right.

Faffing about trying to 'roll your own' is never going to have the same results as a COTS device.

I strongly suggest you contact a sensor manufacturer and discuss this with them.

The only company I've dealt with is IFM, so I suggest you send an email to http://www.ifm.com/ifmgb/web/mailstart.htm

They will probably put you in touch with IFM Australia, who will probably send a rep round with a bag full of different sensors, and a demo rig !.

Don't waste time reinventing the wheel.

Cheers,

Buzby
 
marzan

marzan

Senior Member
Buzby said:
Hi Marz,

This sensor is the most critical part of the system, so it needs to be right.

Faffing about trying to 'roll your own' is never going to have the same results as a COTS device.

I strongly suggest you contact a sensor manufacturer and discuss this with them.

The only company I've dealt with is IFM, so I suggest you send an email to http://www.ifm.com/ifmgb/web/mailstart.htm

They will probably put you in touch with IFM Australia, who will probably send a rep round with a bag full of different sensors, and a demo rig !.

Don't waste time reinventing the wheel.

Cheers,

Buzby
Click to expand...
Sounds like a good idea Buzby. Also thought I could check out what is on our newer machine and see if there is a part number
 
marzan

marzan

Senior Member
I remembered that I had some old sensors in a tub somewhere so I dug them out. They are Allen Bradley 872c-d3np8-d4 sensors. I looked up the data sheet and they are 3mm sense detect. Hooked them up and they see a staple at 2mm so that`s not too bad.I figure a 10mm sensor should see at 6 mm clearance which should give enough leeway when running if I drop it down to 5.. Thinking the best plan is to hook the sensor up to a picaxe and run 50 books and get the picaxe to count the pulses to see how well it works.
Marz.
 
D

Dippy

Moderator
Sounds like a good plan !
Obv I assume you've bench tested first to check 'compatibility' with PICAXE input?
 
marzan

marzan

Senior Member
Dippy said:
Sounds like a good plan !
Obv I assume you've bench tested first to check 'compatibility' with PICAXE input?
Click to expand...
No. not yet. I thought get the sensor that I want to use first. But I could try the other one while I am waiting. I will be using 12v to power the sensor. Can I use a simple voltage divider to the input
pin or is there a better way to drop the voltage?
Marz.
 
D

Dippy

Moderator
I don't know what the mystery sensor puts out.... logic 12V or 5V or an analogue or what... stick it on the kitchen table , apply the correct power, stick a multimeter or 'scope on output and wave something in front of it. Help us and we can help you ;)
 
marzan

marzan

Senior Member
Dippy said:
I don't know what the mystery sensor puts out.... logic 12V or 5V or an analogue or what... stick it on the kitchen table , apply the correct power, stick a multimeter or 'scope on output and wave something in front of it. Help us and we can help you ;)
Click to expand...
he sensor

I put 12v (range of sensor is 10-30v) to the sensor. Output was 10V when activated.
 
techElder

techElder

Well-known member
Is the sensor ACTIVE LOW or ACTIVE HIGH? In other words, is there an open collector output or does it have an internal driver? Data sheet will help there.
 
You must log in or register to reply here.
Top