Small Scale Automation Adventure in the British Isles!

[Zookes]

Hello, programmable logic controllerers! I must first apologise for my crippling ignorance on the subject of automation, I got the idea into my head to automate a process without really knowing a thing about it, as such I have been bouncing around trying to get guidance about what technologies to focus on. I started with CNC as I thought that was more or less a blanket term for automation... Not so much it turns out. My initial fumblings can be seen in these threads on other forums: http://www.linuxcnc.org/index.php/english/component/kunena/?func=view&catid=38&id=19484 http://www.cnczone.com/forums/general_electronics_discussion/153651-amazing_automation_aventure_uk.html Where I learnt that programmable logic control is probably a much more appropriate control system for my application. Here is as good a nutshell as I can muster right now: Basic Process: 1. Release measured amounts of Resin, Catalyst and pigment into mixing chamber. 2. Seal Mixing Chamber. 3. Decompress Mixing Chamber. 4. Send Mixer into mixing position. 5. Mix. 6. Return Mixing Chamber to atmospheric pressure. 7. Move Resin Solution to Separation Tray. 8. Push Resin Solution into Cavities. 9. Wait for Resin Solution to cure. 10. Open Cavities. 11. Pull Result out of Cavities. 12. Push Result off of Stake. 13. Repeat. Also, here are the elements the system requires to be controlled as far as I can see: 3x Valves/dispensers for primary liquids to mixing chamber 1x Valve for vacuum pump to mixing chamber 1x Motor to drive mixer (edited) 1x Vacuum pump for chamber 1x Pump system for cylinders 1x Valve/dispenser for mixing chamber to separation tray 2x Pneumatic cylinders for resin plungers 2x Pneumatic cylinders for push prongs 6x Pneumatic cylinders for plate pushers My thanks to those that have made it this far, sorry about the list-o-rama, but I have learnt that asking questions about a fresh automation project piecemeal doesn’t necessary give answerers the information they need to give detailed answers, so there you go. My questions are numerous, but I’ll start with the basics: Is a PLC system the most appropriate method for a system such as this? I mean that in terms of being: -Reasonably simple to set up and program (for someone who has no prior knowledge specifically, but a little programming experience). -Repeatably accurate. -Economical. -Given to working well with simple sensors and writing logs and the like. In comparison to other automation systems. If so, what would constitute a general estimate of the kind of controller I would be looking at for a job such as this, and please factor in that this is a very small scale operation and that I would prioritise value over features. Thanks in advance answerers! And sorry for the deluge of text! Edited May 12, 2012 by Zookes

[plcdp]

Hello, Hopefully I can answer a couple of your questions...yes this would be ideal for a PLC. I wouldn't say it is simple to set up and program, with no PLC programming experience you will spend a lot of time with programming manuals. Yes the PLC is highly repeatable and accurate. Processor power varies depending on your requirements, but even the slower PLC's nowadays are very powerful. As far as economical, price can vary quite a bit depending which PLC brand you choose. I currently have a machine that the processor for it cost over $3,000.00. On the flip side a have a small chemical mixing station using a click PLC from Automation Direct that cost under $100.00. All depends on your system requirements. To estimate what you will need you will have to come up with an initial I/O count. Figure out how many digital inputs and digital outputs you will need. Also you must decide if you need analog control. (Temperature, analog in, or analog out.) Once you have an I/O count and know your analog requirements you can start looking at PLC's that support the number of things you need. Hopefully this helps point you in the right direction. -Dave

[Zookes]

Right, an I/O count, will do. My instincts tell me that analogue control is unnecessary, as everything in the system is pretty binary as far as I can tell. As what you chaps like to call a 'ballpark estimate', what would a digital controller with say 20 I/Os set me back? Doesn't have to be fast, the process itself will take about half an hour from start to finish. It has indeed! Thank you very much. Edit: I notice this forum has a thumbing up/down system, does anyone bother with that? It is beneficial to members in some way? Edited May 12, 2012 by Zookes

[plcdp]

Based on your requirements, you could get a click PLC for 69.00 that has 8 inputs and 6 outputs. Then 1 expansion input module with 16 more inputs for 45.00, 1 expansion output module with 16 more outputs for 45.00, a power supply for 39.00, 42.00 for a programming cable, and the software is free. Totaling just over 200.00. That will give you a little more than 20 inputs and 20 outputs. -Dave

[Steve Bailey]

20 I/O is way too low a count for the syatem you've described. Your list of items to be controlled is a minimum of 19 outputs. Do either of those two motors that drive the mixer need to rotate in both directions? You need two outputs for each motor that can run in two directions. About those ten pneumatic cylinders. Depending on the style of solenoid valves you choose to control the air flow to the cylinders, you could need two outputs per cylinder. You haven't mentioned anything about inputs. Are you going to install switches to indicate the position of any of the components? How about pushbuttons to start and stop things? Are you going to use an HMI panel? My first impression is that the Click is a bit underpowered for this application. If you're going to use something from Automation Direct, I would start with a DL-06. Edited May 12, 2012 by Steve Bailey

[Zookes]

In what way? The process is slow so speed shouldn't be an issue, unless maintaining pressure with cylinders is more taxing than I realise. Thank you for your contribution Mr. Bailey. Edited May 12, 2012 by Zookes

[OkiePC]

I may disagree with Steve's assessment of the Click PLC. It is surprisingly powerful, has more memory than many other "brick" type PLCs and it is expandable. Free software is a big bonus, too. The big drawback to the Click is lack of analog. There are analog CPU units but they're limited to two inputs and two outputs of analog data. Hopefully, they will add analog modules to increase that capability. Along with an I/O count, you should list them as inputs and outputs separately, along with voltage and power requirements for each in order to select the right PLC/IO modules. Adding an HMI may be a very good idea. A good HMI will allow you to replace real push buttons and switches with a virtually unlimited number of touch buttons and also allow your machine to include adjustable values without the need for connection to an external PLC. A good HMI will also provide data logging capabilities which can be sent to a PC for further analysis/backup using a variety of methods. There are many different makers of quality HMI devices, also called OIT (operator interface terminal) and several other acronyms. HMI is probably the most commonly used term at present. Edited May 12, 2012 by OkiePC

[Steve Bailey]

That's what I had in mind when I said the Click is "underpowered" for this application. From the description, I can envision this project to need some analog signals before it's done, specifically, vacuum level, maybe temperature, maybe analog speed control for one or more of the motors. Since the OP isn't really sure about what his needs will be, I think starting with a platform that gives him more room to grow is a good idea. Incidentally, I'm not sure how Automation Direct prices compare with the competition in Great Britain, or exactly how Koyo products are marketed there.

[Zookes]

I'll admit, I'm highly reticent to add cost to the unit, but I suppose that if the workstation the PLC is connected to is off, then there will be no memory in the PLC to store logs and such? Therefore I'll need a HMI to maintain proper records and monitoring? Also, can a PLC be connected to a PC through a bog-standard router via Ethernet? As to parts, I am being advised here: http://www.cnczone.com/forums/general_electronics_discussion/153651-amazing_automation_aventure_uk.html that small air cylinders will more than suffice for my needs. That being so, does anyone know of a compressor/pump that may be used with an air cylinder system, I don't know what keywords to search for and so on, so can anyone start me off with an example? Edited May 13, 2012 by Zookes

[Zookes]

Update the first! So, I have been doing a little research and I feel that I have the pneumatic system more or less figured out. In short, I will use something like this: http://www.toolbox.co.uk/sealey-sac0610e-6l-direct-16397-116820 to compress the air, something like this: http://www.ebay.co.uk/itm/Pneumatic-Double-Acting-Compact-Air-Cylinder-SDA-12x25-/300700548056?pt=LH_DefaultDomain_3&hash=item4603263bd8 to actuate my plates and something like this: http://www.ebay.co.uk/itm/1-8-Bsp-Flow-Control-Valve-Uni-Directional-Alluminium-body-B300-/120880605020?pt=LH_DefaultDomain_3&hash=item1c250ba75c to control the flow of air. Does anyone have any qualms about such a set-up? What I am unsure of is how I will get the uniform compression I need. The actual cavities that need to be sealed are made out of silicone, so they can distort under too much pressure. My instinct is to have the surrounding plate utilise a hard stop which can be identified by the PLC via a limit switch of some kind. Is it possible to have a limit switch which can work together with a typical valve to maintain a very constant pressure, or would it be easier to have the cylinders simply compress to the hard stop and maintain pressure? Would there be any possibility of damage if a hard stop were utilised, if I made sure to adjust the pressure on the compressor to a reasonable level? I notice that the discussion sort of fizzled out a bit, have I committed a forum faux pas of some kind? Should I be posting new topics for new questions? Thanks in advance anyway. P.S. I will also be using the Siemens LOGO! series of PLCs as they seem to be ubiquitous over here and they have a nice for programming PLCs which looks perfect for a novice like me. Edited May 19, 2012 by Zookes