C
I forgot to explain why I switched from 32 in 16 out to an even split. In my mind the outputs and inputs have to be on opposite sides of the board to prevent confusion. 36 terminals at .2" each (4 groups of 8 plus commons) plus space to fan out that wacky header connector layout with industrial spacing would mean a very long card with more empty space. ($$) As it is it's not as small and handy as I'd hoped. For this number of ports the PLC vendors typically cable out to terminal strips. (Usually with a special $80.00 connector) I thought about doing this with standard 25 pin IDC connectors but didn't see anything reasonable for the other end. Comments, ideas are welcome. Oh, one other thing. The reason I am so driven on this project is that I _need_ this for my day job. We will be installing a cell with Linux control. I had hoped to use LPLC for this but for now will be using my little demo mapping engine and compiled C logic to handle the digital IO. The engine will run in the background and the vision/lathe control/robot control portion will share the map. I hope to migrate to LPLC soon :^). Regards cww Curt Wuollet wrote: > > By way of a reply and an announcement: > > What started small has grown a bit: > > In a few more hours we will have full industrial strength IO for > Linux. Rather than compromise, I took a survey of PLC vendors IO > specs and produced a fully compatible design. > > Pro Version: > > Design features: > > 24 Fully optoisolated sinking/sourcing inputs. May be strapped in > groups of 8 for sinking or sourcing use with provision for > fast or slow filtering. 5000 V. isolation group to PC. The option > exists for adjusting logic thresholds. > > 24 Fully optoisolated NPN OC outputs with integral protection > diodes. Common strapped in groups of 8 with 5000 V isolation to > PC and XXX group to group depending on construction. Outputs > sink .5 ADC. Very generous duty cycle and # outputs on specs. > TBD but should be better than the big guys. External fusing. > > Provision for fixed or removable rising elevator contact terminal > strips. > > Industrial layout and design rules. All through hole devices. > Board should be sprayed with clear acrylic to maintain isolation > in dirty environments. > > Hobby/lab version: > > Cheap, for when the PC is the machine and isolation is not a > problem. > > 24 sinking voltage divider inputs with provision for filtering. > > 24 NPN OC sinking outputs with .5 ADC capability. > > May be used with or without terminals. Believe it or not, > decent terminals are the single biggest expense for these > boards.. The Hobby/lab artwork should appear in a week or two > after I fab and test the pro board. There are a lot of lines > and connections and there may even be an oops or two so I > want to verify the artwork/design before I derive the second > board, > > I will be offering the Pro version to the folks at my day job > for fab and testing to meet needs we currently have. This is > how I plan to test the Pro version. The Hobby/lab version will > probably need to be fabbed and tested by the user. I will > request that the proto house I use retain artwork/etc so boards > can be ordered with fast turnaround. > > Artwork for both versions will remain the property of the LPLC > project. They are produced in pcb, the free open source layout > tool and are freely editable. The input and output cells can be > used for other projects (if they work :^)) saving considerable > time and effort to support other boards. > > Regards > > cww > > Andrew Kohlsmith wrote: > > > > > Where you're coming from is not gelling with me but I've been > > > laying down tracks on the DIO48 thing for hours and nothing makes > > > much sense tonight. I'll look at it again in the morning. > > > > I'd been watching this for a while... the DIO48 is the IO board discussed > > last week? > > > > Regards, > > Andrew _______________________________________________ LinuxPLC mailing list [email protected] http://linuxplc.org/mailman/listinfo/linuxplc
