Facebook
Google
GitHub
Linkedin
Beginner with a project, unfortunately a more ambitious one than I meant, but that's okay, because I have access to the internet and no deadlines in mastering the sufficient amount of knowledge.
I have a matrix encoder board that scans a 16x8 matrix for switch actuation. That is, 24 pins. 128 switchpoints, like a full computer keyboard would be.
I’m trying to design a modular system where you can chain multiple individual switch modules (or keyboards) to the encoder board, or remove them from the chain as you like. And yes indeed: since the actions programmed to the switchpoints are what they are, some modules would contain switches that perform the same action. On the other hand, whenever a given action isn’t relevant to a given module, the wiring would simply run through it and on that module, there’s no switch in that place that you could actuate; it’s ignored, but the encoder doesn't know that there isn't even a physical switch that you could actuate if you wanted to.
Some modules will be permanently connected and available at all times, but some will be hot-pluggable. Anyhow, the encoder would always “think” that there’s only one keyboard, being unaware of the fact that the switches might equally well be located across an arbitrary amount of chained modules.
After a lot of research, I understood that the principle that I need to implement is called a data bus; The main serial cable of 24 wires are permanently connected to the 24 pins on the matrix encoder board and at the end of the cable, I could have the module that does not need to be unplugged but stays connected at all times. But along that cable, I’d have to have receptacles that maintain a connection to the main bus but allow an additional branch to be hot-plugged in as required.
A classic rainbow ribbon cable with IDC (insulation-displacement contact) connectors looks like a simple and cheap option; You can make it a short ribbon that can be neatly tucked away inside whatever enclosure I will make for the project, and leave the branching receptacles ready and available on the outside wall. But there’s two problems with that:
One is, those IDC receptacles don’t look like something that you’d leave visible. Do some kind of neat bezels or plates exist, that makes these individual receptacles safe and secure to mount to stick through the wall of the project box, while allowing them to be spaced as snugly close as they can? I find these face plated versions with screws for some types of connectors, but not for these...
Another question is, what kind of cable I could use for the module that connects to that female receptacle on the bus that branches out. I don’t know of any other option than again, using a 24p ribbon cable with those very bulky pin blocks. These modules are things that are supposed to be taken from a drawer when required, put away when no longer necessary, and they are used on an office desk. I can’t see a 24p ribbon cable being a good choice for this, as they are rigid, massive, ugly, and there can be several of these modules in use at any given time. That’s a lot of ribbon cable in plain sight. I think I need a more sophisticated solution but I don’t know what, I’m torn between the choices because I don’t know if these fancier standards actually even support the type of communication that this project would utilise. (A matrix encoder scans for all the rows and columns one by one, in quick succession; Physical key press causes a row and a column to short which lets the encoder know where the key press happened. Locating it on the coordinates is similar to how the blocks on a chess board, or a spreadsheet, are named.)
Something called IEEE-488 or GPIB is currently one that I’m especially interested in; As a specialty, there’s a stackable version available: the connector at the end of the thick cable has male on one side, female on the other. This would make building of a bus a breeze and the connectors would feel robust enough to be handled regularly. It has exactly 24 pins but the pin names confuse me: Pin names aside, are cables like this anyway merely the same kind of “dummies” as a ribbon cable is but the pin labeling would merely refer to what each of those pins traditionally connects to *inside the devices* that officially adopted the connection type as their standard? Or is there something going on under the hood that I don’t know of, such as some conductors being in contact with each other inside? I have such a heavy Wikipedia hangover by now, I also may be ignoring completely viable candidates just because they were commercially used for something very specific only. And the newer the technology is, something as simple as cables seem to get more complicated, using less wires to achieve more, and some cables have logic of their own which seems like something that would hinder the simple connection that I'm trying to do. "Simple" – not in configuration perhaps, but as logic, because all of the logic is in the encoder board and the rest is just paths.
Thank you for your time!
I have a matrix encoder board that scans a 16x8 matrix for switch actuation. That is, 24 pins. 128 switchpoints, like a full computer keyboard would be.
I’m trying to design a modular system where you can chain multiple individual switch modules (or keyboards) to the encoder board, or remove them from the chain as you like. And yes indeed: since the actions programmed to the switchpoints are what they are, some modules would contain switches that perform the same action. On the other hand, whenever a given action isn’t relevant to a given module, the wiring would simply run through it and on that module, there’s no switch in that place that you could actuate; it’s ignored, but the encoder doesn't know that there isn't even a physical switch that you could actuate if you wanted to.
Some modules will be permanently connected and available at all times, but some will be hot-pluggable. Anyhow, the encoder would always “think” that there’s only one keyboard, being unaware of the fact that the switches might equally well be located across an arbitrary amount of chained modules.
After a lot of research, I understood that the principle that I need to implement is called a data bus; The main serial cable of 24 wires are permanently connected to the 24 pins on the matrix encoder board and at the end of the cable, I could have the module that does not need to be unplugged but stays connected at all times. But along that cable, I’d have to have receptacles that maintain a connection to the main bus but allow an additional branch to be hot-plugged in as required.
A classic rainbow ribbon cable with IDC (insulation-displacement contact) connectors looks like a simple and cheap option; You can make it a short ribbon that can be neatly tucked away inside whatever enclosure I will make for the project, and leave the branching receptacles ready and available on the outside wall. But there’s two problems with that:
One is, those IDC receptacles don’t look like something that you’d leave visible. Do some kind of neat bezels or plates exist, that makes these individual receptacles safe and secure to mount to stick through the wall of the project box, while allowing them to be spaced as snugly close as they can? I find these face plated versions with screws for some types of connectors, but not for these...
Another question is, what kind of cable I could use for the module that connects to that female receptacle on the bus that branches out. I don’t know of any other option than again, using a 24p ribbon cable with those very bulky pin blocks. These modules are things that are supposed to be taken from a drawer when required, put away when no longer necessary, and they are used on an office desk. I can’t see a 24p ribbon cable being a good choice for this, as they are rigid, massive, ugly, and there can be several of these modules in use at any given time. That’s a lot of ribbon cable in plain sight. I think I need a more sophisticated solution but I don’t know what, I’m torn between the choices because I don’t know if these fancier standards actually even support the type of communication that this project would utilise. (A matrix encoder scans for all the rows and columns one by one, in quick succession; Physical key press causes a row and a column to short which lets the encoder know where the key press happened. Locating it on the coordinates is similar to how the blocks on a chess board, or a spreadsheet, are named.)
Something called IEEE-488 or GPIB is currently one that I’m especially interested in; As a specialty, there’s a stackable version available: the connector at the end of the thick cable has male on one side, female on the other. This would make building of a bus a breeze and the connectors would feel robust enough to be handled regularly. It has exactly 24 pins but the pin names confuse me: Pin names aside, are cables like this anyway merely the same kind of “dummies” as a ribbon cable is but the pin labeling would merely refer to what each of those pins traditionally connects to *inside the devices* that officially adopted the connection type as their standard? Or is there something going on under the hood that I don’t know of, such as some conductors being in contact with each other inside? I have such a heavy Wikipedia hangover by now, I also may be ignoring completely viable candidates just because they were commercially used for something very specific only. And the newer the technology is, something as simple as cables seem to get more complicated, using less wires to achieve more, and some cables have logic of their own which seems like something that would hinder the simple connection that I'm trying to do. "Simple" – not in configuration perhaps, but as logic, because all of the logic is in the encoder board and the rest is just paths.
Thank you for your time!
