Timeline for Daily chain sensors with HIGH/LOW signal output
Current License: CC BY-SA 4.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 1, 2022 at 4:23 | comment | added | jonk | @JeremyGillick If you are going to do this so-called 'bare metal' programming at each board, each of which are running with a local MCU that has its own clock source and therefore are essentially asynchronous from each other, then you may want to look at the APIC chipset bus used by Intel. It allows completely asynchronous clocking systems to compete for the serial bus, use it, then release it. There's a short burst of IDs on a wire-OR line for the purpose of determining which gets to own the bus next. It may be worth a little look-see. | |
| Nov 29, 2022 at 7:26 | comment | added | Jeremy Gillick | Cool. Thanks @jonk. The more I think about it, the more I think you're right. Cheers. | |
| Nov 29, 2022 at 3:23 | comment | added | jonk | @JeremyGillick I think you are better off just coding up a cheap MCU for each board. Dirt cheap, not a lot of code, and it will do exactly what you want it to do and, should something novel come up later, you can make a small change to incorporate that without changing the board -- only a software rev needed. And you aren't locked into some boutique chip, either. This is one of the kinds of thing I contract to do (and enjoy doing) for those less experienced in bare-metal coding -- making custom ICs out of MCUs. It sounds to me as though you already have those skills, as well. I'd go that way. | |
| Nov 29, 2022 at 1:57 | comment | added | Jeremy Gillick | But, to distill it down a little further, I'm looking for an IC that likely has two inputs and one output. Inputs: serial input and sensor signal input. The serial input would contain the sensor data from all the sensors before it. Output combines the serial input plus the sensor signal input. | |
| Nov 29, 2022 at 1:52 | comment | added | Jeremy Gillick | These sensors will be within a few inches of each other in a line. They will be hard-wired to each other for power and data, but ideally, each sensor wouldn't need an independent wire back to the main MCU. There's also a WS2812 LED on each sensor board, so the wires between boards will also include the required lines for the LED protocol. Ideally, the MCU would have wires connected to the first sensor board, and then that board would be wired to the next sensor -- so on and so forth -- with the last board connected back to the MCU. | |
| Nov 29, 2022 at 1:12 | comment | added | jonk | It's not clear to me, Jeremy, just exactly how these sensors are physically located and/or how many may be co-located (close enough for direct connection to 'something'.) But I take it from what you wrote and your comment that no two sensors are close enough to each other to share a mux. So that's why you write "bus" instead. I gather you know it is hard-wired either way. So you are not escaping wire. Just you won't want to use wire to an IC mux chip, I guess. You want some kind of 'thing' that will work over distance. Is that about it? Or am I missing something important? | |
| Nov 29, 2022 at 1:08 | comment | added | Jeremy Gillick | If all the sensors were on a single board, I could use something like a parallel-to-serial converter to take in the signal data from each sensor as a binary HIGH/LOW and output serial data. However, since each sensor is on an independent board, that method is tricky as each board would need a wire back to the main board with the parallel-to-serial converter. | |
| Nov 29, 2022 at 0:58 | comment | added | Christopher Moore | What do you mean by hardline to the multiplexer and what are the issues with it? | |
| Nov 29, 2022 at 0:46 | history | edited | Jeremy Gillick | CC BY-SA 4.0 |
edited title
|
| Nov 29, 2022 at 0:35 | history | asked | Jeremy Gillick | CC BY-SA 4.0 |