I know one can live without it and there are numerous clever workarounds but it has always bothered me that the I/O connections to the controllers did not use standard sorts of latching connectors (that use the same pins and sockets). The RoboRio looks to continue this design issue (in my humble opinion). So I was thinking about a quick design that would fit over the RoboRio. It would include, at a minimum, these features:
Latching versions of the 10 DIO, 10 PWM, 4 AIO and 4 RLY connectors
Latching connectors for the 16 MXP digital ports (PWM/DIO/I2C/SPI)
Latching connectors for the 4 MXP analog ports
Latching connector for the RSL port
TTL to RS232 level conversion circuit
2 DB-9 serial connectors
Clearance for LEDs, Ethernet, USB, Power, SPI and CAN ports
Clearance for reset and user buttons
Can attach to 2 mounting holes near the Ethernet/USB connectors
What else? Perhaps an Arduino circuit for custom LED string control, some diagnostic LEDs, a small display?
Comments, suggestions? I’ll probably open source the PCB design. But if we sold it, what might you pay for such a board (assembled or kitted out)?
All outputs from CUSTOM CIRCUITS shall connect to only the following:
other CUSTOM CIRCUITS,
input ports on the Digital Sidecar,
input ports on the Analog Breakout Board,
the RS-232 port on the cRIO,
the Ethernet network connected to either Port 1 or Port 2 of the cRIO,
the CAN-bus if and only if all Jaguar motor controllers on the CAN-bus are wired in full compliance with R67and R68, or
the sensor inputs on the Jaguar motor controller.
As an LRI, I would probably call your description (having not seen it and not having next years rules in front of me, I can’t make a definitive ruling) a custom circuit - even if the latching connectors are just pass-throughs, I would need a Q&A to allow it AND time with the bare circuit to verify it. It would be better to limit yourself to a circuit on the MPX with some plastic latching features that extend out to the other IO without actually interfacing with the IO pins, similar to the latches that extended up and over the lip of the PWM cables on the DSC and breakout boards.
Thanks for the input Jon - sounds like someone at FIRST would need some time with a circuit like this. Maybe I’ll try to get some time on it with a couple beta teams. It is passive (with respect to the discrete I/O) and the connector pins are the same (only the housing is latching) so maybe approval would be quick. The serial level converter, LEDs and Arduino-ish features meet the letter of the 2014-R71. The plastic latching features would be more complex to design and make fit properly - out of my league. I’ve never done any plastic work, I’m an EE - wouldn’t know where to start.
There was some serious discussion about this at the last test weekend. The rules are still under debate, even for completely passive boards. Also watchout, the DIO & PWM are shared on the MXP.
For the latching plastic, I’d recommend buying headers with it build in. The original proto boards used Molex KK series connectors.
In 2013 3574 used a wiring system based on direct interface boards on the crio, 10 pin ribbon cable, and 8 breakout boards. The ribbon cable carried 8 digital io signals as well as power and ground. The wire-to-board headers were polarized and latch locked.
It’s been over a year now, but it was explained to the judges and we were cleared to compete.
My project is very similar in concept (though I was not going to consolidate any signals). The connectors I plan to use are latching and polarized but use the same 0.025 square pins. Thanks for the pictures!
Will do, thanks for the heads up - I see the extra 16 DIO lines are multiplexed (in function) with PWM/I2C/SPI functions. The KK series are nice but only friction-locking (I think). I might go one step further and use latching polarized connectors.
Any of these, except possibly the LEDs, is more likely to get the breakout ruled as “active”. If you wanted to do something like these, the arduino, etc. should plug into the breakout rather than be part of it.
Another neat connector for this would be one designed so that a ribbon cable could bridge directly to a COTS 5V 4- or 8-port 5V relay module for LEDs or other light-duty custom circuits; many of them have very similar pinouts. (The module would need to get 12V power from the PDP, of course).
