2018 Robogames Mech Target Plates/Transponder

[GhengisDhon]

We've made a few changes to the target plates for 2018. For the most part, these changes will be transparent to the end Mech user.

As in 2017, R-Team will be providing the target plates to each competitor just prior to the match start. In addition, we will also be providing the transponder as well. I repeat, for 2018 Robogames, competitors will NOT need to provide a transponder. The transponder is integrated into the back target plate. The mounting method, of the target plates to the Mech will be identical to 2017.

Info on how to mount the target plates can be found here:
http://forums.trossenrobotics.com/sh...-Target-Plates
https://github.com/ghengisdhon/rbots/blob/master/Target%20Plates%202017/2017%20Robogames%20Target%20Plates.pdf

The major difference (for Mech Builders) to the 2018 plates from the 2017 plates is that they now use 4-pin Dynamixel connectors. We highly recommend that competitors integrate the harness/connectors through their mech. Basically, you need a harness from each target plate location on your mech, to the back target plate location on your mech. (We will have limited loaners available.)

The 4-Pin Dynamixel connectors can be bought from Trossen here:
http://www.trossenrobotics.com/store...e-10-Pack.aspx
http://www.trossenrobotics.com/store...e-10-Pack.aspx
http://www.trossenrobotics.com/store...e-10-Pack.aspx
http://www.trossenrobotics.com/store...e-10-Pack.aspx
http://www.trossenrobotics.com/store...e-10-Pack.aspx

Hope this isn't to confusing... if you have questions, let me know and I'll try to clarify.

[jwatte]

Is there some kind of interactivity between the scoring system and the 'mech, such as telling the 'mech where it's hit and what the score is?
Or, barring that, is the network protocol going to be finally robust and easy to use? :-D

(Not that I'm building a new mech or anything. I don't have time for that. Just checking.)

[GhengisDhon]

Is there some kind of interactivity between the scoring system and the 'mech, such as telling the 'mech where it's hit and what the score is?
Or, barring that, is the network protocol going to be finally robust and easy to use? :-D

(Not that I'm building a new mech or anything. I don't have time for that. Just checking.)

Such privileged information only goes to those who are ACTUALLY building a mech! Honestly though, there used to be a pin on the transponder board that indicated the hit & location. We've respun the boards a few times and have been trying to minimize the board foot print, so I'll have to check with Artans and see if it has survived all the revisions.

[jwatte]

In the best of worlds, that pin wouldn't use the hundreds-of-milliseconds pulse protocol, but instead send a UART byte at 115,200 bps or something like that ... A value something like:

(0x80 if counted, 0x00 if not counted) | ((panelid & 3) << 5) | (remaining hitpoints)

And if you're out of serial capability, doing the same with a soft-serial at 9600 bps would still be alright :-)

[artans]

In the best of worlds, that pin wouldn't use the hundreds-of-milliseconds pulse protocol, but instead send a UART byte at 115,200 bps or something like that ... A value something like:

(0x80 if counted, 0x00 if not counted) | ((panelid & 3) << 5) | (remaining hitpoints)

And if you're out of serial capability, doing the same with a soft-serial at 9600 bps would still be alright :-)

To detect hits, it's still the legacy pulse. The pulse will be 50-200 ms based on which plate was hit. Unfortunately right now the pulse only happens when it scores a hit, so it doesn't tell you when you're getting hit and it doesn't count. The transponder works off 3.3V now using a 1s LiPo. The signal with the pulse is open collector now.

If we redo it, I'll consider implementing software serial.

[jwatte]

Thanks for the info!