NI releasing/designing new controller for FRC

[Meshbeard]

I talked to the guys at the Cross the Road Electronics booth today. I can confirm that they are planning on having two versions of the Talon in the future: a pwm version and a CAN version.

I also took pictures of the informational posters they had out about the new control system (what they were allowed to say about it at least).
Control System Overview: http://i.imgur.com/eA3Bvfu.jpg
Power Distribution: http://i.imgur.com/bWDHpSt.jpg
Pneumatics: http://i.imgur.com/276hmyB.jpg
New things: http://i.imgur.com/YEtaHrp.jpg

I can try to elaborate more if someone asks for particular information.

[Peter Johnson]

Originally Posted by Meshbeard

I also took pictures of the informational posters they had out about the new control system (what they were allowed to say about it at least).
Control System Overview: http://i.imgur.com/eA3Bvfu.jpg
Power Distribution: http://i.imgur.com/bWDHpSt.jpg
Pneumatics: http://i.imgur.com/276hmyB.jpg
New things: http://i.imgur.com/YEtaHrp.jpg

Outstanding work by CTRE and a big thank you! This will make things like custom circuits/coprocessors significantly easier in the future (one of the ugly hurdles has always been the extra volume/weight of power conversion modules, the VRMs sound like the perfect solution). I love the additional CAN modules, particularly the pneumatics one and the addition of data logging. CAN has been a robust solution in the automotive industry for many years and it's good to see it gaining more traction in FRC.

I remain cautiously optimistic about NI's Athena.. while the new form factor is a really good idea (merging digital sidecar + crio = great!), and having integrated USB and CAN available is excellent news, what I really want to see is (a) if it's Linux based (rather than VxWorks) and (b) if boot times have been significantly improved. Linux would make life so much easier for development (out-of-the-box excellent USB driver support, robust TCP/IP stack, easier porting of tools/languages, a decent interactive shell prompt, non-kernel-mode code for easier debugging, better memory management, code reloads in any language without rebooting--just kill the user process and restart it, the list goes on), and we all complain about the current cRio boot time. However, I'm not holding out hope for either at this point given the "cRio platform" reporting so far.

[MrForbes]

I really need to get to the vendors booths today!

I like the idea of the size reduction of combining the processor and interface board. But the downside is that if you blow an interface, you have to replace/repair the whole thing. We've fried a couple sidecars, no damage to the cRio.....consider the relative cost of the two parts

[Alan Anderson]

Originally Posted by MrForbes

I like the idea of the size reduction of combining the processor and interface board. But the downside is that if you blow an interface, you have to replace/repair the whole thing. We've fried a couple sidecars, no damage to the cRio.....consider the relative cost of the two parts

One common way to ruin a Digital Sidecar is to put battery voltage on any of its "ground" pins and fry its reverse power input protection. Integrating it could make that specific reverse power protection unnecessary* and thus remove a failure mode.

* It could be combined with a protection circuit for the entire device that is less susceptible to permanent damage.

[Bennett548]

This new control system has the potential to make many robots more "robotic" rather than "RC cars with arms". Control systems are a very tricky concept, even for many in college, so I think that the move to make them more accessible to high schoolers is a great idea.

I had been pretty excited about AM's new shifter, but this definitely takes the cake.

[Steven Sigley]

So the pneumatics will be CAN, and the Power Distribution Panel, is there any way to integrate sensors like encoders into the CAN network without PWM cables in the future?

[Chadfrom308]

What if it is going to be arduino based
probably not, but you can code arduinos in labview

[mman1506]

It's not, you cannot do vision anything on a arduino

[FrankJ]

The NI Booth had a prototype. I almost got a picture of it. In addition to Ethernet, it has client/server USB ports & a high speed Canbus. It is running a dual core processor & a bigger FGPA. Although the guts are based on standard NI products, it is specifically designed for First. It is NI's hope that the programming tools for it will be backwards compatible to the CRIO. It looks to be novice friendly while having expansion opportunities for the teams with the resources to take advantage of them. No comment on rather or not the current CRIO's will be competition legal in 2015. Cost is predicted to be in line with the current CRIO.

A big thanks to National Instruments for continuing to support First in the way they do. I hope we get to Beta test it. (We are a java team)

PS
This is way beyond arduino.

[Meshbeard]

Originally Posted by Steven Sigley

So the pneumatics will be CAN, and the Power Distribution Panel, is there any way to integrate sensors like encoders into the CAN network without PWM cables in the future?

Most of the encoders used on FRC robots are quadrature encoders, which means they have four wires. They're usually four individual wires twisted together or a four wire ribbon cable, they're not really pwm wires. The way the digital sidecar is set up, you need to use three wire connectors, which is really inconvenient, but the new control system might have dedicated encoder inputs, which would be great.

[Tom Line]

Originally Posted by Meshbeard

I talked to the guys at the Cross the Road Electronics booth today. I can confirm that they are planning on having two versions of the Talon in the future: a pwm version and a CAN version.

I also took pictures of the informational posters they had out about the new control system (what they were allowed to say about it at least).
Control System Overview: http://i.imgur.com/eA3Bvfu.jpg
Power Distribution: http://i.imgur.com/bWDHpSt.jpg
Pneumatics: http://i.imgur.com/276hmyB.jpg
New things: http://i.imgur.com/YEtaHrp.jpg

I can try to elaborate more if someone asks for particular information.

I'll preface this by saying that I am not a CSE, EE, or embedded engineer. This layout scares me.

Having the most failure prone component (digital sidecar) now built into the robot controller worries me. How often have shorted pins, miswired power leads, and other mistakes caused burned-out sidecars?

I sincerely hope the controller is over-engineered to a level that makes it virtually indestructable. Otherwise teams will be replacing their entire controller when someone shorts a 24V power to a 5V jumper.

[crake]

Originally Posted by Tom Line

I sincerely hope the controller is over-engineered to a level that makes it virtually indestructable.

I would say that's not over-engineering - but engineering to spec

[Andy Baker]

This is a heads up from the peanut gallery:

Everyone interested in the new Athena controller system needs to pay attention to crake (aka: Chris Rake) who posted directly above this post. Notice that his team number is "Athena". I can confirm that his focus to make Athena great for the FRC teams is great and I am confident that this will be a wonderful system. Along with Chris, there are many other folks at National Instruments and other supporting companies who are working hard to make this a wonderful system.

I love what I see already, and I am excited to see the full roll out in August.

Sincerely,
Andy B.

[crake]

Originally Posted by Andy Baker

Along with Chris, there are many other folks at National Instruments and other supporting companies who are working hard to make this a wonderful system.

Thanks Andy - As Ray announced at opening ceremonies this system is a result of collaboration between numerous companies and organizations - all of whom are dedicated to making this the best possible system for this program.

I also have to ask for some forgiveness from the forum. There are a lot of questions, and some of these may have to go unanswered for the time being. But folks won't have to wait for long - August will be here very soon!

[Nate Laverdure]

6mm lugs for the main power terminals on the next-gen PDB. WHY???