CAN Design Kit

[Team1710]

I was wondering if we would be able to use something like this (http://www.luminarymicro.com/products/rdk_bdc.html) if FIRST allows us to use CAN. What would be the benefit of using a CAN design kit or is this unnecessary?

[Jared Russell]

I believe this kit is essentially just a CAN-enabled Jaguar with a motor.

Note: I would hold off on purchasing something like this now; teams are currently beta testing the brand new CAN-enabled "Black Jaguar" that will improve several aspects of the Jaguar.

Here's a FIRST Forums thread (and image) about the new CAN-enabled Jaguar:

http://forums.usfirst.org/showthread.php?t=12983

http://www.freeimagehosting.net/uploads/e1cb7c9330.jpg

[DonRotolo]

The advice to hold off on buying anything is good.

You asked what benefits this kit can offer. The short answer is that if you need to develop something that uses the CAN Bus interface, this will help you do it more easily. Think of it as a small toolbox for CAN control of Jaguars. BUT, it is unlikely that FRC teams will be asked to develop the CAN interface; instead it is more likely that software tools (such as a VM for LabView) will be provided, and teams will merely have to parameterize (fill in the blanks) it to use CAN.

This would be an excellent question to ask Luminary Micro. Their sales support team answers questions like this every day, and if you identify yourself and ask a question, they'll provide an answer. (That's how it's done in the business world).

[dmcguire3006]

We got one of these just to familiarize ourselves with CAN. This particular device simply implements a CAN interface and allows the user to select (and set) the motor ID # and then control the Jag over the RJ11 connector via a simple LeftRightUpDownSelect menu and the OLED display. After learning all we could from this device, we converted into a handheld motor tester inside a RadioShack project box.

[jskene]

To interface with the CAN bus is it necessary to buy the NI 9852 CAN bus interface module?

This has a price of ~$1100, and so would not be legal on a USFIRST robot under previous rules, where the limit on any single component was $400.

[EricH]

To interface with the CAN bus is it necessary to buy the NI 9852 CAN bus interface module?

This has a price of ~$1100, and so would not be legal on a USFIRST robot.
I can think of a way it would be legal. Several, actually.
1) KOP item. KOP items and spares for them are not counted in cost accounting (until you get more than KOP quantity).
2) Discounted price (<$400) available to ALL FRC teams.
3) Change in the cost accounting rules to allow it.

[Joe Ross]

I can think of a way it would be legal. Several, actually.
1) KOP item. KOP items and spares for them are not counted in cost accounting (until you get more than KOP quantity).
2) Discounted price (<$400) available to ALL FRC teams.
3) Change in the cost accounting rules to allow it.

Option 4 would be not to use the NI CAN interface, but use another method. See http://forums.usfirst.org/showthread.php?t=13440

[jskene]

This RS-232 to CAN interface would be much more economical:

http://www.factorycomm.com/gc-can-rs232.html

[Ted Weisse]

Option 4 would be not to use it. See http://forums.usfirst.org/showthread.php?t=13440

Thanks Joe and 330 for your great work with the beta testing...

I have found that you should listen to beta testing teams and read the forums especially the FIRST beta testing one.
http://forums.usfirst.org/forumdisplay.php?f=1260

CAN will provide great possibilities but it appears that it won't happen this year.

We can still hope.
http://forums.usfirst.org/showthread.php?t=13313

Thanks all beta teams for all of your help!

[ajlapp]

What is going on with Tucan? The ethernet to CAN device I saw from cross the road electronics....

They demoed some impressive technology at IRI that would make talking to your CAN-enabled Jaguars a snap.

Also, I think all Jaguars are CAN enabled, even the kits items from last year. Can anyone verify this?

[Mike Copioli]

What is going on with Tucan? The ethernet to CAN device I saw from cross the road electronics....

They demoed some impressive technology at IRI that would make talking to your CAN-enabled Jaguars a snap.

Also, I think all Jaguars are CAN enabled, even the kits items from last year. Can anyone verify this?

Anthony, Jaguars have always been CAN enabled even last year. Teams were just not allowed to use the CAN interface.

Attached is a picture of the 2CAN. It will be released shortly. The price will also be announced shortly but I will say it has more features, more power and is cheaper than the Chicklet(129.99). The pricing is for FIRST teams only, as this device has many applications outside of FIRST. The web site has not been updated to show any information about the 2CAN. It will be updated soon and will have links to firmware updates, user manual and example drivers for the cRIO and possibly a link with live video stream that users may control motors over the internet with. The picture does not show dimensions. They are 3" x 2.5". this does not include the mounting tabs which make the overall 3.25" instead of 2.5". It is smaller than Jaguar but bigger than a Victor.

[AdamHeard]

Anthony, Jaguars have always been CAN enabled even last year. Teams were just not allowed to use the CAN interface.

Attached is a picture of the 2CAN. It will be released shortly. The price will also be announced shortly but I will say it has more features, more power and is cheaper than the Chicklet(129.99). The pricing is for FIRST teams only, as this device has many applications outside of FIRST. The web site has not been updated to show any information about the 2CAN. It will be updated soon and will have links to firmware updates, user manual and example drivers for the cRIO and possibly a link with live video stream that users may control motors over the internet with. The picture does not show dimensions. They are 3" x 2.5". this does not include the mounting tabs which make the overall 3.25" instead of 2.5". It is smaller than Jaguar but bigger than a Victor.

Will it be legal for the season?

[AustinSchuh]

Will it be legal for the season?

I'm guessing that we won't know for sure until the rules come out... I'm sure I've heard that phrase around here somewhere on the forums before by a certain individual :rolleyes: Anyways, this looks like a very nice device.

Even if it isn't legal, it opens up a huge number of options. All you need to drive a robot would be the wireless bridge and some custom software on a networked computer driving it. Depending on what data is accessible over the CAN bus, you could even implement some closed loop control and use sensors. A lot less expensive than a second CRIO, and it would work for probably 90% of the robots out there without any loss in functionality.

[ozrien]

All you need to drive a robot would be the wireless bridge and some custom software on a networked computer driving it. Depending on what data is accessible over the CAN bus, you could even implement some closed loop control and use sensors. A lot less expensive than a second CRIO, and it would work for probably 90% of the robots out there without any loss in functionality.

Actually one intended use for the 2CAN is to be the backbone for a modular based control system. Basically it opens up ethernet to the world of CAN and vice versa. So your point about any ethernet device being able to drive a robot is absolutely correct.

Furthermore the 2CAN was not intended to be a replacement for the CRIO. However, with some small changes to the enable/disable protocol, it could free up the FPGA for teams to use (I suspect the FPGA is used for this purpose and this is why teams do not have access to it).

[EricVanWyk]

Even if it isn't legal, it opens up a huge number of options. All you need to drive a robot would be the wireless bridge and some custom software on a networked computer driving it. Depending on what data is accessible over the CAN bus, you could even implement some closed loop control and use sensors. A lot less expensive than a second CRIO, and it would work for probably 90% of the robots out there without any loss in functionality.

Thats the part I'm really excited about. I'll bet we'll see a lot of practice bots that are an off the shelf wifi router/bridge/whatever and a 2CAN. Can't beat that!

[AdamHeard]

Thats the part I'm really excited about. I'll bet we'll see a lot of practice bots that are an off the shelf wifi router/bridge/whatever and a 2CAN. Can't beat that!

What will sensors be hooked up to?

[Andy Baker]

Thats the part I'm really excited about.

So are we at AndyMark. Here is the 2CAN page (http://store.andymark.biz/am-0545.html).

This device should be available before Christmas.

We are happy to work closely with Mike and Omar at CTRE (http://crosstheroadelectronics.com/).

Andy B.

[EricVanWyk]

What will sensors be hooked up to?

I was thinking more of box on wheels type practice bots to teach new drivers, so there wouldn't be any sensors. But on second thought, the Jags do have a couple sensor inputs. I'm not saying the 2CAN could replace a full system, but it really does lower minimum the cost bar for partial-bots.

Knowing Mike, I wouldn't be surprised at all if he made a CAN sensor node. Maybe the 2CAN itself will have a few inputs. I haven't gotten the chance to play with these guys yet, but it could happen.

Bear in mind that I tend to believe that more options is never bad, at worst neutral and usually good.

[AdamHeard]

I was thinking more of box on wheels type practice bots to teach new drivers, so there wouldn't be any sensors. But on second thought, the Jags do have a couple sensor inputs. I'm not saying the 2CAN could replace a full system, but it really does lower minimum the cost bar for partial-bots.

Knowing Mike, I wouldn't be surprised at all if he made a CAN sensor node. Maybe the 2CAN itself will have a few inputs. I haven't gotten the chance to play with these guys yet, but it could happen.

Bear in mind that I tend to believe that more options is never bad, at worst neutral and usually good.

Ah, I was thinking more along the lines of full functionality.

Could a gyro be hooked up the to analog input on a jaguar and have the operation of the two be unrelated? The Jaguar's analog input just being used to pass the value to the processor.

[Mike Copioli]

Ah, I was thinking more along the lines of full functionality.

Could a gyro be hooked up the to analog input on a jaguar and have the operation of the two be unrelated? The Jaguar's analog input just being used to pass the value to the processor.

The 2CAN is only one component of what would be a modular control system. It would be the core communications module. Other modules would be needed such as CAN to GPIO(general purpose input/output), CAN gyro, CAN to analog. CAN speed controllers such as the Jaguar, CAN accelerometers, CAN encoders. These devices could be developed by teams or purchased form a third party vendor. The advantage to using a device like the 2CAN is that a developer could use ANY processor using ANY platform to rapidly develop a fully functional robot. You could use a laptop, an arm processor, a PIC, a basic stamp. Any language c,c++, c#, basic, Java so long as the processor module supports either CAN or Ethernet or Both. Furthermore this device opens a door that will allow teams to develope their own CAN modules.

For years the primary focus of a FIRST team during design and the robot development cycle was mechanical and programming with some allowances for custom circuitry, leaving a very important field of engineering behind, Electrical. We mentor students on forces like inertia, and momentum, power to weight ratios with little or no focus on things like Ohm's Law, Thevenin's Theorem and Kirchhoff's laws. We do not have rules that require you to use a specific size axle shaft or sheet metal thickness because we want this to be part of the "Problem" that needs to be solved. But we have rules that require teams to use a certain gauge wire and specific breaker size. We disallow the use of power management devices in fear of fire. I argue a wheel flying off of the robot or an end of arm tooling that was under engineered is just as if not more dangerous than a rouge electronic component releasing the magic smoke in an arena with enough fire suppression capabilities to put out the earths core. Allowing teams to build their own speed controllers is as important as allowing them to build a transmission. Teaching them to use PCB layout and design software such as EAGLE or PADS or Altium is just as important as teaching them how to use Solidworks or Autodesk.

I hope the 2CAN is the gateway to these changes.

[kamocat]

Sounds great!
I have some ideas on a CAN relay module.
But... I don't really want to deal with the low-level CAN handling.
Is there any chance you could make a "CAN prototype board" sort of thing?
Most of the IO mapping can be done on the cRIO in the communication VIs.
But a device that would have several prescribed analog PID inputs and outputs, and some general analog and digital inputs and outputs.

I can see several possibilities with this already:
Wire the pressure switch for the air compressor directly back to the CAN relay module that controls it.
Servo controllers for a camera pan, tilt, and zoom, all in one CAN module.

[Tom Line]

The 2CAN is only one component of what would be a modular control system. It would be the core communications module. Other modules would be needed such as CAN to GPIO(general purpose input/output), CAN gyro, CAN to analog. CAN speed controllers such as the Jaguar, CAN accelerometers, CAN encoders. These devices could be developed by teams or purchased form a third party vendor. The advantage to using a device like the 2CAN is that a developer could use ANY processor using ANY platform to rapidly develop a fully functional robot. You could use a laptop, an arm processor, a PIC, a basic stamp. Any language c,c++, c#, basic, Java so long as the processor module supports either CAN or Ethernet or Both. Furthermore this device opens a door that will allow teams to develope their own CAN modules.

For years the primary focus of a FIRST team during design and the robot development cycle was mechanical and programming with some allowances for custom circuitry, leaving a very important field of engineering behind, Electrical. We mentor students on forces like inertia, and momentum, power to weight ratios with little or no focus on things like Ohm's Law, Thevenin's Theorem and Kirchhoff's laws. We do not have rules that require you to use a specific size axle shaft or sheet metal thickness because we want this to be part of the "Problem" that needs to be solved. But we have rules that require teams to use a certain gauge wire and specific breaker size. We disallow the use of power management devices in fear of fire. I argue a wheel flying off of the robot or an end of arm tooling that was under engineered is just as if not more dangerous than a rouge electronic component releasing the magic smoke in an arena with enough fire suppression capabilities to put out the earths core. Allowing teams to build their own speed controllers is as important as allowing them to build a transmission. Teaching them to use PCB layout and design software such as EAGLE or PADS or Altium is just as important as teaching them how to use Solidworks or Autodesk.

I hope the 2CAN is the gateway to these changes.


To a point, I agree. Picking up a hammer or a drill are life-skills, not just engineering skills, and a lot of the kids coming into FIRST have never even put a drill bit in a chuck due to their suburban - urban lifestyle and the throw-away mentality a lot of people have nowadays. I can definitely see an argument where the mechanical side of FIRST is more applicable for a larger subset of people.

The only way, however, to get MORE of those engineering skills out there is to do what you're doing. Which is pretty darn cool. Kudos Mike. That's a really cool bit of Kit that I think we'll try to use too :D

[Mike Copioli]

Thanks for the kudos Tom.

Picking up a hammer or a drill are life-skills, not just engineering skills, and a lot of the kids coming into FIRST have never even put a drill bit in a chuck due to their suburban - urban lifestyle and the throw-away mentality a lot of people have nowadays. I can definitely see an argument where the mechanical side of FIRST is more applicable for a larger subset of people.

The mechanical world is tangible.
The electrical world is abstract.
The programming world is lonely and misunderstood however sometimes the keyboard is mightier than the drill!:D

I prefer a combination of all of the above. As far as the life skills argument is concerned these are my thoughts; Many times when a friend or family member is in need of help with a project it is my soldering iron or oscilloscope that is required not my lathe or hammer. Hey can I borrow your truck to move has turned into hey can you fix my kids X-box. What does this tell you? It tells me that in a throw-away world these people are not willing to throw away these devices. But they lack the skills or know-how to make the repairs themselves. often times they even provide me with a link to a site that tells you EXACTLY how to fix the problem but because they are intimidated by the size and misconceived fragility of the components they seek help from others who CAN. Although taking a hammer to it might make you feel a lot better it will not fix your kids X-box. I believe that everything that you use through life regardless of the frequency is a life skill. When I was in high school i did not realize until my junior year that I had an aptitude for electronics. The primary reason......People were showing me how to use hammers instead of soldering irons. Then one day I saw a device that intrigued me. I had no idea how it worked. I found myself in a position that required me to solicit the help of others. The person I found was the high schools electronic teacher. The next year I enrolled in his class. The rest is history.

One argument that I will agree with is that the robot requires more "Manpower" to deal with the mechnical aspects then the electrical/programming.

[EricVanWyk]

One argument that I will agree with is that the robot requires more "Manpower" to deal with the mechnical aspects then the electrical/programming.

I have to agree with you now, but I hope I can disagree with you soon. I think the problem is that there is a large gap in the electronics for FIRST; there isn't enough in between the "connect a to b" and "full custom" categories. On the mechanical side, there is a much fuller continuum from "kit bot" to "CNC porn".

I really hope that the Cypress PSoC board for the driver station acts as a small step in bridging that sparky divide. It comes with a bread board, which I'm hoping will allow teams to begin to experiment with simple things like RC filters. It also has some neat bits like analog outputs and capacitive sensors, which may or may not result in some awesomeness.

I don't know how to fully bridge the gap, but I'm excited for the possibilities.