[FRC Blog] 2015 Control System and Kit of Parts Survey

[magnets]

The roborio is only $435. That's really nice, but is this a once per year deal like the cRIO is?

Also, the chart has some errors. They list the number of cRIO Digital I/O incorrectly.

The power distribution board at $200 is pretty ridiculous. The current monitoring features are great and all, but I don't think most teams will be taking advantage of them.

My biggest dislike is the lack of relay ports. We usually don't use pneumatics besides for shifting, so we never have a compressor on board. If we use one spike for shifting drive (so we don't have to buy a $90 board that is way more complicated than it needs to be), then we only have three spikes left for motors. We usually use at least that many.

This has the potential to be a really awesome control system, but these people really need to be more open with us. If they showed us what WPIlib and the driver station looked like now, then there wouldn't be such a shock in December, when we buy the systems, and we don't know what we're doing.

[ILAMtitan]

Originally Posted by StAxis

I certainly hope that the red cypress board being in the picture means that the driver station software will be readily compatible with it next year.

You might have missed it since it sold out so quick, but there was a board called the TI Stellaris Launchpad available in FIRST Choice for teams. We offered this board as a replacement for some of the more complex game pad emulation systems that teams typically use and have problems with. All you have to do is wire in buttons, flash the firmware to the board, and plug it into your driver's station. The whole systems enumerates as two USB game pads making it super easy to use in LabView, or whatever language you prefer.

David Yanoshak (dyanoshak around here) wrote up a post late last year with a lot more details: What is the TI Stellaris LaunchPad in FIRST Choice?

David also links to where you can get the board via DigiKey or the TI e-store if you missed out on FIRST Choice.

TI is working on building a lot more Launchpad based examples and applications that FRC teams will be able to utilize in the future as well. If you want though, you can check it out the ecosystem on your own: ti.com/launchpad

We'll keep you guys posted when we have more available, and please let me know if there is any kind of custom circuit or embedded application that we can help you build.

[Andrew Schreiber]

Originally Posted by magnets

The roborio is only $435. That's really nice, but is this a once per year deal like the cRIO is?

Also, the chart has some errors. They list the number of cRIO Digital I/O incorrectly.

The power distribution board at $200 is pretty ridiculous. The current monitoring features are great and all, but I don't think most teams will be taking advantage of them.

My biggest dislike is the lack of relay ports. We usually don't use pneumatics besides for shifting, so we never have a compressor on board. If we use one spike for shifting drive (so we don't have to buy a $90 board that is way more complicated than it needs to be), then we only have three spikes left for motors. We usually use at least that many.

This has the potential to be a really awesome control system, but these people really need to be more open with us. If they showed us what WPIlib and the driver station looked like now, then there wouldn't be such a shock in December, when we buy the systems, and we don't know what we're doing.

I'd much rather have those features for the additional $11.

[R.C.]

Originally Posted by Andrew Schreiber

I'd much rather have those features for the additional $11.

And the slightly smaller size!

[Michael Hill]

Originally Posted by magnets

The power distribution board at $200 is pretty ridiculous. The current monitoring features are great and all, but I don't think most teams will be taking advantage of them.

I think you, to quote Dubbya, "misunderestimate" most other teams. I saw the new current sensing ability and immediately saw the benefit. Remember how a lot of teams are having issues tripping the main breaker with 6 CIMs and a compressor? That can now be a problem of the past.

[Pratik Kunapuli]

Originally Posted by Michael Hill

I think you, to quote Dubbya, "misunderestimate" most other teams. I saw the new current sensing ability and immediately saw the benefit. Remember how a lot of teams are having issues tripping the main breaker with 6 CIMs and a compressor? That can now be a problem of the past.

Although I think that the compressor will be run in a closed loop without any way of code affecting it, you are certainly right. We had 6 cims on our drive train this year and we often browned out the cRIO with about 20 seconds left in the match. Having the ability to see the current being pulled would let us stop that from ever happening again.

[Michael Corsetto]

Originally Posted by Michael Hill

I think you, to quote Dubbya, "misunderestimate" most other teams. I saw the new current sensing ability and immediately saw the benefit. Remember how a lot of teams are having issues tripping the main breaker with 6 CIMs and a compressor? That can now be a problem of the past.

Although this is partly true, current draw isn't the whole story to tripping the 120amp main breaker. The main breaker, which is a thermal breaker, will heat up and become more susceptible to tripping as it gets hotter.

Thus, tripping the main breaker is a function of both current draw and temperature. We have set up our practice bot with a thermocouple on the main breaker and inductive current sensors on the ground pole of the PDB. We have been logging current and temperature values in an attempt to isolate tripping conditions, but even then it can be difficult to really narrow down exactly what is happening the moment the main breaker trips.

I only have a limited knowledge on this subject, I'm hoping the students on our electrical team will post a whitepaper on the subject soon.

Regardless, excited to see current sensing integrated into the PDB.

-Mike

[Tom Line]

Originally Posted by ayeckley

I guess I'm not wild about the fact that what we used to be able to accomplish in three modules (cRIO, PDB, Digital Sidecar) now takes four (roboRio, PDP, PCM and VRM). I get that it's less expensive overall and that each of the components is physically smaller, but it's not a slam-dunk improvement in terms of the architecture. I guess CAN (and all of the benefits and drawbacks thereof) is now mandatory if a team wants to use pneumatics.

I'm also a bit surprised that the Bridge/Access Point is USB-based. When I saw the pre-Alpha unit at NI Week last year it wasn't obvious that this was the route they would go.

I guess it is what it is...

Currently, the 'access point' is a USB Dongle with drivers on the rRIO. It's a huge improvement in size and weight, and is easy to place anywhere on the robot now.

[Michael Hill]

Originally Posted by Michael Corsetto

Although this is partly true, current draw isn't the whole story to tripping the 120amp main breaker. The main breaker, which is a thermal breaker, will heat up and become more susceptible to tripping as it gets hotter.

Thus, tripping the main breaker is a function of both current draw and temperature. We have set up our practice bot with a thermocouple on the main breaker and inductive current sensors on the ground pole of the PDB. We have been logging current and temperature values in an attempt to isolate tripping conditions, but even then it can be difficult to really narrow down exactly what is happening the moment the main breaker trips.

I only have a limited knowledge on this subject, I'm hoping the students on our electrical team will post a whitepaper on the subject soon.

Regardless, excited to see current sensing integrated into the PDB.

-Mike

I understand that it's a near integration of current draw, but this now very easily lets programmers sense when they get into a pushing match. From that point, they can now decide to shift automatically or to cut off a CIM on each side. When in a pushing match, 6 CIMS easily draw about 450 Amps. The main breaker can only sustain that for a fairly short period of time before tripping (couple of seconds). However, if you drop down to 4 CIMs while in a pushing match, they'll stall out at about 350 Amps. The main breaker can sustain that for a much longer period of time.

[Joe Ross]

The submission date in the blog post has been updated to May 30, which matches the flyer that was handed out at the championship.

[Dale]

The roborio is only $435. That's really nice, but is this a once per year deal like the cRIO is?

At championships they said that the Roborio will be sold through AndyMark with no limits on the number ordered. I suppose they might decide to limit initial purchases to just make sure they don't run out before everyone gets one or two but that's not the plan going forward.

[Tom Line]

Originally Posted by magnets

The roborio is only $435. That's really nice, but is this a once per year deal like the cRIO is?

Also, the chart has some errors. They list the number of cRIO Digital I/O incorrectly.

The power distribution board at $200 is pretty ridiculous. The current monitoring features are great and all, but I don't think most teams will be taking advantage of them.

My biggest dislike is the lack of relay ports. We usually don't use pneumatics besides for shifting, so we never have a compressor on board. If we use one spike for shifting drive (so we don't have to buy a $90 board that is way more complicated than it needs to be), then we only have three spikes left for motors. We usually use at least that many.

This has the potential to be a really awesome control system, but these people really need to be more open with us. If they showed us what WPIlib and the driver station looked like now, then there wouldn't be such a shock in December, when we buy the systems, and we don't know what we're doing.

There have been Alpha test teams working on this since September of last year, and per the last post from FIRST they plan on starting a wide-ranging Beta test this year 2 months early and getting stuff to folks in July.

The Alpha test teams working on the system (and let's exclude mine for a moment) include some absolutely top level programmers from World Champion teams - folks who have had working relationships with NI and FIRST, and many of whom have done Beta testing every year it's been offered.

Hardware details are quite literally, released as soon as they have them. In many cases they are still ironing things out and making changes. In fact, we were given Rev C roboRios at Champs that we are now shipping back from some improvements.

Alpha Teams are traveling back to New Hampshire with our robots in a little less than a month after just having finished champs. New software updates were just release a short time ago. We'll be spending every weekend between now and then testing all the new changes. It would be pointless to release a WPILIB to the community that isn't even completed yet and still has bugs.

I don't mean to be snarky, but there are a whole lot of people working with this stuff since last September non-stop, including setting up demos, bringing it to competitions, and holding a seminar with NI folks who took their personal time to fly up to Michigan and put it on, etc.

FIRST, NI, and CTRE and the other Alpha testers are doing a phenomenal job testing a product and getting information out to testers as fast as possible. They won't stick up for themselves out of professionalism, but I've got no problem saying flat out that they're doing a great job.

For anyone who wants to get involved and see the stuff 'before December', by all means sign up for the Beta testing and be part of the solution. The more experienced teams who test, the better. Keep in mind though that part of being a Beta team means holding a seminar for other teams, so none of the information comes as a shock to anyone.

[Jefferson]

Originally Posted by magnets

The roborio is only $435. That's really nice, but is this a once per year deal like the cRIO is?

I don't work for NI, but my understanding is: no. You will be able to buy as many as you want.

Originally Posted by magnets

The power distribution board at $200 is pretty ridiculous. The current monitoring features are great and all, but I don't think most teams will be taking advantage of them.

It's $11 more and it has data logging as well as current sensing.

Originally Posted by magnets

My biggest dislike is the lack of relay ports. We usually don't use pneumatics besides for shifting, so we never have a compressor on board. If we use one spike for shifting drive (so we don't have to buy a $90 board that is way more complicated than it needs to be), then we only have three spikes left for motors. We usually use at least that many.

I'm not sure if you'll be able to do this. Teams currently have to charge the pneumatics using the control system, even if the compressor isn't on the robot. I'm not sure if the current compressor object in WPILib will survive the upgrade, although you could always program your own with a relay and digital input.
I'm not sure how it's "way more complicated than it needs to be". It has power, pressure switch, and CAN bus inputs. And compressor and solenoid outputs. It's as dead simple as it can be.
And, you're talking about using $140 in spikes to avoid buying a $90 board that can power twice as many solenoids.

Originally Posted by magnets

This has the potential to be a really awesome control system, but these people really need to be more open with us. If they showed us what WPIlib and the driver station looked like now, then there wouldn't be such a shock in December, when we buy the systems, and we don't know what we're doing.

You'll not be shocked by the updates to either of these items. If you know what you are doing now, you'll know what you're doing with the new parts. If you have specific questions about anything, there are a lot of people on this board who have a spent some time Alpha testing. Feel free to ask.

[magnets]

Originally Posted by Jefferson

I'm not sure how it's "way more complicated than it needs to be". It has power, pressure switch, and CAN bus inputs. And compressor and solenoid outputs. It's as dead simple as it can be.
And, you're talking about using $140 in spikes to avoid buying a $90 board that can power twice as many solenoids.

I charge my pneumatics with a relay that is many years old. I buy a single $34.95 relay for shifting. I would not like to purchase $180 worth of pneumatics boards, instead, I would like to purchase 1 new spike relay for use on the competition robot, and use an older relay for our practice robot. That is spending 5 times as much money to accomplish the same task. In my personal opinion, providing CAN functionality to the pneumatics board is too complicated.

[AdamHeard]

Originally Posted by magnets

I charge my pneumatics with a relay that is many years old. I buy a single $34.95 relay for shifting. I would not like to purchase $180 worth of pneumatics boards, instead, I would like to purchase 1 new spike relay for use on the competition robot, and use an older relay for our practice robot. That is spending 5 times as much money to accomplish the same task. In my personal opinion, providing CAN functionality to the pneumatics board is too complicated.

Then use a spike relay off one of the relay ports. If you feel you need more than 4 spikes, some of the general purpose digital i/o can be set up to run relays (probably).