FRC Blogged: - Motor Controllers

[Hallry]

Found here: http://www.usfirst.org/roboticsprogr.../blog-10-03-12

FRC Blogged - Motor Controllers

Blog Date: Wednesday, October 3, 2012 - 12:11
Motor Controllers

You may have noticed that traditional retailers of the Jaguar motor controller have either stopped selling Jaguar or have limited inventory. This blog post is to explain why and what you can expect going forward.

First, what I think you are most interested in knowing: the Jaguar will continue to be an available and legal motor controller for the 2013 FRC season. Please understand that full, detailed rules will not be disclosed until Kickoff, and supersede all other documentation – like blog posts. However, you can be assured that Jags will be legal.

Now, some background. The Jaguar motor controller was developed and designed by FRC mentors who worked for a company in Texas called Luminary Micro. Jaguar was offered as an alternate motor controller with different features for FRC and adopted in to the FRC control system and the FRC Kit of Parts. Shortly after the release of Jaguar, Texas Instruments purchased Luminary Micro and integrated Jaguar in to the Texas Instruments product line.

Texas Instruments recently offered to license Jaguar to FIRST for free, and FIRST accepted. This means that the Jaguar design, software, firmware – everything - is free for FIRST to use in and modify for FRC. Although the Jaguar design is open source already, this allows FIRST access to certain proprietary components such as source code, etc.

Because FIRST doesn’t have the resources to maintain and service a full product line in-house, we released a Request for Quote earlier this year soliciting a partner who could continue Jaguar as a product for FRC. Several excellent companies responded with compelling proposals. Ultimately FIRST partnered with Innovation First (IFI) for Jaguar production and distribution.

Since then, Texas Instruments has worked with IFI to transfer documentation, fixtures, knowledge, code, etc. to IFI so that they can effectively adopt Jaguar.

Going forward, IFI will be manufacturing, selling, supporting, and donating Jaguars to the 2013 Kits of Parts. There are two notable changes in the Jaguar motor controller for the 2013 FRC season. First, the PCB will be conformal coated. Second, the firmware will be modified to remove the current limit protection (this is for black Jaguars only and will not apply to the grey Jaguars). Details about the product itself will be posted on the IFI website here: http://www.vexrobotics.com/products/.../217-3367.html

Regarding the 2013 Kit of Parts, IFI is generously donating 2 Victors and 1 Jaguar to each Rookie Kickoff Kit plus a motor controller Product Donation Voucher (PDV) for two motor controllers to all 2013 FRC teams. Ultimately, the PDV allows teams to choose between getting 2 Victors, 2 Jaguars, or one of each, based on the team’s motor controller preference, for the cost of shipping.

While Texas Instruments no longer supplies Jaguars, they will make every attempt to honor the warranty provided through the specific supply channel upon the completion of a product review using their quality tracking process.

FIRST is excited about this new opportunity for more customization of kit items. We are very grateful for the support and generosity from both Texas Instruments and IFI to make this new initiative successful. Thank you to these two important suppliers - this commitment to FIRST illustrates that they truly support our mission.

I’ll blog again soon.

Frank

[Akash Rastogi]

I like this new Frank guy. He knows what the people want to hear.

More IFI is best IFI.

Nothing about the Talon? I hope it becomes FRC legal.

[Akash Rastogi]

Originally Posted by SuperNerd256

Nothing about the Talon? I hope it becomes FRC legal.

Teams are Beta Testing it.

[andreboos]

The post mentions a conformal coating - am I right to hope that this would reduce their vulnerability to stray metal shavings?

[jason701802]

What is the benefit of removing the firmware current limit? This seems like great way to blow a lot more Jags.

[BrendanB]

Good to hear!

[PAR_WIG1350]

Originally Posted by jason701802

What is the benefit of removing the firmware current limit? This seems like great way to blow a lot more Jags.

It did have much tolerance for excursions from the 40 amp limit despite the fact that the hardware was perfectly capable of handling it. This had the nasty consequence of disabling motors mid match and preventing the robot from performing effectively.

[jason701802]

Originally Posted by PAR_WIG1350

It did have much tolerance for excursions from the 40 amp limit despite the fact that the hardware was perfectly capable of handling it. This had the nasty consequence of disabling motors mid match and preventing the robot from performing effectively.

So wouldn't the real solution be to increase the current limit instead of completely disabling it?

[Brandon_L]

Props to TI for making the transition seamless!

[andreboos]

Originally Posted by PAR_WIG1350

It did have much tolerance for excursions from the 40 amp limit despite the fact that the hardware was perfectly capable of handling it. This had the nasty consequence of disabling motors mid match and preventing the robot from performing effectively.

I'm somewhat worried about this change from a control perspective - last year there were reports of Jaguars losing their state when they lost power, so any CAN configuration commands (PID, voltage ramping, e.g.) had to be re-sent. This will become much more common if overcurrent results in tripping breakers and resetting Jaguars rather than just disabled motors.

I am hoping that this can be solved by enabling the overcurrent protection by default, but allowing it to be disabled via CAN if a team knows what they are doing, or they do not care if the Jaguar resets. This way, if it does reset, the overcurrent protection will be reenabled when it does.

[Joe Ross]

Originally Posted by Akash Rastogi

Teams are Beta Testing it.

I assume you are talking about 1718? They are testing the Talon, but not as part of the FIRST sanctioned beta test. The FIRST Beta test hardware has not been announced.

[Paul Copioli]

Originally Posted by jason701802

So wouldn't the real solution be to increase the current limit instead of completely disabling it?

We are thoroughly testing the firmware changes with some of the most vigorous testing we know (and believe me, the guys we have here have beat up some FRC robots in their day). The software limit was overly conservative protecting FETs that just really didn't need that kind of protection.

We are confident these changes (conformal coating and firmware changes) will make the Jaguar perceived match robustness much higher. The firmware current limit caused Jags to cut power to motors when they just didn't need to.

We have had extensive technical conversations with the engineers responsible for the jaguar development and are confident this is a positive change.

Paul

[Tom Bottiglieri]

Originally Posted by andreboos

I'm somewhat worried about this change from a control perspective - last year there were reports of Jaguars losing their state when they lost power, so any CAN configuration commands (PID, voltage ramping, e.g.) had to be re-sent. This will become much more common if overcurrent results in tripping breakers and resetting Jaguars rather than just disabled motors.

I am hoping that this can be solved by enabling the overcurrent protection by default, but allowing it to be disabled via CAN if a team knows what they are doing, or they do not care if the Jaguar resets. This way, if it does reset, the overcurrent protection will be reenabled when it does.

There are plenty of teams who know what they are doing and elect to not use the CAN bus, ourselves included. The current limit seemed to be a bit too "safe" and was pretty annoying.

[JesseK]

Originally Posted by jason701802

So wouldn't the real solution be to increase the current limit instead of completely disabling it?

The real solution would be for teams to stop designing mechanisms that consistently hit the 40A + mark for extended periods -- which means more teams need to do the math.