Jaguars

[rhp3794]

Hi everybody we are trying to buy new jaguars. but searching in different pages. the jaguars appears that doesn't exist anymore for buying it. searching in vex. and searching in other pages.

Do you know about something?

[Thad House]

http://www.vexrobotics.com/vexpro/motor-controllers/217-3367.html

[MichelB]

Most everyone is switching to Talons. This seems to be the way to go for the upcoming season.

[MechEng83]

Most everyone is switching to Talons. This seems to be the way to go for the upcoming season.

What data do you have to support this claim?

[geomapguy]

What data do you have to support this claim?

I second this, many teams are still all-victor or all-jaguar.

[brennonbrimhall]

What data do you have to support this claim?

I personally have no data to back MichelB's claim up, but Talons have been a significantly more pleasant experience to work with than Victors or Jaguars. Not only are they reliable and pretty indestructible, but they have the smallest footprint too. And most of the time, you don't need a fan on them. As a programmer, they feature a more linear response, so I don't have to compensate for that in my code.

It's my opinion that unless I'm using CAN, Talons are the way to go.

That being said, IFI has definitely addressed the community's constructive feedback on its products, but I'm not well-versed in the new(er) Victor 888 or updated Jaguars to comment on this.

[Jacob Bendicksen]

I personally have no data to back MichelB's claim up, but Talons have been a significantly more pleasant experience to work with than Victors or Jaguars. Not only are they reliable and pretty indestructible, but they have the smallest footprint too. And most of the time, you don't need a fan on them. As a programmer, they feature a more linear response, so I don't have to compensate for that in my code.

Agreed. I don't have any hard evidence to back it up, but they're just nicer to work with. Smaller, lighter, hardier, and unless they're on a drive motor they don't need a fan. They're also close to impossible to smoke with metal shavings unless you're trying.

[MichelB]

What data do you have to support this claim?

I, personally have not worked with Talons but the concensus seems to be that they are easier to work with and a better piece of hardware overall.

[IndySam]

We will never use a Jaguar again, they just caused way too many problems.

We used a mix of talons and victor 888 last year and were very happy with the performance of both. Unless there are rule changes we will do the same this year.

[MrBasse]

How can everyone claim that Talons are tougher than another speed controller. We have been using the same Victors for four years straight and haven't had a failure yet. Talons have only been around for a year, so time will tell. Jaguars on the other hand...We've cooked 9 of those (one with actual flame coming out of the vents) and avoid them as much as possible unless we run out of Victors. I'm fairly sure that I can get smoke to come out of one just by looking at it, but it might also have something to do with our shop environment...

In my view, availability of the Talon during season has been the biggest issue with it, so we have yet to get to play with one. I'm excited to get one, but we don't have the funds to pick up enough to use without mixing speed controllers.

[MichaelBick]

While we wouldn't use jaguars on the comp bot, I've heard they are really good for protos because of they build in PID and current sensing.

[Anupam Goli]

You can purchase additional jaguars here: http://www.vexrobotics.com/217-3367.html

I always thought the CAN bus capabilities of the jaguar were very interesting and if the hardware was more reliable, I'd use the jaguar more often.

[MechEng83]

We ended up using all 3 controllers at various points on our 2013 robot. (Though only 2 at any one time.)

Jaguars are great because of following reasons (and we utilize each one):

Current sensing
Internal PID control
Forward/Reverse Limit Switches
CAN

Jaguars have the following drawbacks (each of which we've experienced):
Relatively low current limits (Though it seems to have been raised with the latest firmware update)
Poor CAN (physical) connection robustness
Highly sensitive to swarf (metal debris) -- though the new ones have a conformal coating.

[brennonbrimhall]

How can everyone claim that Talons are tougher than another speed controller. We have been using the same Victors for four years straight and haven't had a failure yet. Talons have only been around for a year, so time will tell.

That's great if your Victors have lasted that long -- we just haven't had that experience, especially with Jaguars.

I don't know if 'tougher' is the better term, as much as I think that Talons are the better product. Let me explain:

When we made the switch to Talons, it combined the linearity of the Jaguar with the durability of a Victor (pre-Victor 888) plus metal shaving resistance. We've found that CTRE's attention to minor details really help set Talons apart: the PWM connection to the Talon is more secure than either Victors or Jaguars, the small footprint is helpful when squeezing electronics onto the chassis, the 4% deadband saves headaches from old joysticks, and the Smart LED helps debugging code.

[Pault]

Jaguars have the following drawbacks (each of which we've experienced):
Relatively low current limits (Though it seems to have been raised with the latest firmware update)
Poor CAN (physical) connection robustness
Highly sensitive to swarf (metal debris) -- though the new ones have a conformal coating.

Don't forget about a giant footprint.

And all those extra features can be a disadvantage too, because you now have the temptation to use them when often you would be better off without them.

In 2012 (rookie year) my team used Jaguars with CAN. I will just say that even the word CAN now instills fear into the hearts of our veteran students. In 2013 we wanted to use Talons, but the sold out, so instead we used Victor 888s, and had no complaints. This year we have bought 4 talons to put on the electronics test bed we are building, and if they really are as good as everyone says they are we will put them on our 2014 robot.

[yash101]

We used Victors last year, and kept cooking them, (because our motors kept shorting out, damaging their FETs. We haven't been having those problems anymore, because our shooter is disabled so we barely use it. Victors, themselves, are quite nice. They were lightweight last year, compared to jaguars, and they work nicely. I do not have too much experience with Jaguars, but it seems though the feature set is quite great. The CAN interface is nice, along with the PWM interface. These can be networked through an RS232 interface too! There are a lot more features that I won't list, but here's the manual: http://content.vexrobotics.com/docs/217-3367-VEXpro_Jaguar_GettingStartedGuide_20130215.pdf. As mentioned before, a big problem in these motor controllers is their maximum continuous current. In the product page (http://www.vexrobotics.com/217-3367.html), here is the specs table:

Never Limit Progressive Limit Immediate Limit
Pre v107 40A 50A 60A
v107 40A 50A 92A


Here are the Talon specs:
Page: http://www.andymark.com/Talon-p/am-2505.htm

Input voltage: 6-28 VDC
Continuous current: 60 A (above 40A continuous we recommend adding this fan)
Peak current: 100 A
Input PWM signal: 0.9-2 ms @ 333 Hz
Input resolution: 10-bit (1024 steps)
Output resolution: 10-bit (1024 steps)
Output switching frequency: 15 kHz
Talon SR: Synchronous sign-magnitude rectification
Smart LED, blinks proportional to throttle, now with obvious change from 99% throttle to 100%
Simple calibration
User selectable brake/coast
4% neutral dead band
Linear throttle response

I hope this helped. This is a comparison between Talons and Jaguars.

[Jared]

The two most important features of a speed controller are reliability and cost. We don't want to worry about swapping out controllers after a match, and we really want to avoid failures during the match. Reliability is also important because we use previous years speed controllers on the practice robot. This year, our practice bot had victors from 2003 and 2004 that have already been used on at least two robots. Victors are the only controller with this type of reliability. They are also the cheapest.

As for the extra features, I think they're a waste. We end up wrapping the speed controller class to add a few utilities, so it's easy to add a pidcontroller and and the function to make the output linear. The jaguars don't really have any features that can't be done on the crio.

[yash101]

I agree. Victors will take a lot of abuse before failing. The only failures we had were because of a shorted motor. If they added temperature sensors to the MOSFETS, they could allow it to do an emergency shutdown before damage!

[MechEng83]

The jaguars don't really have any features that can't be done on the crio.

Please let me know where the current sensing feature is on the cRIO. My team has been unaware of this feature.

[techhelpbb]

Please let me know where the current sensing feature is on the cRIO. My team has been unaware of this feature.

With minimal hardware the feature could exist but I doubt it would be legal under the FRC rules as it would require measuring the voltage drop over a specific length of wire using the analog bumper or some other analog accessory. It can be done with a series resistor like in the Jaguar but in doing so you'd be in the path of the current with more than wire. I doubt it would be legal on the output side of an electronic motor control as it would make a connection between the bridge in the electronic motor control and the motor back to the control system. It might not be legal on the power input side of an FRC electronic motor control either. Course on the power input is both the current to the motor and the electronics in the motor control (which are small in comparison to the motor). There's also the issue that the current at the power input point will have aspects not conforming to the current at the output to the motor (the bypass diodes, various features, the resistance of the MOSFETS).

Don't think of trying this trick unless you understand the difference between a single ended and differential A/D input.

[MechEng83]

With minimal hardware the feature could exist but I doubt it would be legal under the FRC rules as it would require measuring the voltage drop over a specific length of wire using the analog bumper or some other analog accessory. It can be done with a series resistor like in the Jaguar but in doing so you'd be in the path of the current with more than wire. I doubt it would be legal on the output side of an electronic motor control as it would make a connection between the bridge in the electronic motor control and the motor back to the control system. It might not be legal on the power input side of an FRC electronic motor control either. Course on the power input is both the current to the motor and the electronics in the motor control (which are small in comparison to the motor). There's also the issue that the current at the power input point will have aspects not conforming to the current at the output to the motor (the bypass diodes, various features, the resistance of the MOSFETS).

Don't think of trying this trick unless you understand the difference between a single ended and differential A/D input.

That wasn't actually a serious question, but thanks for the answer! For the Talons we used on our drive train, we experimented with using a wire coiled around one of the wire leads and got some values in a custom circuit. I'm not sure of all the details, our electrical mentors worked on it.

[Munchskull]

As was wondering, at the rick of sounding ignorant, what is the performance difference between Spikes, Jaguars, and Talons? I know the Jaguars have CAN but what that does is a little of a mystery to me.

[Ether]

what is the performance difference between Spikes, Jaguars, and Talons?

A Spike is simply a couple of relays, whereas Jag and Talon are speed controllers.

Performance curves for Jag, Talon, and Vic may be found here:

http://www.chiefdelphi.com/media/papers/2720


I know the Jaguars have CAN but what that does is a little of a mystery to me.

In order to help you, it would help to know if it is a mystery because you've read and don't understand the documentation, or because you can't find the documents.

[techhelpbb]

That wasn't actually a serious question, but thanks for the answer! For the Talons we used on our drive train, we experimented with using a wire coiled around one of the wire leads and got some values in a custom circuit. I'm not sure of all the details, our electrical mentors worked on it.

It sounds like your team measured the electromagnetic field around the wire with a circuit that works like a current sense transformer. It doesn't actually connect to the circuit it merely robs some of the magnetic field around the wire (which is generally there as long as current flows) and converts it to something you can measure. That might let you skate on the FRC rules about getting in the circuit path between the electronic speed control and the motor. Plus when done with a current sense transformer you bypass the grounding issues as one leg of the output can be tied to that electronics ground with little impact on the electronic motor control or the motor. Using a current sense transformer it's important to note that measuring AC and DC currents require different considerations.

There are Allegro hall effect sensors that can measure circuit current if you dig around they were recommended up on ChiefDelphi before (oddly almost precisely 3 years ago):
http://www.chiefdelphi.com/forums/archive/index.php/t-87568.html

I am interested if these are legal in FRC in competition. However that might be a side track for this topic.

[Jared]

Please let me know where the current sensing feature is on the cRIO. My team has been unaware of this feature.

It's true that there isn't a direct way to measure current on the crio, but for the price of a jaguar, you could buy (at least in 2012) a victor and a current sensor.
Per the 2013 rules you are permitted to have a low impedance sensor to measure current.

[techhelpbb]

It's true that there isn't a direct way to measure current on the crio, but for the price of a jaguar, you could buy (at least in 2012) a victor and a current sensor.
Per the 2013 rules you are permitted to have a low impedance sensor to measure current.

Please state the rule you are referring to. There are some caveats about what you can do this with. For example you can use a low impedance sensor to measure current but not to the D-Link radio while in competition.

[Ether]

It doesn't actually connect to the circuit it merely robs some of the magnetic field around the wire

Well, it actually is connected to the circuit, albeit inductively via the transformer formed by the lead wire and the coil wrapped around it.

Since the current to the motor may contain a large ripple under various operating conditions due to the interaction of the speed controller PWM switching, the motor inductance, and the motor commutation, there may be some issues involved in interpreting the measurements.

[techhelpbb]

Well, it actually is connected to the circuit, albeit inductively via the transformer formed by the lead wire and the coil wrapped around it.

Since the current to the motor may contain a large ripple under various operating conditions due to the interaction of the speed controller PWM switching, the motor inductance, and the motor commutation, there may be some issues involved in interpreting the measurements.


Entirely correct. I just wonder if it is even allowed in FRC competition external to an approved FRC electronic motor control (regardless of the issues of making it work).

Of course by extension of being connected by inductance anything that passes through the magnetic field around the wire and can have a current induced in it is connected in a similar way albeit poorly. Parts of the robot that are not designed to be electrical but can carry a current for example. So generally not recommended to wrap a nice piece of copper wire around your other wires to act like a wire guide or retainer.

[Jared]

Please state the rule you are referring to. There are some caveats about what you can do this with. For example you can use a low impedance sensor to measure current but not to the D-Link radio while in competition.
I am referring to
[R47]

Custom circuits shall not directly alter the power pathways between the battery, PD Board, speed controllers, relays,
motors, or other elements of the Robot control system (including the power pathways to other sensors or circuits).
Custom high impedance voltage monitoring or low impedance current monitoring circuitry connected to the Robot’s
electrical system is acceptable, if the effect on the Robot outputs is inconsequential.

I was trying to point out that it is possible to measure the current draw of a motor like a jaguar does without having to use a jaguar. You can't measure the current draw of the radio in any way.

[techhelpbb]

I am referring to
[R47]
I was trying to point out that it is possible to measure the current draw of a motor like a jaguar does without having to use a jaguar. You can't measure the current draw of the radio in any way.

I have no doubt that this works (with the expected technical issues). I merely doubt as to whether or not it would be considered legal on a competition robot connected to a motor (at first glance R47 would not lead you to believe the radio is excluded either but it is excluded I asked in the official Q&A before this year). The closest topic I found from 2009 was this:
http://www.chiefdelphi.com/forums/archive/index.php/t-74138.html

It merely states that the current sense resistor needs to be a very small value. It does have mention (by Al) of using the wire as a current sense resistor. Has anyone actually fielded this on a competition robot using the cRIO to process the measurements? There seems little sense to me to go back and forth over the myriad details (as often happens) if it might not be legal on the field.

Back on the topic of the Jaguar itself: is there a poll by team number of the teams that intend to use Jaguars on competition robots this year? I was helping a student make something that was an accessory to the Jaguar and with all the admirable effort IFI is putting in to these units I wonder if the number of teams using them is actually getting smaller. I am neither for or against the Jaguar in particular myself.

[TogetherSword8]

I am unaware of any poll on motor controllers that would be FRC wide. Although it would be interesting to see, especially if you could organize it by region (with Michigan being one of the major regions to watch).

As for the Jaguars, we used them exclusively this year, but it was mainly because of the CAN functionality. Talking about metal shavings, we have fried two this way, but in the Finger Lakes regional, we chopped off a full inch of off the top of our robot with the Jaguars still on (not advised, we forgot to remove all the electronics while doing this, lucky us), and they did not fry from any shavings coming off of the supports.

We also have a large number of Victors, and we used to be an exclusive Victor team, but we use the build in PID on a Jaguar a lot and since we know the Jaguar well from using it on every shooter wheel, we have used them on Drive Train since we know how to use them.
We have never used Talons, but based on what we have heard, it sounds like they would be better for drive train purposes on our team, but due to the built in PID on a Jaguar, we would still be using Jaguars for manipulators.

[Mike9966]

Hi everyone,
Our team, which is in it's 6th year, has used only Jag's for all of our history. We started with them because they were more interesting, and sophisticated. I think the the first year was PWM just because the old tan Jag's didn't have the 232-CAN converter capability that the black ones that came out the next year did.

We have fried I think 5 in the last 5 years, and I only know of one that wasn't our fault, it was the driver chip in that version of Jag's.

I encourage the team to NEVER do metal work with the electronics mounted, and I think that's good practice no matter what you are using.
Of course sometimes it's unavoidable, and then we cover things up, and use a shop vac to suck things up as we are drilling, etc.

We like the CAN, although I will say it can (no pun intended :ahh: ) be trying once and a while.
The feedback, and control it offers is great to work with though.

We did learn one lesson last year and that was not to use 10 or more Jag's on the CAN bus at a time when using the first black Jag as a serial-CAN converter. There were timeout issues and the robot wouldn't work properly half the time with 11 Jags.
We backed it off to 8 Jag's and everything suddenly worked great!

Anyways, we've already bought some more anticipating this years season.

A poll, or survey would be a great idea, how many use the 3 different brands of controllers, and how many use CAN?

Next year will the Talons have CAN? That'd be cool.

I was thinking about one of the posts about max current output. If one liked Talons for instance because they could put out 100 A continously, and one had 4 CIMs on the drive train, that'd be 400 A draw. I'm thinking your wiring and so forth wouldn't last long if you did that all the time.....
Maybe you don't want that capability?

Maybe you'd use the sensor in the Jag for current feedback to the software in the cRio, and when it approched say, 50A, a warning would come on and a routine might back off the current so as to not burn things up.
Sometimes drivers need to know when to stop pushing the joystick forward if the bot isn't moving....... lol I'm just saying :)

Whatever you use, have fun and be safe,

Mike

[FrankJ]

It will be interesting to see what motor controllers are in IFIs product unveil on 12/11/2013. We have used talons, victors & jags. I would put them all in the reliable category. Most of our failures have not been really a drive issue. (reversed power connections, swag, etc are not controller failures). If you running motors at high stall currents (IE prolonged pushing matches you are probably better off with a talon or victor. We have sucessfully used canbus on the jags with few issues.

I expect to see more canbus on the 2015 platform. Maybe one the Alpha testers could comment on that.

[Joe Ross]

I expect to see more canbus on the 2015 platform. Maybe one the Alpha testers could comment on that.

The 2015 power distribution panel is CAN enabled to allow reading of status. The Pneumatics Control Module also uses CAN for controlling solenoids and configuring the module. CAN Jaguars will be supported, but there's still some bugs in the current implementation (that's why we're alpha testing). No other CAN Speed controllers have been announced.

[brennonbrimhall]

No other CAN Speed controllers have been announced.

I believe I saw the CTRE folks at their IRI booth demoing a CAN-Based Talon (controlled via a 2CAN and hooked up into the 2015 PDB), on track for being legal in the 2015 season. Can anyone else confirm this?

[hzheng_449]

We did learn one lesson last year and that was not to use 10 or more Jag's on the CAN bus at a time when using the first black Jag as a serial-CAN converter. There were timeout issues and the robot wouldn't work properly half the time with 11 Jags.
We backed it off to 8 Jag's and everything suddenly worked great!
Mike

Hi Mike, just out of curiosity was your team using a RS-232 CAN adapter or a 2CAN? I ask since my team is currently experimenting with using a 2CAN, and it would be concerning (to say the least) if timeout issues still happen w/ the 2CAN.

[s1900ahon]

Hi Mike, just out of curiosity was your team using a RS-232 CAN adapter or a 2CAN?

As his post said, he was using the RS232-to-CAN bridging capability of a black Jaguar (the first in the chain).

I sent Mike a PM indicating that reliability could be improved by ensuring that the CAN bus was terminated with 100 ohm resistors (one at each end). Yes, I know that the CAN bus specifies 120 ohm termination at each end, but the 100 ohm values was specified by LMI/TI to improve bus timing in the change from dominant to recessive.

One thing that I'd suggest is that teams use CAN for the things that need CAN and use PWM elsewhere. PWM is "free" CAN bandwidth.

[rsegrest]

We will never use a Jaguar again, they just caused way too many problems.

We used a mix of talons and victor 888 last year and were very happy with the performance of both. Unless there are rule changes we will do the same this year.

We agree completely. We blew up too many jags doing bad things to them. Went all Victor and Talons last year. Think we may go all Victors this year with Talons to supplement. Overall we were pleased with both and MUCH more pleased with either over the jags.