View Full Version : Jaguars and Talons
FS Robotics
16-01-2014, 21:37
Are Jaguars and Talons compatible? Nothing that I have found shows that there are any key differences that would create problems if we used both on the same bot, but I'm curious if anyone has insight on this. Thanks!
geomapguy
16-01-2014, 21:38
Are Jaguars and Talons compatible? Nothing that I have found shows that there are any key differences that would create problems if we used both on the same bot, but I'm curious if anyone has insight on this. Thanks!
Depends how you want to use them.
-Mixing in the drivetrain - NO
-Elsewhere - Sure, unless the two motors are sharing the same load
Depends how you want to use them.
-Mixing in the drivetrain - NO
Why not?
-Elsewhere - Sure, unless the two motors are sharing the same load
Why not?
geomapguy
16-01-2014, 22:07
Why not?
Why not?
Well Ether,
I would suspect because the motor curves are different??
What's the exact explanation or correction?
Well Ether,
I would suspect because the motor curves are different??
What's the exact explanation or correction?
He means the controller curve. A Victor has a different curve from the Talon, which has a different curve from the Jaguar. Mix up the motor controllers and some motors are not spinning as fast as others, causing strafing in the drivetrain, or abrasion in a motor set, when a motor needs to fight the less power being received by a different motor. In real life, while these issues are quite minimal because a lot of work is done to make the curve more linear, It isn't what you want for precision operation!
The code is something that hasn't been discussed too much. Using the same motor controller for at least a specific component will ensure that the code stays cleaner and easier to understand. You don't want to look at programming jargon, where half a subsystem runs code for a Victor/Talon, and the other half runs the Jaguars. That just becomes very messy and can be hard to understand.
I would suspect because the motor curves are different??
It's OK to combine motors with different motor curves.
For example, a CIM and a mini-CIM on the same gearbox.
geomapguy
16-01-2014, 22:30
It's OK to combine motors with different motor curves.
For example, a CIM and a mini-CIM on the same gearbox.
My bad, as Yash 101 said, I meant controller curves
My bad, as Yash 101 said, I meant controller curves
So why is that not OK,if it's OK to combine motors with different curves?
geomapguy
16-01-2014, 22:36
So why is that not OK,if it's OK to combine motors with different curves?
Hmmm good point. I just always assumed it was just bad practice.
SO does it have much effect???
Hmmm good point. I just always assumed it was just bad practice.
SO does it have much effect???
If you're trying to do closed-loop control there might be situations where it might be preferable to match the controllers.
Hey CD, any teams out there who used 1 Jag and 1 Talon for each 2-CIM gearbox on each side ?
Rynocorn
17-01-2014, 13:38
Slightly different but same principal I guess: This year we had 3 884 victors from vex and then 1 884 from Ifi and there was a significant difference in our drive train as it would turn ever slightly to the side. However, those were not jaguars or talons so not sure about that combination of those two. Also, all of the victors were tested beforehand and worked well. Just figured this would help a bit even if it wasn't what you were looking for exactly
geomapguy
17-01-2014, 13:51
3 884 victors from vex and then 1 884 from Ifi
Well........IFI owns VEXPro so essentially it was just an older 884 and a newer one. There was a year there where they rebranded their stuff. So accordingly, we have IFI spikes and VexPro Spikes
This year we had 3 884 victors from vex and then 1 884 from Ifi and there was a significant difference in our drive train as it would turn ever slightly to the side.
Unless you swapped the two left-side Vics with the two right-side ones and observed that the problem switched sides, you can't conclude that the turning was caused by a difference in the Vics.
nuggetsyl
17-01-2014, 15:49
You could run unmatched tires on your car but I would not recommend it.
You could run unmatched tires on your car but I would not recommend it.
Not analogous.
FS Robotics
17-01-2014, 16:50
Not analogous.
Are you saying Victors and Talons are compatible? We have two victors and two talons (only one jag, that was a mistake). Would having half Victors and half Talons cause issues?
nuggetsyl
17-01-2014, 16:54
Not analogous.
And now 90% of the students are looking up the definition of analogous.
Talons and jags were built to do different things. Just like street tires and snow tires were built for different tasks. Sure they will drive your car but are they best suited for the task together?
Talons and jags were built to do different things.
They were both built to control the voltage applied to brushed DC motors.
They both use the same switching method and the same switching frequency.
Let's explore this together. What do you think would be some of the possible problems with controlling a 2-CIM gearbox open-loop with one Talon and one Jag, on each side of a skid-steer bot?
Are you saying Victors and Talons are compatible?
No. Not saying that.
I am searching for the source (hard data or well-reasoned theoretical considerations) of the apparent conventional wisdom that Jags and Talons are never compatible in any application.
We have two victors and two talons (only one jag, that was a mistake). Would having half Victors and half Talons cause issues?
I think that's a question worth discussing (and possibly testing). I wouldn't recommend risking your 2014 season on it.
Rynocorn
17-01-2014, 17:44
Unless you swapped the two left-side Vics with the two right-side ones and observed that the problem switched sides, you can't conclude that the turning was caused by a difference in the Vics.
We replaced all of the victors today with all 888s and no other changes and it is working perfectly now so unless we didn't have something screwed in tight then it probably was that as we are driving perfect now. I'm not sure though
kevin.li.rit
17-01-2014, 19:16
No. Not saying that.
I am searching for the source (hard data or well-reasoned theoretical considerations) of the apparent conventional wisdom that Jags and Talons are never compatible in any application.
I think that's a question worth discussing (and possibly testing). I wouldn't recommend risking your 2014 season on it.
I'm willing to try 1 Victor 888, 1 Talon on each side of my drivetrain. I really don't see a problem in doing it. What kind of measurements should I try and take?
Hmmm good point. I just always assumed it was just bad practice.
SO does it have much effect???
The effect would be neglegible. However, the different handling of motor controllers means that one motor would go through much more abrasion than the other.
While there are safety diodes, one motor may generate enough backflow because of reverse electro-motiv force, to destroy MOSFETs on these motor controllers. However, this is typically negligible!
It's funny how basic physics can explain so much :D. #ILovePhysics
kevin.li.rit
17-01-2014, 19:31
The effect would be neglegible. However, the different handling of motor controllers means that one motor would go through much more abrasion than the other.
While there are safety diodes, one motor may generate enough backflow because of reverse electro-motiv force, to destroy MOSFETs on these motor controllers. However, this is typically negligible!
It's funny how basic physics can explain so much :D. #ILovePhysics
How would one motor go through much more abrasion?
How would one motor go through much more abrasion?
It won't be pushing as much, so one motor will have to work harder to speed the other slacker up
kevin.li.rit
17-01-2014, 19:49
It won't be pushing as much, so one motor will have to work harder to speed the other slacker up
Okay, but in this situation the motors are mechanically linked and spinning at the same speed, so how much does this difference in voltage/current cause much more abrasion?
I'm willing to try 1 Victor 888, 1 Talon on each side of my drivetrain. I really don't see a problem in doing it. What kind of measurements should I try and take?
Skidsteer drivetrain, with 2 gearboxes with one Vic and one Talon per gearbox?
Or four gearboxes with one motor per gearbox?
Assuming the former, here are some ideas, in no particular order:
Just drive it around and see if it behaves normally. No weird noises or odd behavior.
If everything seems OK, do hi speed, lo speed, pushing, turning, reverse, fast throttle changes, etc.
Feel both motors on both sides and see if there's a noticeable difference in temperature (e.g. are both Vic-driven motors hotter than their Talon mates, or vice-versa).
If you can instrument your bot to measure current, that would be excellent, but may be more work than you have time for.
It won't be pushing as much, so one motor will have to work harder to speed the other slacker up
This happens all the time. All motors are not created equal. Neither are all motor controllers.
While there are safety diodes, one motor may generate enough backflow because of reverse electro-motiv force, to destroy MOSFETs on these motor controllers. However, this is typically negligible!
It's funny how basic physics can explain so much :D. #ILovePhysics
Not quite. There are NO DIODES (on the outputs) on the controllers we use. The diodes would waste roughly 10% of the power by dissipating it as heat.
Also, the MOSFETs are set up so that if the robot is off, and you push it along, the power goes back and powers on the robot, no harm done. If you run while pushing the robot, you can get the cRIO to turn on.
kevin.li.rit
17-01-2014, 20:10
We're planning on a skid steer. 1 gearbox on eachside.
kevin.li.rit
17-01-2014, 20:11
This happens all the time. All motors are not created equal. Neither are all motor controllers.
I was thinking the same thing.
The Talon seems to be a bit more efficient (less voltage loss through the controller), providing effectively a higher motor output than the Jaguars do. Also, back current is handled differently if the wheels were manually turned on the drivetrain.
Having said that, there is a slight loss in efficiency overall but its mostly negligible, and certainly not dangerous. I would think the variance in motor performance due to wear/damage and the imperfections in the drivetrain (gear meshing, etc) would outweigh the difference in speed controllers.
To tell you the truth, it really comes to aesthetics, using the same component everywhere, and tidying the code up a bit by using similar code!
The Talon seems to be a bit more efficient (less voltage loss through the controller),
Not disagreeing, but I'm wondering what you mean by "seems to be". What observations gave you that impression?
back current is handled differently
Can you provide a link to the source of this info? I'm wondering if it's still true for the new Talons (changed switching method).
Having said that, there is a slight loss in efficiency overall but its mostly negligible, and certainly not dangerous. I would think the variance in motor performance due to wear/damage and the imperfections in the drivetrain (gear meshing, etc) would outweigh the difference in speed controllers.
That's my working hypothesis, until disproven (by hard data or sound theoretical reasoning).
Everyone up to this point hasn't touched on what I would think is a very important(and not too difficult) solution for any problems this might cause: gyro correction code. As Ether said, not all motors or motor controllers are created equal. In fact, no motors or motor controllers are created equal. There will always be minor differences even between parts of the same model. If you add a gyroscope to your robot, however, along with some code to automatically correct your output voltages if it starts to veer off course, you solve this problem pretty effectively. It allows you to mix and match motor controllers(we try not to, but it's handy for emergency repairs), and our drivers can always drive in a straight line without having to compensate for any possible difference in the motors.
So now I'm interested: how do other teams feel about this?
kevin.li.rit
17-01-2014, 22:48
Everyone up to this point hasn't touched on what I would think is a very important(and not too difficult) solution for any problems this might cause: gyro correction code. As Ether said, not all motors or motor controllers are created equal. In fact, no motors or motor controllers are created equal. There will always be minor differences even between parts of the same model. If you add a gyroscope to your robot, however, along with some code to automatically correct your output voltages if it starts to veer off course, you solve this problem pretty effectively. It allows you to mix and match motor controllers(we try not to, but it's handy for emergency repairs), and our drivers can always drive in a straight line without having to compensate for any possible difference in the motors.
So now I'm interested: how do other teams feel about this?
Well the original subject was whether or not using a talon and jaguar to drive 2 cims on each side of the drive train (with pparently 1 gearbox on each side) would cause any problems. Then subsequently the OP asked about 1 talon and 1 victor 888 combo.
In this situation the veering you're talking about wouldn't be caused by the motor controller combination.
In this situation the veering you're talking about wouldn't be caused by the motor controller combination.
I don't see why it couldn't be--your correction code doesn't care what causes the veering, it just adjusts the magnitudes to compensate.
kevin.li.rit
17-01-2014, 23:13
I don't see why it couldn't be--your correction code doesn't care what causes the veering, it just adjusts the magnitudes to compensate.
Well I'm assuming the same motor controller combination on each side(left and right), i.e, 1 talon and 1 jaguar, 1 talon and one victor 888, or 1 victor 888 and 1 talon. Same motors, on each side, same gearbox etc. Then the veering would be cause by variations in individual motor controllers of the same or different make, motors, gearboxes etc. and not really the combination of the two different motor controllers. With the motors mechanically linked they would always spin at the same speed.
That being said I do agree that correcting for veering is useful as I have done this in the past to help correct for the variation that mainly arises after my motor output.
cadandcookies
18-01-2014, 23:58
To tell you the truth, it really comes to aesthetics, using the same component everywhere, and tidying the code up a bit by using similar code!
To borrow from Ether, why do you think that it comes down to aesthetics?
I'm interested in this question too. My gut tells me that you might have some programming fun (I don't know what the controller curves are like), and it would be more work, but you could probably do it. That being said, I can't say electronics and motors are my forte, so I'll be looking into it a little bit more as time allows.
kevin.li.rit
15-02-2014, 22:41
Skidsteer drivetrain, with 2 gearboxes with one Vic and one Talon per gearbox?
Or four gearboxes with one motor per gearbox?
Assuming the former, here are some ideas, in no particular order:
Just drive it around and see if it behaves normally. No weird noises or odd behavior.
If everything seems OK, do hi speed, lo speed, pushing, turning, reverse, fast throttle changes, etc.
Feel both motors on both sides and see if there's a noticeable difference in temperature (e.g. are both Vic-driven motors hotter than their Talon mates, or vice-versa).
If you can instrument your bot to measure current, that would be excellent, but may be more work than you have time for.
We did some low speed maneuvers today with 1 talon and 1 victor 888 on each side of a 2 CIM configuration on the left and right sides. No noticeable problems and neither motor felt warm to the touch despite a loose bolt in one of the wheels grinding on the chassis.
We observed a while ago that running motor open as jaguar would not drive a talon. But opening as a talon would drive a jaguar. Does this make sense?
Tim
Joe Ross
17-02-2014, 11:20
We observed a while ago that running motor open as jaguar would not drive a talon. But opening as a talon would drive a jaguar. Does this make sense?
Tim
I would expect a Talon run as a Jaguar to hit full speed well before the end of the travel of a joystick, but otherwise work.
DjScribbles
17-02-2014, 12:28
I think the OP's question is much simpler than the discussion that it's generated, so I'm going to try to clear up a few things (I hope).
Your robot can have any mix of motor controllers, you do not have to use a single type; if you have 2 talons, 2 victors, and 2 jaguars, you can put them all on your robot.
There is nothing to prohibit you from mixing your motor controllers on your drive train or on appendages that share a load (such as an arm with two motors that share the task of rotating it), they will still function just fine.
However, conventionally, people balance their motor controllers and motors when they need to stay synchronized (such as your drive-train, or an arm). Just as two different types of motors will turn at different speeds/torques when an identical voltage is applied, two different motor controllers will output different voltages when a PWM signal is applied (although they are somewhat close), hypothetically, a command of 71% power may yield an output 8.3V on a Victor, and 8.8V on a Jaguar, not a big deal if they are doing different tasks, but if they are driving the left and right sides of your drive train, your robot will veer to one side unless it is corrected in code (or by the operator).
In a 4 motor drive train, if you wanted to use 4 CIM motors with 2 jaguars and 2 talons, you certainly could, and any arrangement of these would make the wheels turn, but putting 1 jag and 1 talon on each side would typically yield better results.
Ultimately, the speed controllers are all the same, they take a PWM input and a voltage input, and use it to yield some voltage output as a percentage of the input (the PWM defines the percentage target). The differences are in how they interpret that PWM signal and how accurately and precisely they are able to control the output voltage; this is often discussed as a speed controllers curve, or linearity, which shows the relationship between the actual output % compared to the requested output from the pwm.
Hopefully this clears things up for the OP, and explains why his simple question resulted in quite a complex discussion. I've probably gotten a few things wrong, so feel free to enlighten me :)
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.