Speed Controller Preference

Technical Electrical
#10

During competition season we used 4 Jaguars (2 drive, 1 pickup, 1 dump) and 1 Victor (for the second dump motor). We have to use the one victor simply because we ran out of room and needed a second motor for our dumper to work effectively.

We had no problems with the Jaguars all season…

Then when we started making enhancements for IRI we replaced the 4 Jaguars with 4 Victors. We made this switch because we needed the extra real-estate to make room for traction control hardware. We fit 5 Victors (with room for one more) in the same foot print the 4 Jaguars previously occupied.

We haven’t made a determination on what we’ll use for 2010 since we don’t know what will be available and what our size constraints will be.

#11

I’m really excited about CAN. I think that about 10% of teams will really dig into it this year, and that they will create some solutions with it that (although not game changing) will make us rethink a few common design memes. Hopefully the year after, newer teams will benefit from “how you CAN make your robot drive straight” tutorials.

I’m not saying that CAN will let you do things that are impossible now, I’m just hoping it lowers the ((entrance)) bar for some of the fundamentals.

#12

Most of the features that the CAN bus allows us to do are easily doable now.
The speed (Voltage) control mode, is the mode we are using through PWMs.
The Analog position input (Potentiometer)
has been done by almost everyone using an arm (PID control)
The Encoder position control
Can be done this year, use a PID controller and have your setpoint be a # of ticks
Limit switches - Digital input,
if(input && (abs(speed)/speed) == limitDirection) motor.set(0.0);
else motor.set(speed);

Though, the real advantage of using CAN is that the Jags handle the monitoring, and control internally.

Which can provide the fastest response possible in a closed loop.
(At least with the motors & controllers we can use)

And any faults can be read by the RIO,
ex: does that red slow blink mean the limit switch has been triggered or is the jaguar dead?
Conditions like: over-current, under voltage (ie 6v from battery), limit switch, or over temperature; can easily be read from like the driverstation LCD instead of guessing.

Also having access to all the motors variables (current, voltage, torque, speed) without having to put in a large coil of wire for a current sensor is nice.

http://www.luminarymicro.com/products/mdl_bdc.html
BD-BDC-DS-01 Datasheet for Brushed DC Motor Control Module

#13

We used 8 Victors, simply because Jaguars would not fit, and were not aesthetically pleasing in our design.

EDIT: Also, we already had plenty of Victors from previous seasons, whereas we would have needed to purchase more Jaguars.

#14

We used both last year, Jags on the drive train, victors on everything else.

Victors; Tried and true, fairly bullet-proof.

Jaguars; Nice linear response, motors were much quieter with higher frequency response, held up well so far. Waiting for CAN bus to hopefully simplify wiring.

We’re happy with both, hopefully we will be allowed to continue to choose which controller we wish to use for our specific applications.

#15

We ended up using Victors last year, mostly because all of our Jaguars fried pretty quick. We’ll probably be using Victors again this year, as well.

#16

2834 used 4 jaguars:

Talking to other veteran teams in our area, we decided that the jaguars would work best, despite the slightly bigger size. Besides, there were 4 jaguars in the KoP, but there were not 4 victors.

#17

wow seems to be about even in the poll.

Might change assuming they DO allow the CAN on the Jags this year.

#18

Last year on 1776, we used mostly Victors. The Jaguars were used only for cost avoidance (by which I mean real money rather than Bill of Materials cost accounting money). We didn’t need fine control at low speed for the design we were using, and we did need more space and less mass.

Given the CAN bus, we might have done a different design to use it.

#19

Our team used victors for the competition. 2 for drive and one for turret rotation. Also a spike for the compressor. This year, we plan to use Jaguars as we got them working on our robot for drive. If we are feeling adventurous and we have enough, we will also use Jaguars for other things we need.

#20

Our team used all jaguars this year and I was pleased with both their performance and price. The only issues that our team had with the jaguars was, our jaguar that controlled the motor that moved our spiral which moved the moon rocks to the top, stop working in the forward motion twice. Unfortunately this happened in the finials causing us to lose a match. The weird part of this is that the jaguar stilled worked in reverse as well as the other functions, the only thing that did not work was the forward motion. When we opened them up on both occasions we found that the circuit board had melted in one spot near the fan. We sent them back to Luminary micro for analysis but we have not herd any thing back. Our belief is that it something unique about the way we had the motor set up and the amps it drew. Other then the problem above we had no problems and all the original jaguars are still on our machine. I would suggest that teams start using jaguars if they have not already. I also look forward to exploring the new features that we will be evadible this year.

Matthew Simpson

Team 75 VP

#21

4 Jags, 4 Victors. The Jags were upside down underneath the electronics board and powered the ‘twitch’ drive we had. We had zero issues running full foward to full reverse in less than a second when the twitch modules changed positions. We still used clear nail polish to keep the PWM’s though :wink:

#22

This type of failure accounted for over 40% of all specific failure reports according to Luminary Micro so it is by no means unique.

The Jaguar Failure Analysis report can be found here

#23

We used a mix of both. We found the Jags to definitely be smoother and more stable for low speed and position holding. A Victor on our turret drive would cause the turret to twitch with a tremor when trying to hold a position, and not be repeatable when homing to a position. The Jag in the same function would hold nice and steady, and repeat positions more accurately. We did however have a Jag fail on us at a critical time going into elims.

#24

2502 used 4 Jags, no failures. They did seem pretty responsive on the low end of the spectrum for our brushes, and they did also allow very slow movement from the drive train which were both nice.

#25

Victors for us. The Jaguars took up way too much space, and we didn’t need the control for our dumper. Didn’t need the hooks because we staple the wires down. The max current is 60 but our max breaker is 40. The limiter for keeping the motors from burning out isn’t needed because we limit the arm with programming. It is also a bit heavier than victors, which can add up.

#26

Team 1245 used Victors, mainly because our Jaguars burned out. Victors have higer overall output, but Jaguars have a smoother control curve.

#27

I really need to address a few things (misconceptions) here. The hooks referred to on the Jaguars (for the casual reader) are hold downs for the PWM control wires. (I assume that “stapling” indicates your robot used a wood electrical base. Generally staples are less than desirable as there is 5 volts and power common in the PWM cable and an errant staple could bring down your robot if it should short those two lines.
The max current specifications for the two controllers is based on the specifications of the FETs used and refers to the continuous current. (The Jaguar uses a FET that is slightly better in this regard) In both controllers and on our robots, current in CIM motors can achieve 129 amps in stall. There is nothing that limits that current, including the breakers (assuming wiring is not undersized or lengthy). The breakers will pass nearly 200 amps for short bursts without tripping.
The Jaguar has implementations that make it more linear at lower throttle values (primarily due to it’s higher switching frequency) and it has a slightly lower series “ON” resistance due to the FETs used. However, these differences may not be evident in your implementation. (For the price of a search, you will find a long discussion on this subject elsewhere on this forum)
For both controllers, full throttle means they are supplying the full battery voltage to the motor they are controlling. There is no modulation of the output at full throttle. The Jaguar does contain a small resistor that is used for current sense, it is equivalent to 6 inches of #10 wire. There is a voltage drop across this resistor that can limit the available supply voltage at higher currents and the current sense may cause the Jaguar to interrupt current flow for a short period of time.

#28

I thought there was a teensy bit of time where the Jaguar output is shut off even when at a nominal 100% output. Something to do with recharging a capacitor that keeps it on? I suppose it doesn’t count as “modulation”, though.

#29

This is true on the Grey Jaguars, but put emphasis on teensy. I’m not sure if it is true for Black Jaguars.