Jaguar Issues at the Last Minute

Technical Electrical
1

Intro: The Jaguars are no longer working, although they once were. They blink yellow even when the code is deployed and the robot enabled. Here is a list of all the things we have tried, as well as the circumstances surrounding the sudden loss of function.

For a start, the robot was working and driving during the first week. Recently, though, we built the final electrical box and it was rewired. Now, the Jaguars never change from blinking yellow always.

This is not an obvious solution. We have triple and quadruple checked every connection and cannot find the issue:
-battery is connected to PDD
-PDD is supplying 12V power w/ 20 amp breaker to digital sidecar
-Digital sidecar has two PWM cables each going to a jaguar
-Each jaguar is plugged into the PDD with a 40 amp breaker using large gauge wire.
-The black is going to black and the red is going to red.
-The V+/V- port wires are going to the PDD.
-The M+/M- port wires are going to the motors.
-All lights are good to go on cRIO, DLink, PDD, and digital sidecar.
-The large signal light is on and does respond to the robot being turned on.
-The PWM cables are plugged in correctly: black to ground and white to signal.
-The arguments for the constructor in the code(Java) are correct.

And so on and so forth. Solutions we have tried include:
-Replacing the PWM cables
-Replacing the wires going into/out of the Jaguars
-Replacing the Jaguars entirely
-Rolling back to old code when the robot was first working.

Further, I have dashboard values for the jaguars: when I move the joysticks(yes, they’re plugged in and coded), the values change. It seems that my input is being sent, but lost somewhere along the way between the laptop and the DLink.

Lastly, the relay lights still work; the compressor and solenoid relay buttons work completely, so the problem doesn’t appear to be with the digital sidecar.

If anyone can offer advice, I would be very thankful indeed.

2

From what you said it sounds like the PWM value isn’t getting to the Jaguar. Do you have a servo or a different speed controller you could test with?

Did you try downloading default code or some known working program.

Have you replaced the Sidecar 37 pin cable?

3

Yes, the 37-pin connector from slot 2 of the cRIO to the sidecar has been
fixed. Else, it would not have worked before.

We have tested with a total of four Jaguars; none of them ever stopped blinking orange.

The problem is definitely not with the code; there were literally zero changes to the driving portion from the time before when it still worked. All additions have been to other components: rangefinder, small motors, compressor/solenoid, etc.

4

Have you replaced the cable since it stopped working? Even a ribbon cable that worked at one point can break and stop passing PWM signals.

If you think you have swapped everything try swapping everything else. PD Board, Digital Module, and anything else that could even remotely fix the problem. Do them one at a time so you can isolate where the problem is.

Also double check the PWM connections and cables, change to different motor ports that you are using on the sidecar.

If you have a second digital module put in slot 4 (assuming 4-slot cRIO) and try to run of of that one.

5

Remove and turn the sidecar over, and shake. Maybe something got in there, this will shake it out.

Disconnect everything from the Sidecar, then just connect one Jaguar. There may be something weird with the outputs.

Do you have an oscilloscope? If so, look at the PWM signal itself.

Just grasping at straws here, but you never know.

6

Thank you for the suggestions; we’d thought of those, but we decided to consider alternatives before redoing everything.

Also, I just thought of another potential problem: is the PWM cable going into the Jaguar supposed to be. . .well, loose? When we insert a PWM cable into a Spike Relay, there is a good satisfying click. With the Jaguar, though, it just sort of slides in and slides out easily.

7

Yeah, the Jaguar PWM cable is a lot looser than the spikes. Are you sure that you are putting them in correctly, with Black on the right, and Red in the middle, and White on the left? Assuming the hooks are on the side i call the front. the pwm slot is towards the left.

8

I put tabs on the front of the Jaguar so that you can slide the PWM wire under the tab and then up and over into the PWM connector. That helps keep them in and also is strain relief. The connector is slightly larger to accept multiple types of 1x3 0.1" connectors.

-Scott

9

Yes, I was able to discern the tiny (-),(+), and S marks on the PWM port. Black is going to negative, red to positive, and white to signal.

I am now posting from home, our group having decided to just call it a day. After some discussion with the electrical fellas, though, we decided that the prime suspect is the 37-pin connector for various reasons.

1: There is no light display on the digital sidecar to verify that the connector works
2: We are using a light ribbon cable, which is less sturdy than the bulky ones used by computer monitors
3: In addition with the aforementioned reason, the ribbon cable was twisted a lot recently due to the (mistakenly) too small electrical box.
4: It would account for both the Jaguars not working and is less inconsistent than previous reasons (both jaguars broke, both PWM cables were frayed beyond repair, etc.)

On the morrow, we shall spend all day troubleshooting various issues. I thank you all for the help and I’ll post the aftermath of this situation.

10

*Put the robot up on blocks so you can run it in-place.

Unplug the PWM cable at the Jag and measure the voltage between signal and ground with a cheap digital voltmeter on the DC setting.

You should get a reading of roughly between 1 and 2 volts DC, and you should see the voltage change within that range when you move the throttle.

11

Just for the record: the middle (red) wire is not required for the Jag (and the same is true for Vic and Talon).

12

We use putty (think silly putty or similar) to make sure our PWM cables stay in place on the jaguars, along with those little clippies on the case.

Please let us know if the 37-pin cable does the trick!

13

It sounds like that ribbon cable. But if not, I would try sending the default code, see if the problem remains. That’s the fastest way I know to figure out if the problem is electrical or in the software. But I would bet on that cable.

14

(Skip ahead to the bold if you just want to find out what happened. Otherwise, enjoy this sordid tale of love and loss)

I have just returned from a very long session in which the entire team tried to, among other things, get the goshdarn Jaguars working again. In the end, we were successful, but the tale is one of such stupidity that it should be engraved in drool upon a failed math test.

Titanic efforts were made in every aspect to resolve the issue; we tried resetting the cRIO, completely redoing wiring that everyone knew was right, exhausting every online source even vaguely related to our issue, and so on. Historians will speak of the countless hours spent wasting time in this manner, for the real issue, unknown until the eleventh hour, was not due to a lack of competence, but a staggering amount of incompetence.

You see, on Monday everything was fine and dandy. On Tuesday and Wednesday, work was being done on the final electrical box, and I, as the team programmer, decided to take those days off. Woe to me that I should have been absent, for it was then and there that the unthinkable happened: non-electricians were assigned to re-do the wiring.

Understand, fellows, that the amount of work required by the regular electrical people had dwindled steadily once the configuration was figured out. Much effort in the first few weeks was spent trying to figure it out, as we are only a second year team, with most of the heavy lifters having graduated last year. Towards the final weeks, though, the electrical people stopped attending as often. Thus, the irregulars were assigned the all-too-important task of re-wiring everything.

So basically, the wiring guys here messed up. A lot. Fried the bajeezus out of the digital sidecar, but only to such an extent that the PWM ports didn’t work. Due to the relay ports still working, everyone was misled endlessly. We just switched out the digital sidecar and everything went back to being fine and dandy. Fortunately, they didn’t break the cRIO.

The lesson to take from this? Be steadfast and vigilant in observing whoever gets to work on the electrical components. Beware the dangers of reverse polarity, of voltage oversupply, of DSC power port going into analog breakout. Resist the time-saving demons, for they will only consume you in the long run. Heed the fragility of the electrical components.

15

Have someone experienced and competent inspect all wiring before allowing anyone to turn the main breaker on.

16

This can not be emphasized more. Before we ever apply power to the robot, we go through an exhaustive check of the entire system:

  • The electrical team checks and double checks each connection visually.
  • The electrical team checks each connection with a multimeter to ensure there are no breaks in the wires.
  • The captain of the electrical team goes over everything personally
  • The mentor of the electrical team goes over everything personally, both visually and with a multimeter.

This was learned from expensive mistakes. Like wiring a Jaguar up backwards and frying it… back when they were $100 to replace!