Re: how many people have had failing jaguars
 

Joe, are there any tricks or precautions to using the new power distribution panel, which may be important here?

Re: how many people have had failing jaguars
 

All I can say is I LOVE the Jaguars, they're so much more efficient than the Victors. Our sister team (team 2881, the Lady Cans) accidentally flipped the input power wires to one of their jaguars and it released all of its magic smoke, and the capacitor was about to burst, so don't plug it in backward. But that's the only problem with them that we've had on either team. They were even fine when someone left a jumper on the PWM port going to one of them (the jumper that provides servo power to the middle pin). Nothing happened. I wouldn't suggest doing that, but they were okay.

The interesting part is that when they replaced that Jaguar with a Victor temporarily (and it was programmed as a Victor, so that wasn't wrong), that one CIM out of the four they were using for drive (holonomic) got really warm, while all of the others were very cool. That hadn't happened with the Jaguar. They're much more efficient, and their power curve is nearly perfectly linear, as apposed to the Victors, which have a sqaure-root power curve. I say use them.

Re: how many people have had failing jaguars
 

One of ours just failed tonight. Worked for weeks then just stopped after cylcling the power for the programming. The fan constantly runs and when the PWM cable is plugged in it flashes a slow red. When you unplug the cable the cable or theres no code (labview) it flashes slow yellow. We had one other that came without the PWM connector but we fixed that.

Re: how many people have had failing jaguars
 

Thank you (again) to the teams that have answered my call-to-action. We had some more units returned to us last week - nearly doubling the number of units that we are able to analyze. :) We've still only received about half of the units back that we have replaced - so if you are sitting on a failed unit that we've already sent you a replacement for, please send it back so we can perform FA.

Here is the (updated) status as of Friday (Feb 13th) evening:

Since November, we have received 62 reported failures that we have immediately replaced. Here is a pareto of the complaints:
• DOA:
- Neutral LED – 6
- PWM Missing – 3
- PWM function – 6
• Reverse only – 19
• Forward only - 1
• No voltage on output – 4
• Incorrect output voltage - 4
• Stopped working – 12
• Other – 7

Of those 62 reported failures, we have received 34 units back to us – which we have analyzed. Here is a pareto of the analyses:
• Flash not programmed – 2
• Customer-induced EOS – 2
• Gate driver U6 failed – 22
• 5V regulator failed – 2
• Missing PWM connector – 3
• Other – 3

While our engineering teams continue to analyze returned units and continue to attempt to recreate failures that some of you have witnessed, I’d like to point out a piece of feedback I received from the team. Apparently, a few units have been returned back to us with some visible swarf inside the enclosure. While the Jaguar's plastic enclosure is designed to largely protect the electronics from debris, the Jaguar is not immune to the risk of swarf entering the unit. I am certain every team’s mentors have reminded each teammate at least 15 times about how to properly operate a drill (i.e. wear eye protection, never drill above a Jaguar/Victor, never carry a freshly drilled metal apparatus across your electrical without properly dusting it off, etc). However, I’d like to offer yet another risk for swarf: Jaguar screw terminals.

We recommend using spade terminals to connect to the colored screw terminals on Jaguar. We colored the screws to make it easier to identify which terminal to connect which wire (see Getting Started Guide at http://www.luminarymicro.com/jaguar). That said… at least two problems surface when one unscrews one or more of the screw terminals (to attach wires with loop terminals, for instance). First, it’s easy to mistake which screw went to which terminal – and we’ve had some honest teams tell us about mix-ups occurring upon reattaching the screws (poof!). Second, (and the reason for my mention) the screw terminals were not designed for the screws to be removed completely. If the screw is removed completely, swarf could deposit from the screw and/or the screw terminal housing into the Jaguar.

So… we have not concluded anything definitive yet about the 62 reported failures, nor am I trying to cause panic for teams that did remove the screws completely at some point. This is just something our team thought to tell everyone ASAP when we noticed swarf in some returned units.

My echoing reminder: If you have experienced a failure, PLEASE contact us at support@luminarymicro.com. When you do, PLEASE provide us as much information about your failure as possible (see my questions list here: http://www.chiefdelphi.com/forums/sh...0&postcount=57 – and send us pictures of your system, which are exceptionally helpful for us in understanding your use model). Finally, if you do receive an RMA replacement unit, PLEASE ship us the Jaguar that is giving you trouble. As you can see from above, we have still received only about half of all RMAs we have issued, and ideally, we’d like the opportunity to analyze every failed unit.

Good luck on your final days of the build season!

Regards,
Scott

Re: how many people have had failing jaguars
 

We've had 2 Jaguars fail for no apparent reason. One of them functions in only one direction (although it lights up in both directions), and the other doesn't function in either direction (although again, the light functions in both directions). They are both in relatively low load, low use scenarios. The failure in one direction occurred with probably ~2 minutes of run time. The failure in both directions had a significantly longer run time, although it certainly doesn't exceed a few hours.

We're busy finishing the robot, but we'll email you with more information once the robot is crated up.

Re: how many people have had failing jaguars
 

1/2 of ours (2) burned out during testing, quite literally, smoke and all.

Re: how many people have had failing jaguars
 

Well, yesterday we had a Jaguar on our drivetrain fail. It only works in reverse now so for the time being we've swapped it out for one of the Jaguars controlling our shooter CIM's since it only needs to move in one direction anyway. I've already sent an e-mail to LM answering the questions that were posed in utlinebacker's post. I hope that the problems are solved since I really like the Jaguars.

Re: how many people have had failing jaguars
 

We had one, but I think there were some metal shavings that got inside. So it's not product faulty, it was user error.

Re: how many people have had failing jaguars
 

so we had one burn out on use today it was on our conveyor...

I am e-mailing luminary micro now...

Re: how many people have had failing jaguars
 

192 has had 2 fail that I know of. I was there when the second one was changed out and had a student remove the cover. Inside we found small metal particles. They cover the electronics when they have to drill. We had a failure analysis meeting before starting the robot and one student mentioned the dirt on the moon rocks could be falling down on the jaguars. A plastic dust shield has been installed and the robot has run 20+ hours since then. I have not been informed of any problems since.

Re: how many people have had failing jaguars
 

To clarify more, we have had 2 jaguar failures so far. The first one, which Zach had reference, is assumed to be a user error with metal shavings, but I'm not quite sure it's just that. Upon opening it, the only bad part looked like the capacitor, so we tried to replace that, but that does not fix the problem. After powering back up with a new cap, smoke emitted from the smd chip at U6 on the silkscreen...

As for the 2nd jaguar failure, this one had the same symptom as mentioned in posts above...it would not go in reverse, forward only. Upon opening this Jaguar There was visible signs of one of the FETs overheating by the sign of melted plastic around it. Since the other jaguar was unrepairable, I had taken off one of the FETs and put it on the burnt out FET. To this day there has been no problems with that specific jaguar, and it functions correctly now.

Hopefully those clarifications can assist in diagnosing the problem.

Re: how many people have had failing jaguars
 

We haven't had any break yet, but we sometimes get a flashing red light (fault condition, perhaps current overload) and the jaguars don't work for about 3 seconds.

We have never had this happen with victors on any of our past robots, and since these are on our drive, we don't want this to happen, and we plan to swap them out with victors.

We aren't jamming the joysticks at full forwards and then suddenly full reverse.

I was wondering if anyone else has had similar issues.

Thanks!

Re: how many people have had failing jaguars
 

We have been running practice bots and driving for the full 6 week build period without a single Jaguar failure. We are planning on using Jaguars on our the drive system for the increased linearity.

Re: how many people have had failing jaguars
 

612 is actually using 8 Jaguars and no Victors. Surprisingly we have had problems communicating with the Victors through the new system so we completely stopped using them. But after reading this thread we have made sure to have spares on hand. But as far as programming for our drive system, we are enjoying the new features.

Re: how many people have had failing jaguars
 

we did unluckily before shipping but it's a lot better than at competition !