Jaguar Meltdown

[Al Skierkiewicz]

This is the destruct problem with conductive debris. Remember that the full power supply is available on the FETs when the power is turned on. If a chip should contact the Source and Drain pins on one FET 12 volts will cause full current to flow limited only by the action of the breaker feeding the controlller.

[Racer26]

I am currently holding in my hands a Jaguar that smoked about 90 minutes ago, and looks inside very similar to the above pictures. Upon disassembly, no obvious metal debris is present. Email already sent to LM for RMA.

EDIT: upon closer inspection of the above pictures, this one also appears to have smoked the chip on the lower right edge of the ring of FETs.

[Dad1279]

Originally Posted by Al Skierkiewicz

This is the destruct problem with conductive debris. Remember that the full power supply is available on the FETs when the power is turned on. If a chip should contact the Source and Drain pins on one FET 12 volts will cause full current to flow limited only by the action of the breaker feeding the controlller.

I think there is more going on here. We have not smoked a victor in 6 years, always very careful. Today we mounted a Jaguar, never drilled near it, and it smoked after less than 5 minutes of operation. No way there was debris in it, unless it was left over from manufacturing.

I will contact Luminary Monday, and send it in.

[Al Skierkiewicz]

Dad,
It is common for debris to vaporize in this kind of failure. So it is next to impossible to determine whether there was something or not that a foreign body was the cause. Make sure to contact luminary and return it. They are doing forensics to determine failures at this point.

[MrForbes]

Al--With the plastic piece surrounding the pins of each transistor, it looks pretty unlikely that a shaving could short the pins. It might be able to short the mounting tab of one transistor to another, or to some other part of the circuit.

[Al Skierkiewicz]

Jim,
It would look that way but on closer examination you will see that the plastic has sufficient clearance between the FETs and the plastic to allow chips to fall into that area. Also the shroud is tapered on the bottom side so most of the pin area is exposed. The shroud only touches the board at the outer edges.

[MrForbes]

It's possible, but it still looks unlikely. The openings in the top of the case are very small, the gap between the transistor and the plastic part is very small, so it seems that for a shaving to get in there, it would have to be so small that it would not be large enough to short the pins together. There are other ways for electronic components to fail, than to be shorted by foreign matter, aren't there?

That said, we haven't had any problems, we've been using 4 Jaguars since the end of December. We also haven't been doing any drilling of metal above them, since they're located in a wooden enclosure that is mounted vertically and has it's opening facing away from the rest of the robot.

[Al Skierkiewicz]

With the shape of the shroud, there are large open spaces between transistors and lots of turbulent air under the fan to push chunks around and into the FETs along with vibration running hard wheels on hard floors. When in place the underside center of the shroud is almost 1/4" above the board.

[Gdeaver]

This may not be debris. This year with the slippery wheels, a jag can go from full forward to full reverse in a very short period of time both electrically and mechanically. This could be a problem with them dealing with back EMF. Does any body know what FET they are using? While modern FETs are tough, the intrinsic diode can be the weak point and when the diode blows FETs look allot like the picture. Didn't the victors have a separate diode to deal with back EMF? Does the Jag? I've noticed that some high power drives that can handle these surges use a special type of MOV to soak up and dissipate the energy from a rapid reversal. Other speed controllers that are used for battle bots warn to allow a time period before applying a reversal. Also, the jags thermal sensing is not part of the FETs and can not react real fast. There are drives that are based on intelligent high side switches that have current and thermal protection built in to the package. It's a more difficult design when using multiple FETS per leg of the H bridge, however they are much tougher. Our team is using a joystick rate of change filter that gives a pronounce decay and ramping up when the driver changes direction rapidly. If jags have the back EMF problem, our filter should help allot.

[Alex Golec]

In keeping up with current events, we had the same problem today.

We caught the problem early on, but the Jaguar had already failed at that point. We could smell "warm plastic," and when we disassembled the Jaguar in question, the smell was confirmed: the plastic around one of the FETs had begun to melt.

There were no metal chips to be found, and I have no reason to believe any would have ended up in there. In our testbed, this Jaguar was mounted sideways on a panel above all driven components, and had been opened up fresh out of the box the night before. This was in regular use (http://www.youtube.com/watch?v=FkuxZKA6Cl0).

Hopefully we'll be able to find the source of the problem since this Jaguar is nowhere close to incinerated. I'll be in discussion with Luminary.

[Team2191]

Where should we send broken Jaguars?

[seannoseworthy]

Originally Posted by Team2191

Where should we send broken Jaguars?

You can find instructions here on how to RMA a jaguar.

We had this exact same problem last year. In the report Luminary Micro released. They say the problem is Excessive heat in MDL‐BDC MOSFET(s) with the symptoms of Excessive current draw by MDL‐BDC with the visuals of Single low‐side MOSFET melt‐down with damage to plastic ring Possible blown or fractured PCB traces. To prevent this Luminary Micro reccomends Mitigate risk for swarf (for example, avoid using ring terminals; take extreme precaution when creating metal debris), avoid positioning module for direct impact vulnerability during competition.

Two days ago we had another Jaguar fail the status led flashed red and BDC-COMM reported a VOLTAGE BUS error. Which is most likely "Voltage regulator Failure (REG): Failures of the 3.3 V or 5.0 V linear regulator could be due to one of two conditions. The most likely root cause is excessive current draw from the power rail at the interface connector. The regulator is specified to shut down internally under these conditions; however, our testing has shown that this regulator feature does not work correctly. The second possible root cause is input over-voltage. The regulators have a maximum input voltage of 15 V, which is reflected in the absolute maximum rating of the Jaguar module. Bench tests have shown that the regulators function beyond 15 V with no degradation or failures at less than 17 V. Although there is no measured data showing DC supply voltages greater than 15 V, there is speculation that the supply could exceed this under certain regenerative conditions. Efforts to replicate this failure on the bench using a motor in full-regeneration were unsuccessful. Despite lack of correlation with bench testing, this is the only reasonable explanation for certain regulator failures, and because of the range of installations, not every condition can be reasonably replicated. Luminary Micro will use a regulator with an increased voltage range and functional over-current protection in future revisions of this design."

-Sean

[Dad1279]

We bricked one tonight just updating the firmware.

We really wanted to take advantage of the features this year but the reliability is frustratingly unacceptable.

Ready to throw on the towel and dust of the reliable 5 year old vics....

[seannoseworthy]

Originally Posted by Dad1279

We bricked one tonight just updating the firmware.

We really wanted to take advantage of the features this year but the reliability is frustratingly unacceptable.

Ready to throw on the towel and dust of the reliable 5 year old vics....

If the firmware is corrupt, you may be able to fix it with the JTAG port on the bottom of the device it will be coverd by the sticker. Flash it with the newest firmware found here. You would want to use this cable or a similar cable to flash the firmware.

PM me if you want to sell me your jaguar with corrupt firmware

-Sean

[Phoenix.14]

We experienced this with two of our jaguars in close succession; one of them burned and produced smoke, and had a similar melted plastic piece when we opened it up. The other stopped working just after and produced similar behavior to that of the burned one (running at full speed nonstop, even when the PWM wasn't plugged in), but we haven't opened that one up yet.

Fortunately we still have four working jaguars, so barring another failure we'll be fine.


There is a possibility that there could be metal shavings in the jaguar as we drilled holes in the metal board they were mounted to that day, but they were mounted upside down on the bottom of the board, so I find it doubtful that they would have gone into the fan opening.