Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Yes..we don't know yet. But it is possible that a single pulse from the JAG that mimics the Victor would suffice to break the static friction and drop the friction to a dynamic value. This is what I wanted the testers to try.

If you held a JAG at 100% pwm for duty/120 sec then the current waveform of the JAG and Victor at the same duty would look identical for a single 120 hz cycle. So whatever the Victor is doing to rattle the lock loose can be duplicated by smart control of the JAG. This JAG breakout waveform could precede any manual command for regular control. Hence a lower duty smooth command to the JAG could then be used in theory since the static lock would be broken.

If that didn't work, then conceivably the JAG command could be Victor shaped continuously. The only limitation is the 1khz JAG command cycle. This is doable as Ether hinted if you insist on using a JAG, but why do it if the Victor works.

By the way, I was able to duplicate the waveforms posted by diviney with my recycled Hbridge waveform simulator that we used in the JAG linearity vs frequency discussion last year. I put in the 2010 Denso , tweaked the i_free to be 1.4 amps (spec 1.8 amps) and upped the motor inductance to 500 uh (was 200uh ) to get a pretty good match of the 20% Victor response. See Fig 4 in Diviney post http://www.chiefdelphi.com/forums/at...1&d=1268192689 and the attached Fig 4. What is unknown is the friction drag...so I assumed that static and dynamic friction currents were equal and added some inertia load.

Using these parameters with the attached denso hbridge vi I ran a 100% JAG pulse for 1/240 sec. ( See attached Fig A. ) and a 50% duty Victor (Fig B). As expected, the current plots match very close. So, if a single pulse Victor can shake it off the lock... then why not the JAG?

PS. Sorry the figure titles aren't showing... just mouse over them to get titles

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Chris,
Interesting premise... The only difference would seem to be the refresh pulse for the bootstrap cap in Jaguar. The manual does not list the refresh rate for the charge circuitry. Only this...

"This state only occurs when the motor is running full-forward or full-reverse. The MDL-BDC software intermittently switches to the low-side MOSFETs for a short duration to replenish the bootstrap capacitor. The short duration has no impact on motor speed."

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

This thread has grown cold but, I'm reviving it one more time to update our summer experience. As mentioned, we did have random and intermittent wheel lock up at a off season competition with the locking pins removed. The Denso motors run much cooler with out locking pins. We replaced the jags with Victors before IRI. Since we went to the Victors we have not had a Denso motor problem. The only guess to the problem with Jags and window motors is that the high switching frequency and the motor PTC cause a self heating of the PTC. After this experience I would strongly suggest that any team that uses window motors should only use Victors. We did not test with locking pins in and Victors but, the performance with out locking pins is much better.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

We had a similar experience with a black Jaguar driving a flywheel.
The Jag couldn't handle full current after it heated up a little without faulting.

The problem could be avoided by ramping up slowly, but it wasn't a sure thing. Swapped in a Victor and the issue went away, if a tiny bit noisier as it maintained a set tach reading. The Victor didn't show any stress via excess heat.

I suppose the lesson is to test the Jaguar rigorously in the application ahead of competition if possible.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

What kind of motor was the Jag driving ?

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

A CIM. It was moving too, not just stalled. We added a slow boost to make sure the flywheel was moving before allowing more power, so we were seeing Jag faults at half power around 800 rpm after awhile.
It had a really, really smooth maintain speed though...

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Mark,
800 RPM on the flywheel or the motor? My suspicion is the motor was heavily loaded and controller current on the Jag was approaching it's set point for current overload. Was this for the steering via moving flywheel of 2009? Of the designs I observed, some incredible forces were built up during steering moves.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

It's 800 rpm on the CIM.
This is a kinetic energy kicker from this year. It's geared to be capable of 5000 rpm, but we restrict it and normal operating speed is about 1600rpm from the far zone.
The magnetic fields collapse quickly when it kicks, but we didn't measure a dangerous pulse back into the Jag. Power to the CIM is stopped when it kicks.
The CIM didn't heat up more than slightly warm to the touch.
A combination of the back EMF and spinning up the flywheel is obviously stressful enough on the Jag to warm it up over time and put it close to the over-current. It didn't have any trouble maintaining speed, just on the ramp up.
I am surprised that over current was reached at half speed after a lot of stressful speed changes.
The fault point gradually dropped in testing as time and stress went on.

We had it playing golf at an SBPLI fundraising outing on Tuesday and it was operating from 300 rpm to 450 rpm without noticeable issues. It can repeatedly putt a golfball back into the box it came out of at 40 feet.


The 2009 flywheel for turning you're thinking of was one of our local teams (263 Sachem), and was powered by a victor that overheated and burst into flames before they redesigned the code to ramp up and not try to go full bore all the time. That flywheel was 22 lbs and they were spinning it a bit faster.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Unfortunately, I have never seen a spec on trip current on the Jaguar. My suspicion is somewhere around 60-80 amps sustained. I have no test data on this, just some observation. Of course, back EMF could be playing havoc with the low voltage sense.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

According to the documentation, the jaguar is designed for 40 amps continuous current. They can hold 60 amps for 2 seconds, and a 100 amp start current.
As found in the MDL-BDC24 Data sheet.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

FWIW, the CIM draws 68 amps at max power, according to its engineering drawing.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

If you are quoting this FAQ...
The Jaguar (MDL‐BDC) modules have a self protection feature that does not allow excessive current to be drawn. This feature protects both the motor and the motor controller from damage. The Jaguar (MDL‐BDC) module normally provides up to 40 A of continuous current to a heavily loaded motor. However, it is capable of providing much higher currents, but for shorter periods of time. Jaguar provides 60 A for up to two seconds and provides 100 A for approximately 0.2 s.

I read this as the current protect mode waits 2 sec at 60 amps and 0.2 sec at 100 amps. The stall current on a CIM is 129 amps. If this is an accurate statement, then Mark's experience could have been right around the 60 amp trip point. Remember that fault will prevent output for 3-4 seconds while flashing the fault LED.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Personally, I've experienced the jaguars holding the stall current of CIM's for a little under a second, easily about 100-110 amps (we were running 2 CIM's on 2 jags, so, bit of a voltage drop and thus lower curent). The .2 seconds wasn't even close to what I was seeing before the flashing red lights came on.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

We had a similar problem at Championships. We built a swerve drive using denso motors for steering and 2 cims per drive transmissions. Our students noted that we would lose partial power when reversing and it would continue until the stick was centered. We put our machine on blocks and noted the blinking Jaguars. We then replaced the Jaguars but to no avail. Finally we got a hold of the TI guy (sorry can't remember his name) and he basically told us that the Jags were faulting when we changed motor directions too quickly (This was happening even when the machine was on blocks). Said the cutoff was around 40 amps. No. our drivetrain does not have excessive friction in it everything is properly supported with ball bearing and properly tensioned and it was geared for intermediate speed. With that said our students swapped out the Jaguars at home to victors after verifying that our machine was still doing this and poof! Problem solved. It would seem that the Jaguars have some really neat features, but for some things, they are just too smart for their own good (or they are just too smart for us :) ). I think it would really help if they could raise that trip current for the fault protection. We didn't have any issues in 2009 with the Jaguar going into a fault condition, probably because of the Regolith playing surface.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

The internal breaker in the Jaguars is really annoying in my experience. It tends to trip faster than the PDB's breaker and stays locked off for 3 seconds continuously. A lot of teams rely on drawing more than 40 amps current for just a few seconds for various scenarios involving turning, pushing, etc. so running Victors in a drivetrain automatically gives you performance gains simply because you get more leeway with temporary high current scenarios.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Just to be clear, the Jaguar does not have an internal breaker. The internal micro is sensing the current and making decisions accordingly. The micro also decides to shut down output for the 3-4 seconds it is flashing the fault LED. This is likely to give the FETs a chance to shed some heat.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Are you saying you had 8 CIMs on your vehicle?

(2 CIMs/trans * 1 trans/wheel * 4 wheels/vehicle = 8 CIMs)

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

I believe 368 built a coaxial swerve drive with two transmissions for four wheels.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Just as a suggestion, if you are using CAN (and thus Jaguars) then the fault timeout can be reduced to 0.5 seconds.

Obviously, if your mechanism is pulling so much current that this happens fairly often a redesign could be a good idea (as reducing the timeout leaves the jaguar with less time to cool its FETs).

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Justin,
While this might be a good idea in some designs, I would not suggest it for FRC robots. If a fault is occurring, you may not know it without also programming to watch for fault status. Should someone be trying to diagnose a problem, they would not have the added status of the fault light and timeout to see where the problem is occurring.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

I've been following this thread for some time.. However, I have yet to read anything that definitely shows that Denso Window Motors and Jaguars don't work well together.

In 2007 we had some issues using Window Motors on our lift. As such for this season we are hoping to use the advanced monitoring of the Jaguars and CAN to watch voltage draw etc... so we don't burn out, or over tax, or even gain an indication of a failing motor. Of course our lift will be different, and hopefully significantly better than 2007's.

However, if it is very clear that Jaguars won't meet our needs then we'll switch to victors.

Anyone?

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Michael,
It is well documented that the window motors react with the Jaguar. Several teams have documented the issue but I believe it was GDeaver and John Novak who really tied it down. It would appear to be an issue with the anti-back drive assembly in the window motor. This year's rules allow the removal of that assembly for just this reason.
<R47> Motors and servos used on the ROBOT shall not be modified in any way, except as follows:

C. The locking pins on the window motors may be removed.

Thank you to those that made us aware of the problem. I apologize if I left a contributor's name out.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

is there a procedure for removing this pin?

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Thanks Al, GDeaver, John Novak, and everyone else that has contributed to this.

Given the definitive "issues" we'll be using Victors, as the anti-backdrive is a feature we need in our design.

Again,
Thanks everyone

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

I have not performed this operation. John Novak gave me the impression it was pretty easy once you have the motor apart. I am guessing they are very similar to the pins in the drill motor transmissions we used in the early years or the NBD transmissions.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

The jag denso issue goes beyond the locking pins. There must be a problem between the Jags high frequency PWM and the PTC in the Denso motor. We had problems with premature shut down with both back and tan jags with the locking pins removed. The problem was not as frequent with the locking pins removed. We had no problems after we went to victors. this is a link to our website and a page that shows some pictures of the locking pins.
http://wiki.team1640.com/index.php?t...r_Locking_Pins
After we changed to victors our pivot steering motors work wonderful. Don't think FIRST will go for bypassing the PTC. We did not try this.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

I have to say our experience with Denso motors has been very predictable. Every time we used them with the locking pins in place they exibited the same problem with either Victors or Jaguars. Our most recent application was independantly steering wheel modules. When they encountered a "spongy " load such as rubber tread on carpet or a soft plastic stop, the energy absorbed would backdrive the motor and set the locking pins. The only way to unlock the motor was to drive the pins back out of the lock condition. This was impractical with error based steering algorithms. As soon as the pins were discovered and removed, The motors performed absolutely predictably with Jaguars. This means that if the motor was stalled due to a restriction, the internal thermistor would heat up in a few seconds and the motor torque would be reduced to the point where it would no longer turn. as soon as the thermistor cooled down the full motor output returned. we went through three competitions with this combination and the only time this happened was when a chain came off and jammed one. Under normal conditions the motors never faltered or even warmed up after 20 minutes of constant driving. The same held true on our backup robot with gray Jaguars. CAN was used in volage mode throughout the season.
Thanks to GDeaver for the pictures at the link above of the motor disassembly and pin removal. This is absolutely essential for reliable operation with varying loads. Beware that when the pins are removed the gearmotor will backdrive very easily. It will not support a load although a motor this small probably shouldn't be driving an arm anyway.
I will be glad to elaborate on the pin removal or anything else that is unclear. Just keep in mind the motor power and efficiency of a worm drive when applying these.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Sorry to bring this thread to life again, but... :)

We have a dual window motor arm joint this year. Bench testing was great, we removed the locking pins and had no problems with lock ups. Now that is in the bot, the arm is locking on the way down. If we bump it up, sometimes we can get it to free up and go down. I am not too smart on the inner workings of the worm drive and such. We are at crunch time and need to figure out if this is an internal lockup, or if the speed controller is playing in to this. Drum roll please, we are using the new black jaguars with it.

I don't think we have any more victors available to try, but after reading this tonight I will do another search tomorrow to see if we can dig up 2.

Is it still the general consensus of this thread that the jags should be swapped out for victors when controlling a window motor under load?
Thanks

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

iam,
You might try adding some counter balance to the arm joint to help prevent jamming the worm drive in the window motors. It can take the form of surgical tubing or springs. If you have removed the pins, the only thing left is the worm drive.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

While I'm no electronics expert, I have to say that I can't disagree with the statement that a problem exists specifically while using a window motor powered by a jaguar that doesn't exist while using a victor speed controller. Last evening our arm (powered by two window motors) exhibited the exact issue being referenced in the thread. We switched to Victors from the Jaguars and the problem disappeared completely.

Should we switch to spikes?

Thanks, Charles Baxter

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Charles,
Your mileage may vary but several teams had reported differences last year in window motor operation between the two controller types. John and Gary seem to have researched this very thoroughly at the end of last season. I can see that the mass of the pins could be effected by the high switching frequency of the Jags in certain applications. However, as Gary points out, there is a PTC thermal device that may also have a significant effect. When one rules out everything else, it has to come down to motors themselves. I should have pointed out above that there are differences in the left hand vs. right hand motors. If used in opposite pairs, there may be binding in the two transmissions with these differences. Your design may have a more pronounced effect than your neighbor's robot.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

With our Arm, we are using 2 Window Motors (with Victor's) through a 4.8:1 Gear reduction to a arm that weighs about 15 lbs (I think) and we are delivering about 61 ft-lbs of torque with 2 rubber bands on the back. We have measured the current going to each motor during normal operation and it is around 11 Amps each at full load I believe. According to the motor specifications, the Denso motors will stall at 18.6 Amps.

This may be within your acceptable range, but I would be wary of controlling an arm with no speed control. Unless you are really good with keeping it counter-balanced I think it would be very jerky in operation. We can run our Arm for about 25 minutes before the window motors shut down to temperature. When our fans are running on the window motors (normally) we have ran the robot for 8 hours a few times during driver training and the motors are cold, with no problems at all.

EDIT: We did not remove the locking pins, as we wanted the arm to hold its position without having the motor hold it there.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

That is only on one motor, the opposite type is speced at 21.5 amps stall.

Re: Denso Window Motors, Warning - Don't use with Jaguars !
 

Indeed.....

We are using two Densos through a CIMple box, driving a chain onto the pinned fulcrum of our arm. Previously, we ran them with the new MDL-BDC24s. We had problems with the thermo-resistors cutting out and the arm wouldn't move.

Today, we swapped the Jags with Victors, and they run like a fine watch, or at least as good as you can get with unavoidable backdrive....

Before anyone asks, the code was correctly written both before and after.

Thanks all!