Victor 884 won't go forward

[Manoel]

Hi,

We have an 884 that came with this year's kit that only works in reverse. When we send it a nice 254, the Victor LED does turn green but the voltage is 0 V.
We discovered this problem on the ship day (sure, when else?), when our arm was only going down. We then noticed the fan had stopped and proceeded to replace the Victor.
Today I tested it again, hoping that the long months of dust gathering would have done it some good. Sure didn't.
So we pried it open and there was nothing wrong, apparently. No smell, no burn marks, nothing. We heated the MOSFETs a bit to see if it might have been a problem of cold solder joint, but it didn't help.
Any ideas?

[KenWittlief]

YIKES! how did you heat the Mosfets?

one possibility is the Victor lost its calibration. I think there is a way to restore the factory default setting, and I know it is possible to re-calibrate it (pushing the tiny recessed button while it is connected to the control system, pushing the joystick all the way forwards, then all the way back.... Its in the manual for the control system (or the Victor data sheet).

There are several Mosfets for each leg of the H bridge in the Victor. I find it hard to believe all the 'forward' mosfets could have burned out. But it is possible the controlling electronics in the Victor have failed.

Parts do fail. We use to staple all our failed parts to the wall in our Sparky area. After a few years we had quite a collection on the 'wall of death'.

[Manoel]

Originally Posted by KenWittlief

YIKES! how did you heat the Mosfets?

one possibility is the Victor lost its calibration. I think there is a way to restore the factory default setting, and I know it is possible to re-calibrate it (pushing the tiny recessed button while it is connected to the control system, pushing the joystick all the way forwards, then all the way back.... Its in the manual for the control system (or the Victor data sheet).

There are several Mosfets for each leg of the H bridge in the Victor. I find it hard to believe all the 'forward' mosfets could have burned out. But it is possible the controlling electronics in the Victor have failed.

Parts do fail. We use to staple all our failed parts to the wall in our Sparky area. After a few years we had quite a collection on the 'wall of death'.

Yes, we did calibrate the Victor to the factory default.
We have a lot of dead Victors (wrap it up in electrical tape to identify!), but this one has such a bizarre failure that we haven't labeled it dead... yet.

[KenWittlief]

If you have an oscilloscope, and could get your hands on the schematic for the Victor, it would be interesting to take it all apart and see where the 'forward' signal is dissapearing

Ive never tried to fix a burned out victor. But if you have a couple maybe some clever student could build one or two good ones from the dead pile?

[Manoel]

Originally Posted by KenWittlief

If you have an oscilloscope, and could get your hands on the schematic for the Victor, it would be interesting to take it all apart and see where the 'forward' signal is dissapearing

Ive never tried to fix a burned out victor. But if you have a couple maybe some clever student could build one or two good ones from the dead pile?

Well, if I could get my hands on the schematic I'd be a very rich man
Unfortunately we do not have a scope, and I don't know if it would be very helpful. Pretty much everything is SMD and the board not only has a mask that hides the wire traces, it also has some weird silicone covering that makes it even harder.
Since it is our only broken 884 (though the board is marked 883 revision F), I'll take a look at the (lots) of 883s we have to see if they share the same MOSFETs.
Thanks for the advice,

[Mike Betts]

Originally Posted by Manoel

...We have a lot of dead Victors...

Manoel,

Be advised that I have been mentoring Team 177 since 1995 (before IFI and the Victor existed). We use between 7 and 11 Victors per robot and some years we build two robots.

We have only failed two Victors in all that time. Both failed due to metal shavings falling into the FET heat sink assemblies (beneath the fan).

If I were you I would re-look at the "dead" Victors during the off season...

My students have reported many "failed" parts that were not failed. Bad or loose PWM connections are the primary electrical reason for these reported "failures". Other "causes" have been mechanical binding of the motor, software and cables/wires going where they shouldn't.

In my humble opinion, the Victor is a highly reliable part.

Regards,

Mike

[Manoel]

Originally Posted by Mike Betts

Manoel,

Be advised that I have been mentoring Team 177 since 1995 (before IFI and the Victor existed). We use between 7 and 11 Victors per robot and some years we build two robots.

We have only failed two Victors in all that time. Both failed due to metal shavings falling into the FET heat sink assemblies (beneath the fan).

If I were you I would re-look at the "dead" Victors during the off season...

My students have reported many "failed" parts that were not failed. Bad or loose PWM connections are the primary electrical reason for these reported "failures". Other "causes" have been mechanical binding of the motor, software and cables/wires going where they shouldn't.

In my humble opinion, the Victor is a highly reliable part.

Regards,

Mike

Mike,

Thanks for your advice, but we are very thorough when analyzing a Victor (as I pointed out in my description - I'm not sure there's anything else we can test on this one) and labeling it defective.
I agree that they are very reliable items, but we have had some tough luck before.
I remember only ONE Victor blowing up because of metal shavings. We have some old Victors (the one with stickers on the bottom with calibration procedures), some really old Victors (the ones that required external capacitors) that simply refuse to work. We have received TWO Victors from the Kit that were DOA - the cost and time to send it back turned them into fancy paper weights.
We also have had one Victor blow up straight from the box, mounted in a Lexan board with absolutely no metal shavings.
What intrigues me is this weird kind of failure, refusing to work forward. I had never seen something like that and don't recall ever seeing it on CD.
Again, I'm open to suggestions.

[Matt Leese]

I also second the statement about the fact that Victors seem to be extremely reliable. I've been working with them since they existed and haven't seen too many fail. I should say that there's a cavaet with the extremely reliable line: it was that way until this year.

Perhaps it was simply coincidence, but it seems that the quality of IFI products decreased this year. I did see a speed controller with exactly the same failure characteristics you're noticing. This was at a competition and we asked the IFI personell about it; they thought that some of the FETs were dead. There weren't any metal shavings or anything else to indicate why it had failed. I think you just have a dead speed controller and need to swap in a replacement.

Matt

[Andy Baker]

Originally Posted by Manoel

Again, I'm open to suggestions.

I second the notion of replacing FETS. We have done this and resurrected a few Victors (all 883s).

But... that seems to not be your problem. The Victors up here in North America work well. Yours give you problems. Since you are located in the southern hemisphere, maybe you should mount them upside down.

Hey, you said you were open for suggestions.



Andy B.

[Joe Johnson]

I am looking at a Victor with exactly the same problem you describe.

We were noodling around driving things we probably should be driving with a Victor (I am ashamed to say I was using it a load for a charge circuit for batteries -- Not exactly the design usage) something when pop and no more reverse.

Anyway, it is hard, but not impossible to fry an entire leg a Victor's H bridge*.

Joe J.

*Actually, it is quite easy to toast the MOSFETS in a typical H-bridge (I have a PILE of those next to the bad Victor). It is a tribute to Tony & Bob & the rest of the IFI Gang that they have wrapped these fragile jewels in such a way as to make it quite difficult for us to destroy.

[Matt Krass]

My team has a Victor with the same failure characteristics also, but we've determined the cause to be driving the motor the victor was connected to manually.

In Atlanta,, a student got metal shavings in one of our Victors for the drive, and it blew out next power up. We were left with 3 functioning and a few minutes until our next match. So we decided to go from 4-motor/2-wheel to 2-motor/2-wheel. But in our rush we only disconnected the PWMs to the corresponding Victor on the other side of the drive. After the match that Victor failed to work. We think it's because the disconnected motor was manually spun by transmission (it was geared to a working CIM) and the reverse current flow blew out the forward half of the H-bridge. It's the only logical problem we can find, and seeing as how 2 motors geared together was considerably more popular this year, combined with loose PWM cables, that could explain the apparent rise in Victor failures.

I think.

[sanddrag]

We've pushed our robot on the ground and spun the motors sometimes generating enough current and voltage to make all the lights come on and the fans spin. So, I know the electronics can tolerate a little bit of this. But a whole matches worth I guess not.

So my two questions:
Is is really bad to backdrive motors in turning them into generators in the FRC electrical system and why?
How long can you do this for without causing any damage?

[sciguy125]

Originally Posted by sanddrag

Is is really bad to backdrive motors in turning them into generators in the FRC electrical system and why?
How long can you do this for without causing any damage?

You know, I've actually been trying to figure out why backdriving the motors generates current to turn the system on. I haven't been able to figure it out though. The MOSFETs have diodes, but they're reverse biased, so any current going through them would be in the wrong direction. The other thing is that at some point, the power has to get to the RC to turn it on.

But to answer your second question, we had someone pushing it through a parking lot one time. By the time it got to it's storage location, it had turned itself on. The backup battery was maintaining the RC. It also smelled funny, like burning electronics. But, it still worked fine. I wouldn't advise doing it though.

[Joe Johnson]

Originally Posted by sciguy125

...The MOSFETs have diodes, but they're reverse biased, so any current going through them would be in the wrong direction. ...

But there are 4 of them in an H configuration. These 4 diodes end up working like a rectifier bridge; no mater which way the motors generate electricity, the diodes funnel + voltage to the +batt and the - voltage to the -batt (a.k.a. ground).

This is why the lights come on no matter which way you turn the wheels.

Joe J.

[KenWittlief]

when you spin a motor (backdrive it) it acts like a generator. We use to demo this with the globe motors and the old rotating lights - hooking them together, turning the globe motor with a wrench and the rotating light would lightup and spin

during an event when you let off the joystick, the motors back-feed, and depending on whether you have your victors set to coast or brake, the counter-EMF (generated voltage) from the motors will either be shorted through the victors to act like a brake, or it will be left open circuit and the bot will coast to a stop.

I dont think you can blow out your victors this way unless you were pushing the bot really fast, and the motors were generating a voltage above the Victors rated MAX voltage spec.