RC-Monster Forums

RC-Monster Forums (https://www.rc-monster.com/forum/index.php)
-   Brushless (https://www.rc-monster.com/forum/forumdisplay.php?f=12)
-   -   UBEC Failure (https://www.rc-monster.com/forum/showthread.php?t=22488)

BrianG 07.24.2009 07:12 PM

UBEC Failure
 
I was out playing at the track yesterday with my MBX5/Corr truck, and things were going pretty good... until I landed a jump really hard upside down. Well, the zip tie holding the MMM broke causing one of the motor leads to come undone. The Koolflight BEC apparently didn't like something about that and decided to blow. Unfortunately, it also took out my Futaba S9155 steering servo. And my Spektrum receiver won't power up anymore either. Poo.

When I got home last night, I opened the servo; yup, it's gone. About the only thing it is good for is having an extra servo case.

And hooking up the UBEC to my PS just made it get VERY hot VERY fast while outputting full input voltage.

Expecting the worst, I then opened up the Spektrum, but curiously did not see any visible damage. Now, that doesn't necessarily mean anything, but I started dissecting the unit further anyway. There is what looks like a regulator before the actual Spektrum circuitry, so maybe it's not dead after all? A little continuity test between all the servo pins showed that the ground trace was blown (couldn't see it as that trace is apparently sandwiched). So, I soldered a resistor lead as a bridge across the all the - pins, and for good measure, another across all the + pins (on the solder side of the PCB). Hooked her up and she lives! Yay! That saved me ~$80. So, it looks like the regulator saved the Spektrum circuits from overvoltage after all.

Just thought you guys might want to know this if it happens to you...

suicideneil 07.24.2009 07:18 PM

And thats why you're so respected- incredible electronics wizardry and ingenuity :yes:

fletchoman 07.24.2009 07:39 PM

Quote:

Originally Posted by BrianG (Post 307254)
And hooking up the UBEC to my PS just made it get VERY hot VERY fast while outputting full input voltage.

When I was looking for a UBEC, Snellemin recommended using a BEC that always failed open to avoid the condition you just described...he recommended Dimension Engineering.

http://www.dimensionengineering.com/contact.htm

BrianG 07.24.2009 07:59 PM

Are you sure the DE does this? I don't remember seeing any other device on those BECs that would do as you describe. It would be a nice feature though.

fletchoman 07.24.2009 11:41 PM

Quote:

Originally Posted by BrianG (Post 307262)
Are you sure the DE does this? I don't remember seeing any other device on those BECs that would do as you describe. It would be a nice feature though.

I was specifically warned about the failure mode you described mode by Snellemin...you don't want a BEC that fails by passing full voltage because you will likely lose all electronics...but I will let Snell speak for himself...I bought DE based on his recommendation.

BrianG 07.24.2009 11:54 PM

As far as I know, ALL switching BECs can fail in this way. Unless DE put some kind of catastrophic circuit in that measures the output and breaks the connection if found, it too can do the same thing. However, DE makes good stuff, so even if the circuit is basically the same as any other BEC, the choice of support components, layout, etc can have a dramatic impact on reliability. And the Koolflight one I was using was not a cheapie; this is the first one I've had that failed. I just think that when one of the motor wires came loose, some very high back-EMF voltage was generated and overvolted the input of the BEC.

fletchoman 07.25.2009 12:03 AM

Quote:

Originally Posted by BrianG (Post 307297)
As far as I know, ALL switching BECs can fail in this way.

Yeah...I know that any device cannot 100% guarantee no failures (shorts through components)...but I was warned by Snell that design matters here...DE was his recommendation...I am interested to hear his comments on the subject...

BrianG 07.25.2009 12:16 AM

I'm interested as well. I have several BEC styles here: Chinese cheapies, Koolflight, Castle Creations, DimensionEngineering, & Western Robitics. Koolflight, DE, and WR all performed equally well (temps, output, heat, noise levels, etc). If DE has additional protection, it would be nice to know this. I didn't see anything at the DE site that said either way, and you would think that would be a good selling point.

jasje 07.25.2009 11:10 AM

Slightly offtopic BrianG, but I was checking out your mbx5 conversion and I happen to have a spare mbx5 here too, that I might convert.
I see you use the rcmonster motor mount, for which model is this?
Secondly, why did you make a custom alu top plate, where the normal front and rear brace not enough? Or didn't they fit due to the mods?

Bondonutz 07.25.2009 11:29 AM

Thanks to your curiosity and enginuety you were able to save your RX, glad to hear. Send your servo back to Futaba, they may just surprise you and replace it w/no Q's asked. Stranger things have happened ? A buddy of mine totally smoked a Fasst RX due to a dopey noob mistake and they replaced it,no probs.

lutach 07.25.2009 12:14 PM

This is why I recommend Rx packs.

BrianG 07.25.2009 12:50 PM

Jasje: I'mm 99% sure I used this mount: http://www.rc-monster.com/proddetail...=RCMmugenmount. And I create a top plate in many of my conversions to really stiffen up the chassis. I find that even with the front/rear supports, the chassis tends to "taco" over time. This prevent it. However, it probably isn't good for serious racers as you generall want chassis flex for proper handling, but I value durability over flexibility.

Bondo: I emailed Futaba and explained what happened. The failure was caused by collateral damage, so it's sort of my fault and I don't expect them to cover it for free (but it sure would be nice, lol). I even asked if I could just purchase the control board and put it in myself. I doubt they'll go for this though. We'll see I guess.

Lutach: Yeah, an rx pack would have prevented this from happening, but I don't accept that. Threy are larger, heavier, and just one extra thing to charge. A UBEC is a much lighter/elegant solution. All that is needed IMO, is a simple addition of a little bit more circuitry to make them 100% reliable (or at least not take out everything when they fail).

magman 07.25.2009 12:56 PM

Would an additional cap on a ubec help prevent this type of failure or would a more intricate piece of circutry be in order

BrianG 07.25.2009 01:07 PM

I don't think a cap would be enough. If the BEC failed, the cap would just charge up to whatever voltage it was outputting. No, I think an added circuit that should shunt anything over 6.5v directly to ground would be better. This would blow an on board fuse or something like that. Yeah, the device would still be toast, but would save whatever expensive circuitry it was attached to.

BP-Revo 07.25.2009 02:55 PM

Wow...I don't often hear of UBEC failure (first ever, I think).

Brain, was there maybe a surge or maybe a short when the motor lead popped off? Only think I can think of besides a physical type failure on the BEC's part.

Brian, I wonder rather than integrate the failsafe into the BEC itself, maybe develop one which plugs inline between the BEC and RX. This way it will work with any BEC.

This way you can also make it so that the thing blows a replaceable fuse or something so you can replace the fuse and not the unit.


All times are GMT -4. The time now is 12:57 PM.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2025, vBulletin Solutions Inc.