Thread Tools
Apr 28, 2014, 09:41 AM
Hit my smoke....
Hostage-46's Avatar
Thread OP
Question

In line fuse for servos - scary bench test event


So I was fiddling with my new X2 3.5 this AM on the bench. As soon as the system conneced I get a low voltage warning, hmmm, then I notice my right flap/aileron setup was not working...then came that smell, that burning plastic smell .. I DO NOT LOVE the smell of burning servo in the morning!

That was the first time I witnessed the awsome power of a wide open failed servo. It melted the servo plug of the switch going into the reciever, the servo is toast. It even chared a bit of the wing skin, melted wires etc. For what it's worth, the zepsus switch worked fine, I was able to turn it off, but I'll replace it just the same.

It was brand new servo, this was a routine bench test. If this had happened at any altitude there is no way I'd have gotten down in time to save the ship before the battery (1450 Life) drained at max C melting god knows what before fizzling out.

Got me to thinking, has anyone ever put fuses in line for something like this?
Sign up now
to remove ads between posts
Apr 28, 2014, 12:46 PM
NorCal Electric Soaring (NES)
SoaringDude's Avatar
Quote:
Originally Posted by Hostage-46
Got me to thinking, has anyone ever put fuses in line for something like this?
Glad it happened on the bench. You just described the classical failure for which inline regulators were designed. MKS introduced 2 different regulators (1.5A, 2.0A) designed to limit current from a failed servo. MKS erroneously calls them SBECs but they are just linear regulators that will clamp current to the rated value in the event of a "short circuit" type of failure.

I don't use these but if I did I would cut off the connectors and hard wire them in since extra servo connectors just increase the chance of an intermittent connection failure.

The other thing to keep in mind is these inline regulators will not "save" your battery from being discharged but it will slow the rate of discharge down.

The problem with using a fuse for this is that under high-G turns or zooms servo currents can easily reach 8A or more so you would need to use a slow blow type of fuse. I suspect you would find that even slow blow fuses would blow out on quite a few types of extended high-G maneuvers but having never tried it you may be able to find just the right fuse that would give good protection.

Chris B.
Apr 28, 2014, 01:16 PM
Registered User
dinamich's Avatar
Look at https://www.rcgroups.com/forums/show....php?t=1533260
Apr 28, 2014, 02:39 PM
Hit my smoke....
Hostage-46's Avatar
Thread OP
Quote:
Originally Posted by SoaringDude
Glad it happened on the bench. You just described the classical failure for which inline regulators were designed. MKS introduced 2 different regulators (1.5A, 2.0A) designed to limit current from a failed servo. MKS erroneously calls them SBECs but they are just linear regulators that will clamp current to the rated value in the event of a "short circuit" type of failure.

I don't use these but if I did I would cut off the connectors and hard wire them in since extra servo connectors just increase the chance of an intermittent connection failure.

The other thing to keep in mind is these inline regulators will not "save" your battery from being discharged but it will slow the rate of discharge down.

The problem with using a fuse for this is that under high-G turns or zooms servo currents can easily reach 8A or more so you would need to use a slow blow type of fuse. I suspect you would find that even slow blow fuses would blow out on quite a few types of extended high-G maneuvers but having never tried it you may be able to find just the right fuse that would give good protection.

Chris B.
I was discussing this with a buddy in the shop and we concluded that a slow blow fuse makes the most sense. I terms of high G manuevers, other than than launch mode, it's not an issue as I fly TD and F3J.
Apr 28, 2014, 03:30 PM
NorCal Electric Soaring (NES)
SoaringDude's Avatar
One other thing: not sure what battery voltage you were applying to the failed servo but at least with MKS you have to be careful to not run with much more that 5.5v. The more you exceed the rated voltage (i.e. an unregulated 2S LiFe pack) the greater chance for failure. The MKS regulators put out 5.0v FWIW.

Chris B.
Apr 28, 2014, 05:37 PM
Hit my smoke....
Hostage-46's Avatar
Thread OP
This was an ATX 761. I've been flying these for years with a Hyperion LifeO generally at 6.6 volts.

I've send the servo in, something was most defintely wrong with this thing. Again glad it happened on the bench
Apr 28, 2014, 05:46 PM
Mark LSF # 3792
I always "run" mine in for about an hour before use, 'course if they fail afterwards...
Apr 29, 2014, 12:22 AM
Flying What?!?
fnnwizard's Avatar
I've intentionally tried to kill many servos over the years and actually had some success!

Predominately, it comes down to 2 types of component failures. The motor or the FET controlling the motor. If the motor fails, it's usually by way ofunderperforming or completely quiting due to severe arcing/pitting/oxidation/carbon buildup of the brush/commutator. This type of failure is isolated from the rest of the electronics and often not disastrous to the airframe.

The other failure mode is typically the FET (due to various component voltage issues/defects) and most of the time they fail as a direct short circuit = torching the FET (more like a micro explosion) and whatever is near. The result is that enough current goes through to melt the servo wire (and the battery lead wire, if small gauge was used there) creating an even higher current short. This usually causes complete radio disconnect not to mention a possible fire if power was not removed quickly.

If one really wants to use an inline fuse to protect for this type of failure, for our typical servos, a 2A slow blow fuse at each servo to rx/power connection should work. The wiring we use are plenty adequate to handle >2A for long periods.

However, it might be better (space constraint and connection points) just to strip some of the strands from the power wires of each servo/ servo extension near the rx end. The left over strands then essentially act like a fuse. It should be simple to test the wire to get it to burn through at say, 3A without fear of melting the insulation. If it does ever burn through, then it can be repaired with an inline fuse ( which either way acts as the current bottleneck). The best option might be to get high-temp insulation wire as that might allow the FET to burnout quickly without melting wires to cause a longer term short.

Soarmark's recommendation to cycle the servos (as fast as required of them) for some time before install is definitely a requisite.

@ Chris, since you are the electronics expert here, would you have an idea what would happen if those inline regulators (specifically those ic regs,) tried to regulate a direct short? I'm thinking probably not much since those can dissipate heat more readily until the servo FET burns through the short? Are there any instances of those failing short also?

With the more potent batteries we are using today, the fear of the plane causing a fire afterwards if involve in any incident is always at the back of my mind. So, my bottom line is always being extra careful of control/power components, but knowing the risk is never zero.
Last edited by fnnwizard; Apr 29, 2014 at 12:30 AM.
Apr 29, 2014, 10:53 AM
NorCal Electric Soaring (NES)
SoaringDude's Avatar
Tuan, you bring up some really good points. Regarding servo failure modes while I've done my own post mortem take-aparts on several servos that failed for me in the last few years it would be great to have reps from the popular servo companies tell us what the % breakdown is for failure modes. They obviously have all the numbers...

While talking with electronics wiz Mike Wilson 2 years ago about digital servo failure modes he pointed out that most manufacturers are probably using the same electronics. From the checking I did on several brands it turned out he was correct: they used the same 64611 digital servo control chip and a pair of dual FET driver chips. From my limited experience servo "short" failures came down to the motor winding failure due to overstress aided by too much battery voltage, servo binding, or motor defects. The FET chips may blow but my guess is that it is usually brought on by the motor windings failing. MKS and JR, feel free to jump in and add your .02

Assuming your servo supply voltage is safe I like the fuse protection idea but I would do tests to confirm the slow blow fuse thing under all the high-G moves you typically make in F3B, J, etc. It should work. You could rely on the fusing current of several wire strands as you suggest but that I like the fuse idea better.

Inline regulators are nice since they in effect act like a fast-acting auto-resetting limiter. The linear regulator chips sense the current level and then electrically throttle the current down to keep the regulator in its "safe" region. So an inline reg would even give you very fast temporary current limiting for occasional servo binding. The more I think about it the more I like the idea of adding the inline regs to flaps. Haven't done it yet.

And...since these reg chips are so small it would be peachy for the servo mfgrs themselves to add them to the servo electronics board so we wouldn't have to do it externally.

Your last point: while the inline regulators are self-limiting and pretty darned reliable they are extra electronics that could fail themselves. That's why I wouldn't be a fan of adding them blindly to every servo just for short circuit protection, just to the ones most likely to get overstressed and fail.
Apr 29, 2014, 11:06 PM
Hit my smoke....
Hostage-46's Avatar
Thread OP
I think 2A slow blow makes sense. I'll ping servo city and see if they have anything. otherwise, it seems like one could fashon an inline fuse that has servo plugs on either end. I have a few EE buddies, I'll report back with what we come up with. Something simple that can go in-line.

FWIW I put the Hyperion 1450 life on the charger, it took 550 or so MAH to recharge. I've never discharged more than 350 in a multi hour session. The battery was on for maybe 30 seconds. I'm pretty sure this would have resulted in a fire had I not been on the bench. I also deduce the battery would have lasted no more than one minute from failure to exhaustion.

If this would have happened in flight it would have likely been a lost model.

Mark, looks like a great day to fly on Friday, I'll give you the battery to test, might as well since I sent in the receiver in for a look see and replaced the zepsus
Apr 29, 2014, 11:19 PM
NorCal Electric Soaring (NES)
SoaringDude's Avatar
Dan,

I did some research earlier today and located a nice line of micro fuses that would work great and are very small. I went ahead and posted what I found on my RCG blog. If you think it's appropriate I could present that info here too. I do plan to do some testing with them when I get some soon.

Chris
Apr 29, 2014, 11:36 PM
Hit my smoke....
Hostage-46's Avatar
Thread OP
Thanks Chris, very helpful, I'll run with this....
Apr 30, 2014, 07:50 AM
Mark LSF # 3792
Dan, I have to take my mom for some tests tomorrow, but should be free Friday. It might be a good idea to test the battery also. BTW where are you getting your Hyperions?

Also, I was making up a 6S AA pack for my SD-10G and shorted the lead wires by accident. Amazing how hot the wires got (melted the insulation in seconds) with new uncharged rechargeable AA's!
Apr 30, 2014, 10:02 AM
Hit my smoke....
Hostage-46's Avatar
Thread OP
Yessir I'd like you to test it for good measure. I was getting them from RCLipos but the owner had some health issues. Last one came from AllErc, I'll look up the proper site.
Apr 30, 2014, 10:49 AM
Hit my smoke....
Hostage-46's Avatar
Thread OP

next steps


Chris,

To provide a bit more data, my EE flying buddy and I will conduct a proper test of the output of a catostrophic closed circuit scenerio. This would emulate the worst case scenerio for the Hyperion lifeo. I'll bring in some other pack types for fun.

It will be useful to know how much and how fast we're discharging to properly sized the fuse.

We're also planning on an some in flight logging to accurately measure amp draw during "normal" conditions. Need to source the correct logger, maybe you have a suggestion?

More to follow

Dan in Dallas


Quick Reply
Message:

Thread Tools