Thread Tools
This thread is privately moderated by SoaringDude, who may elect to delete unwanted replies.
May 09, 2014, 12:17 PM
NorCal Electric Soaring (NES)
SoaringDude's Avatar
Thread OP
Discussion

Resettable fuse isolation for servos


This is a follow-up to my blog post "Protective fuses for your servos" immediately preceding this one. While the micro fuses in that post would do a good job protecting your plane from a potentially catastrophic servo short circuit, using a resettable "polyfuse" is even better since it never needs to be replaced if it ever trips. I owe Mike Wilson the credit for suggesting this alternative.

The resettable fuse part number: TE Connectivity p/n RUEF110 1.1A 30V, Mouser p/n 650-RUEF110, $0.32/ea in qty 10. These devices are low cost and small (14x7mm) "radial" package devices that look like capacitors and should be soldered directly into your harness in front of each servo you want to isolate for short circuit faults (described later in this post).

How they work

If you're electronically curious you can read how these resettable fuses work here. The basic idea: current that exceeds the rated trip current causes the polyfuse to heat up and increase its resistance thereby reducing the current to safe levels until either the short circuit terminates or until the fuse is reset via a power cycle.

Bench tests

To test this device I used a Hyperion 2S LiFe 1450mAh battery, a 68 ohm 1 watt resistor to simulate a 100mA average servo load, and a 1ohm current shunt that allowed a digital storage oscilloscope to display the current flowing through the servo load during short circuit conditions. Two types of tests were performed: a continuous servo short and a pulsed short circuit to simulate repeated servo binding that could cause momentary high current spikes.

Continuous short test. As shown in attached Figure 1, at point "1" the battery is switched on, at "2" a short circuit is applied which causes the current to jump to over 4 amps (limited by the battery's internal resistance, the test wiring, and the current shunt), at "3" after 2 seconds the polyfuse has heated up to the point where it starts reducing the current, and at "4" even though the short circuit is still being applied the current has been reduced to a safe level. In a plane this means that particular servo would be dead but the rest of your plane would be flyable and most importantly, with only minimal current flowing through the shorted servo.

Pulsed short circuit test. As shown in attached Figure 2 three 1 second short circuit pulses are applied only separated by a little more than a second of off time. Since the polyfuses trip based on accumulated heat, in this particular test there is not enough time between pulses for the fuse to cool so on the 3rd pulse the fuse trips. This is how the fuse might work if you had a servo binding problem that was happening often: if the duration of the effective short is too long, and/or if the cooling time between binding events is too short the fuse may "soft-trip" (explained below). Also note that if the same test is run with 3 seconds between 1 second short circuit pulses this polyfuse never trips.

5-29-14 Update: To answer a question from Mike Wilson I measured the pre-trip and post-trip resistance of my batch of polyfuses. Results: the batch of 15 untripped fuses measured 80-85 milliohms while the 5 tripped ones measured 105-115 milliohms. So both cases were within range per the datasheet: Rmax untripped <= 100mohm, Rmax tripped <= 170 mohm.

How to reset a tripped polyfuse

In the case of a continuous short circuit, once tripped the polyfuse will stay "latched" at its holding current state until you either remove the short circuit or shut off power. When you power up again if the short is no longer there, and if the fuse has cooled enough, normal operation will resume.

In the case of a pulsed short circuit where there is no continuous short being applied, even if the polyfuse "soft trips," if enough cooling time elapses the fuse will reset by itself. From empirical tests with this particular fuse device and a 2S LiFe battery it can take 15-20 seconds for a tripped fuse in 75 degree ambient air to cool enough to allow normal operation again. This means if you are flying with a binding servo, and if you notice the servo goes dead, after a cooling period it will come back to life until the binding trips it again. This is a nice indicator of a binding problem (as opposed to no overt symptom until your battery runs out of juice in the air).

How to install radial package polyfuses

Cut the +V servo wiring lead, strip both ends of the cut wire, wrap the stripped leads around the fuse leads and solder, and cut off the excess leads. I would also use a small amount of Liquid Tape (available at Harbor Freight) to insulate the exposed solder joints BUT I would not cover any part of the polyfuse body with tape or anything else. The reason is this: if you cover the fuse it will speed up the heat buildup in the fuse element and cause trips to happen faster than is rated. This would also extend the cool-down time before a tripped fuse is reset.

I intend to retrofit my planes with these polyfuses soon. For all new planes I'll be making my own harnesses and directly soldering these fuses near each of the servo connectors in the harness.

Thanks again to Mike Wilson for the idea and component selection. Questions and feedback welcome.

Chris B.
Last edited by SoaringDude; May 29, 2014 at 04:35 PM. Reason: Added pre-trip / post-trip resistance measurements.
Sign up now
to remove ads between posts


Quick Reply
Message:
Thread Tools