kyleservicetech

I've put together a TwinStar with a pair of Hyperion motors, Hyperion ESC's and Spektrum radio.


The receiver is on top of the wing, and extensions were made to plug into the elevator and rudder servos. The servos are a mix of various brands, including a Futaba S9202 for the elevator servo. This setup was tested around the workshop, and out in the traffic circle in front of our home. The temperature was in the mid 60 degrees F. No problems were noted.

Yesterday, the maiden flights were made. The temperature outside was 33 degrees F. The first flight went OK. Before the second flight, the elevator servo was very erratic. The servo overshot its travel and hunted a bit before settling down. I plugged in the servo directly to the Spekie AR6210-x receiver, and all was OK. Two more flights were made.

At home, this setup was connected to my Tektronix oscilloscope to take a look at what was going on. Normally, the receiver servo signal output is a repeating pulse of 3 Volts DC, 1.5 milliseconds wide, and repeating at around 20 Hertz or so.

It was quite apparent that this issue was temperature sensitive, as the erratic elevator servo never happened in my workshop.

The elevator servo signal was full of a lot of "overshooting" where when the signal went to zero volts, it undershot zero volts by about 3/4 volt DC. The receiver was exchanged with a Spekie AR600-x servo, and similar problems were noted. Installing a Spektrum AR7000 receiver eliminated the overshooting, and resolved the problem.

That "Overshooting" was definitely caused by the Futaba S9202 servo. No other servos, including a standard Futaba S3003, S3004, and various Hitec servos had the problem.

I only have one of these Futaba S9202 servos, so don't know if its a bad servo or what. That servo works just fine with shorter servo leads.

So, is there a compatibility issue between the S9202 servo and Spektrum AR 600 and AR6210 receivers? Hard to tell with only one servo sample. But, a word to the wise.

fhhuber

Bad extension or other contact issue between servo and RX will cause this.

kyleservicetech

Yeah, the first thing I did was check contact resistance of my servo extensions with my Fluke 87V digital multimeter with its ability to measure resistance down to 0.01 Ohms.
http://www.amazon.com/Fluke-FLUKE-87.../dp/B0002YFD1K

All leads checked out at around 0.05 ohms, including the servo wires them selves. In my case, servo wires were eliminated from the possible causes.

fhhuber

Temperature dependent ... usually indicates a fault. Put the stuff in the fridge for a while then check it cold.

A bad resistor in the servo is a possibility. Cracked resistors can give the indications you described.

Wildflyer

I can't understand where the voltage to undershoot zero came from. You have an odd one.

That Futaba servo is not a cheap one so it would be hard to just shelve the thing and forget about it. But yet if you know the problem is caused by it. I would have a hard time using it with confidence.

I have seen on large IMAC planes, that some guys used the little ferrite donut chokes on their long servo leads. Maybe that would have helped.

Along time ago, I mixed 3 and 4 wire servos and never had a problem. Some people told me that mix would absolutely not work. I just didn't tell my airplane, it flew just fine.

kyleservicetech

Yeah, I've got a bunch of those toroid donuts that were used on some old brush type motors that worked well.

But in this case, the problem was all low frequency, not radio frequency.

After more testing, I suspect the servo was intermittently shorting out, synced with the servo pulse. I duplicated the voltage reversal on my scope with a 1.5 ohm power resistor stuck across the servo plus and minus connections.

Why did the servo do this? Don't know, but suspicions are that the servo had some moisture inside of it, and at freezing temperatures, that moisture condensed on critical parts of the servo circuit board. Sounds weird, but before retiring at work, we had a lot of problems with this same exact phenomenon. The company spent a LOT of $$$$ resolving this problem. Finally by encapsulating the circuit boards in a solid 3/8 inch of encapsulation. (Our equipment is installed outdoors, all over the world, from Prudhoe Bay, to South America, to the jungles in Central America.)

This is an expensive servo, but I don't trust it. It came from a club member in a box of servos for a good price.

Wildflyer

"This is an expensive servo, but I don't trust it. It came from a club member in a box of servos for a good price."

Well at least you didn't buy it new, at any cost it is not worth it unless you can trust it.
I would maybe run it in a scale electric boat, or somthing else, where it won't hurt to fail.

I've had things go weird before and it drives me nuts until I can figure out what happened.

Good luck on your plane, I love twins. I just bought a pair of rather different motors on Ebay, that are rated in every listing I could find, to put out up to 2000 watts. I will do some testing to check them out.
Right now I can't even think of how big a twin I could have with 4000 watts available.

kyleservicetech

OK

I pulled off the bottom servo cover, and hooked up everything to the Tektronix scope as before. Touching the servo electronics with a very slightly moist finger tip resulted in the servo doing what it did at the field on Sunday.

So, IMHO, their is a fair chance that the Sunday servo problem was due to moisture inside the servo from condensation, after being moved from the workshop to a temperature of 32 degrees F.

Before anyone suggests that servos should not do this, FYI, just about any electronic equipment will malfunction if moisture is present at the "Magic Spot" on the circuit board.

This servo is one of two that came in the box of servos. The first one had problems and was junked. I'm going to do the same with this one.

kyleservicetech

H'mmm
4000 watts total?

Lets see, the power to fly a model goes up very roughly as the cube of the wingspan. My 55 inch wingspan model flies with 900 Watts. So, increasing that to 4000 watts is an increase of 440% of power. Cube root of 4.4 is 1.64. that 1.64 times 55 inches suggests a wingspan of somewhere around 90 inches.

That follows pretty closely with the power input on my Giant Big Stick, with its 80 inch wingspan, and its power input of 2800 Watts. (90 inch /80 inch^3*2800 Watts=3980 Watts)

Doing the same very rough calculation with my Hanger 9 Kantana, with its 59 inch wingspan predicts 1100 Watts. I'm running 1180 Watts on that model.

Seems a very crude formula for calculating watts required for very powerful electric models, with wingspans over 55 inches is as follows:
Watts required = (Wingspan of your model/55)^3*900 Watts.

Of course, your results will vary. My models are capable of vertical pull outs, virtually out of sight straight up.

solentlife

Denny .. I have va bunch of JR servos ... 501, 5001, 101, 1001, 401 etc. - I will not use them on anything except non-critical controls. One caused my biplane to crash into the runway couple year back ... others have shown erratics in weather like yours.

After reading anothers post few years back about the grease in the gear train gelling up in the cold - I had a look at mine - not the above - I mean todays servos ... yep - have to agree. Even though silicon based - there must be an ingredient in it that does not like real cold. So I have removed the great blobs of it generally inside servos. The slowing has stopped. They now work fine in real cold ... dunno if anyone else has noticed it ?

We all assume we can fly in all ... but in truth there are still limitations we need to be wary of.

Cheers
Nigel

fhhuber

5000 watts is in the class of a 50CC. 4000 would be about 35cc size.

But you can jam 5000 watts into something intended for 1.20 glow... (see avatar pic)