wildblueyawner

Advice anyone,

I'm working on a giant-scale model where some servo leadwires may be nearly 15 feet long, so I'm concerned about noise pickup and voltage drop.

Bunch o' Questions:

1. For filtering noise on very long servo leads, which is more effective - ferrite beads or noise / glitch trap circuits?

2. The filter should be placed closest to the servo, correct?

3. Can noise in servo leads somehow affect the receiver?, e.g. generate EMI?

4. With filters in place, can wires to multiple servos be run alongside each other w/o problems, either noise-related or crosstalk?

5. I figured the most reliable way to determine min wire gage would be by measuring voltage drop through various wire sizes ==> Is there a recommended max allowable voltage drop?

6. I can chop the max lead length down to about 5 feet by putting a receiver in each wing to feed the aileron and flap servos. But since I don't want to put a battery in each wing, I'd have to power the receivers thru about 7 - 8 feet of wire from the batteries in the fuse. Bad idea?

Any advice appreciated!
Thanks all

Buz^

My Feedback: (2)

Usually when we try and put a string of beads on our ferret's they just scurry out of them

JohnMac

wildblueyawner.
First point is to use the best quality heavy guage wire you can for your extensions, and preferably have them made for the job. This reduces the number of plugs.
Second, you can expect some problems with RF noise being picked up by leads of this length and upsetting the Rx or in some cases the servos.
On the other hand you may have no trouble at all, so try it (on the ground) first and see. Ferrite beads or rings work by increasing the impedance of the wire, and so making it hard for the RF to get through. These are placed as close to the RX as possible.
A opto isolator is next step up if beads don't work. Normally these go in the same place.
Occasionally you may find just a single servo jitters, not the whole lot. In this circumstance likely as not the servo pulse is getting a little out of shape and there are devices (glitch traps) that reshape the pulse and block any spurious RF at the same time. These are placed close to the servo.
In theory cross talk is possible but I am not aware of any such problems being experienced.
I hope this helps,
Regards,

John.

Whirley Bird

My Feedback: (1)

I'm not an expert at this but you just can't slip any bead on the wire.
Beads come in different powder mix when they are being made.
Maybe Ladyflyer will see this and drop a note
As a ham operator I use beads to suppress RF coming down the line.
You want to suppress noise.
The RX sends a SIG to the servo along with the noise.
You can degrade the the design built into the RX and servos.
I would contact the manufacture for the right way to eliminate any bugs.
15 foot run?? Thats a big bird.
Why chance anything until you get solid info on the beads

Lynx

Ferrite beads are stupid cheap so there's no reason not to use them and circuit filters. You should probably put beads on both ends of the servo wire. With runs that long you should keep the individual servo wires at least several inches apart. Servo's pull enough current if you have wires right next to each other you're bound to get some capacitance between the wires. Having very long leads on the batteries to extra recievers will probably cut down on noise. You'll want to use wire as big as you can (even excessivly so) to avoid voltage drop and generally improve any signal you put through it. You can't use wire too big, you can use wire too small, better safe than sorry. Well unless the weight of the copper keeps your plane from flying =)

Forgues Research

My Feedback: (7)

You may want to look at what I have developped for the very reason you are looking.
Fiber Optic servo extensions.
I can power servos up to 300 ft, not that you will ever need this long, but I have customers that do (the movie industry for special effects)

With these, there are no servos directly plugged in the receiver so they can never feed back into the receiver causing problems such as receiver lock up.

You simply plug the encoder end in the receiver and plug the servo/servos ( up to seven have been tested in one decoder) and connect the two of them with the fiber optic cable.
You power the decoder/servo end with a seperate battery then that of the receiver.
We even have powered a mini sub at 200 feet deep with this setup.

wildblueyawner

Hey all, thx for the helpful comments & suggestions (that even includes you Buz^ )

I can't recall the site referencing the recommended type of beads for r/c use, but they're Amidon #43-2401 ==> 43 is the material type suited for 40-200 MHz. I got 'em for less than $.50/ea here http://www.universal-radio.com/catalog/misc/amidon.html

Yup, I'll definitely consider opto-isolation or the fiberoptic approach (Roger - I'll inquire thru your site).

While reading these posts some more things came to mind - A few more questions for the group:

- Given the anticipated higher voltage drop thru long leads, should a 6V system be used? If so, can I expect higher current drain compared to a 4.8V system, to the point where larger capacity batts should be considered?

- Anyone have experience with buffer boosters? They seem to function like little repeater stations http://www.servocity.com/ServoCity/P..._boosters.html

- If whatever setup I use works noise-free on the ground, can I assume that the system is adequately filtered, and will stay noise-free all the time, in any orientation, location, etc (provided the hardware isn't moved or wires re-routed) ==> In other words, is a consistently successful ground check a definitive "thumbs up" of system noise immunity, or is it just a quick and dirty check with no guarantees?

Thx all.

Lynx

It's RF, there are NEVER any grantee's =-) The more wire in your system the more possible problems you're gonna have, if it's not too expensive the fiber optics is perfect because that removes the wire in between as a source of interference. I've put a link at the bottom here that has every bit of information you'll need to make wire choices if you're using them. If you search the page there's even an equation to find the voltage drop on a given wire.

wildblueyawner

? Links Lynx

Whirley Bird

My Feedback: (1)

Off the topic a bit but I just looked at the above URL.
http://www.universal-radio.com/catalog/misc/amidon.html






Futaba flight pack $135.95
Tower Hobbies the same thing is $109.95
The full 4 CH setup is only $149.95 from Tower Hobbies and thats with the BB servos.
Are the simple 4 CH systems being faded out?
I was also looking at the 6 CH computer radio W/servos.RX/batteries,charger,everything you need $229.00

Lynx

hmm, wonder why the link didn't make it, well here it is.
http://www.epanorama.net/documents/w...esistance.html

strato911

wildblueyawner - With a plane of that size, I assume you don't care if you add a few oz weight to ensure a clean signal to your servos. Here's a link to an electronic noise trap for long servo leads. This is what I would do in you position.
http://www.uoguelph.ca/~antoon/gadgets/noiserx.htm
If I'm right, this circuit should also perform servo reversing for dual servo elevators or flaps.

wildblueyawner

I will certainly consider the noise trap circuit - Simple 2-component device. And yes - Since the plane's wingspan (weight to be determined ) will be 354 inches (29.5 feet), won't care about the trivial added weight of a few chips and mini-boards. Servo reversing would be handy also, as I'll be using redundant elevators and split ailerons.

strato911

They're an inexpensive fix to a potentially nasty problem. Test it out thoroughly on the ground first. If you still get spurrious servo control at those lengths, check out the fibre-optic extensions aerografixs has already suggested. Here's a review where they were tested as 30 foot extensions. http://www.geocities.com/traviscouch/FAQ.html?

Forgues Research

My Feedback: (7)

Thanks for the heads up Dennis.

Recently I have improved on the design and you could actually install a receiver on the tip of one wing of a full size 747, and install the servo on the other tip and have full control with no chance of ingesting EMI. That's about 200 feet.
Roger

wildblueyawner

strato911 ==> good link for the fiberoptic extensions, sums it up nicely, thx much!

Jim Oddino sure makes some convincing arguments to go the FO route. I'm noting especially the immunity to ignition system noise. Being that my GS project is a glider, one normally wouldn't assume ignition systems to be an issue, however the model will include self-launching capability in the form of a big gas engine mounted on a pylon that retracts into the fuse aka "up 'n go."

Since the up 'n go will sit in the fuse around the rear half of the wing chord, it won't be far from the wires going out to the aileron/flap/spoiler servos. Being so focused on the long-leadwire situation, I hadn't really looked at this ignition system noise issue - It makes a total-immunity (FO) setup all the more attractive.

wildblueyawner

Roger - What is the total current drain added by the FO system (encoder, decoder, laser or whatever light source is used, and all other devices comprising the system) aside from the current drain of the servos ==> Being dependent on torque, the servo portion of total current drain should be the same as in a conventional wired setup, correct?

Forgues Research

My Feedback: (7)

The total current drain of the decoder part (where the servo plugs in) is 0.8ma which is insignificant. On a full house like my setup with six of them, your talking 4.8 ma.
The encoder part draws more because we're talking LED's which I'm driving at 10ma/ea, so for my full house system on mine, I'm drawing 60ma for the encoders plus 15 ma for the receiver .
So with a 600ma battery on the receiver you would be good for about 8 hours assuming you don't have any servos on the receiver.

With longer Fiber Optic extensions, you are also gaining in weight control, as the longer the extension, the more copper you are replacing with the plastic fiber cable. which is much lighter the copper.

wildblueyawner

For others interested in the fiberoptic extensions, a couple more links:

http://www.geocities.com/roger_forgues/batteries.html

http://www.geocities.com/traviscouch/pilot.html