Extension Leads Problem

[Tosh McCaber]

On rebuilding my crashed Phoenix 2000 motor glider, I've installed the elevator/ rudder servos in the tail, which will hopefully resolve the nose heaviness problem with the elevator/ rudder sevos under the wing. However, I've run into a problem.

When hooked up wth the original regular servo leads- no problem. However, on hooking up the servos with the extension (600mm long) leads, there's a slight continual juddering from the servos.

Now, the leads from the wing servos- aileron and flaps- both have (soldered) extension leads into the fuselage. These extensions look to have thinner wires than the new elevator/ rudder leads. No problems with that set-up.

The new extension leads are from EBay, but look to be of good quality. The leads are thick- certainly thicker than the leads on the Towerpro servos, so I'm at a bit of a loss?

[Trevor]

When you say the leads are thick, are you sure that's down to the wire gauge not the insulation?

[Tosh McCaber]

Hmm- Not sure!!

[Tosh McCaber]

Just wondered- would a different Rx make a difference?

[Martin Harris - Moderator]

This sort of thing was more prevalent with 35 MHz and long leads - putting a ferrite ring with several turns of the extension lead through it close to the receiver was the cure.

Edited By Martin Harris on 14/08/2020 19:50:30

[Robin Colbourne]

I'm no electronics expert, but could there be crosstalk between the two signal wires since you are running them parallel for some distance? If you could set up the system out of the model with the servos in opposite directions from the receiver, then in parallel, this should demonstrate if this is the problem.

If you know any electronics experts, referring to this may help them identify the problem in your particular case Noise in servo leads

It appears that mixing copper and tinned copper leads can be a source of noise.
 

Edited By Robin Colbourne on 14/08/2020 19:56:11

[Trevor]

First step must surely to try some better quality extension leads. If these are digital servos, they can draw quite high transient currents which, if the leads are resistive, will cause equally significant voltage fluctuations at the servo end which could account for the juddering.

[Tosh McCaber]

My testing was one at a time, outside the model. The servos are not digital, just lowly Towerpros!

Obviously something to dio with the leads- maybe try ferrite rings?

[Keith Miles 2]

I agree with Trevor and Robin.

The problem is either volt drop in the leads, which would be my first instinct or a signal issue which I would have thought less likely given that this is not an unusual set up.

And as Trevor says, don’t let the thickness of the insulation fool you. You certainly wouldn’t be the first, or are likely to be the last to make that error. It’s one thing that annoys me that cable size is rarely quoted, at least in my experience. All that one tends to see is “Heavy Duty” and even that can be a leap of faith unless buying the truly “genuine” article at a premium price!

As suggested, start by trying everything outside of the model, with leads and servos laid out straight and kept well apart. If it’s volt drop, as I suspect, the problem will still be there.

[Tosh McCaber]

I naively thought that extension leads were extension leads- all being equal. Not a question of money! Can anyone give me a lead (pardon the pun!) to somwhere who sells "heavy duty" leads?

Thx!

[Nigel R]

Cheap first shot, try switch cleaner spray on the servo plus and extension lead sockets.

Does the problem happen with just one servo on an extension?

[Denis Watkins]

Posted by Tosh McCaber on 15/08/2020 06:45:39:

I naively thought that extension leads were extension leads- all being equal. Not a question of money! Can anyone give me a lead (pardon the pun!) to somwhere who sells "heavy duty" leads?

Thx!

Properly made leads are just leads Tosh, as you say

I suspect Nigel has a point you need to check as the " weak link " is always the pins and sockets

On another day you would fit long leads without a problem

Electrons love flowing down cables

Edited By Denis Watkins on 15/08/2020 07:15:32

[Neil James]

I bought a crimping tool and connector set some time ago to make my own leads and after some initial confusion on how to use the crimper (no instructions so thank you Youtube) I now make my own leads. The advantage to that is that I can use different gauge wire for power and signal wires, as there’s minimal current on the signal wire I use 10/0.1 TC wire which I bought some time ago from Radiospares. The power wires get 7/0.2 gauge. You could use 14/0.2 for digital servos but I’ve had no problems even with 1m long leads.
The numbers refer to the number and diameter of the wires inside so are independent of the thickness of the insulation and that’s what you should be looking for when buying wire. There’s a wide choice on eBay.

I’ve used this system on large gliders with 2 servos per wing and tapped the power leads for the inner flap servo by stripping back the insulation part way along so only 4 wires are needed. I can then use a smaller 4 pin deans connector to hook everything up.

You need to be careful with the crimping so buy extra pins and practice then test before use! It’s been one of my best investments in modelling tools though, not just for long leads but for making up combined switch/charging points.

[Cuban8]

Occam's Razor..........Tosh - as others have said, don't get bogged down in complexities at the moment, particularly issues of cross talk & ferrite rings, twisted cables etc. There are several reasons why servos might be unstable in a given installation, so for the moment start from scratch and set up your RX and servos out of the model and on the bench and have them powered by a battery with servos directly plugged into the RX. If everything works OK and range tests OK in the garden, then I'd work forward by replacing the original parts from the model one at a time (extensions and BEC) and try again - I think substituting the suspect parts until you get a definite and repeatable solution is the only way. Try it with just the elevator and rudder servos and then with the ailerons plugged in. See what sets off the juddering. If you have a method of monitoring the system's voltage, make sure that the supply is robust and not dropping off.

You say that the problem has only started since moving the elevator and rudder servos to the back of the model on 600mm extensions - I doubt whether it's a catch all 'noise' issue caused by 'long leads' on their own, my own experience has never come across that even with servo extensions of a metre or more, more likely a supply issue to the servos, hence my suggestion of running the whole system from a known good battery with direct connections (no extensions) to the RX to start with. An annoying fault, - will be interesting to know the solution. Good luck.

 

Edited By Cuban8 on 15/08/2020 10:50:47

[Tosh McCaber]

I've just been doing some further investigation. Firstly, I tried to clean the contacts on both extension leads, and servo plugs. Still juddering happens, with the extension leads on. Again, the servos (TowerPro MG 90 metal servos) work, with no juddering when directly plugged into the Rx.

Then I remembered that I have some unused TowerPro SG90 (plastic geared) servos. They work fine on the extensions, with no juddering. However, I don't particularly want to use them on this 2 metre wingspan motor glider. So I'll have to get cracking, and buy some more expensive servos I think. (Mind you, until this little(?) problem, I've had no problem with TowerPro MG90 servos to date.)

Interestingly, I checked the voltage between positive and negative servo leads with a voltmeter, and found the reading was 4.9v, both with extension leads on and using the servo plugged in directly.

Neil, thanks for the crimping methodology. Maybe I'll go out and buy a crimping tool, and servo leads to experiment with. However, it may not be used very much, since I don't build too many very large models requiring long leads!

If I can find a reliable supplier for long leads, I'd prefer to go that way.

Cuban 8, just saw your post after I had posted- thanks for your input!

Edited By Tosh McCaber on 15/08/2020 11:15:03

[Tosh McCaber]

Having read your input Cuban 8, as I mentioned, I ran the system with no extensions- fine. I definitely suspect the servos, although the voltage holds up through the leads. I'll come back after I pick up replacements!

[Keith Berriman]

Hi using long extension leads from my experience on similar installation is that if the leads are the flat type and you have some length left try folding it back on its self to form a little 8 section to try and eliminated some back emfs and small ty rap to hold in place. If I go for long leads I use the twisted wire units

[BackinBlack]

+1 for (tightly) twisted leads, just good practice for longer leads. Nevertheless I can't help but think that the leads are the most likely cause of the problems. As C8 says a progressive test of all components should identify the problem.

[Bob Cotsford]

Noise sensitive amps in the digital MG90s being upset by the capacitive/inductive effects of the long leads? I know these cheaper servos can draw more current than you might expect considering their size, could the pulses generated be enough to cause problems? I've used them quite a few times myself but not on long extensions.

[Martin Harris - Moderator]

Are the MG90s the digital version? Mixing analogue and digital servos can sometimes cause problems. Does your transmitter provide timing options?

[Chris Bott - Moderator]

Some radios (maybe most) provide a servo pulse voltage of only 3.3v. some servos are right on the edge of being able to detect this and read it correctly as they are expecting a pulse as big as the battery volts.

I wonder if, in this case, the extensions are attenuating the pulse just that little bit?

If you Google "servo signal booster", you'll see that there are solutions, therefore it is known as sometimes an issue.

Having said that, it's probably as cheap to buy servos that are less fussy.

[Martin_K]

Posted by Chris Bott - Moderator on 15/08/2020 12:45:06:

Some radios (maybe most) provide a servo pulse voltage of only 3.3v .....

I do not understand what you mean by that? Are you referring to the logic levels used in the RX? (In consumer electronics typically 3.3 or 5V).

Servos are actuated by the pulse width (duration in milliseconds) of the signal, not by the signal's amplitude.

[Andy Stephenson]

Martyn,

What is being said here is that the control pulse amplitude from the receiver is quite often less than the supply voltage and if the servo needs a pulse of a voltage greater than the receiver is supplying it would cause it not to be recognised by the servo. This being independent on the pulse width.

A.

[Chris Bott - Moderator]

Martin you're right, these days most Rx will have a CPU that works at 3.3v so it's output pulses will be 3.3v "tall".

Yes the Servo measures the width of these pulses. It likely does so by having a threshold (slice) voltage, effectively measuring the time that the pulse is above that voltage. In a perfect world the pulses have vertical sides and a flat top so the servo will work if the pulse height is greater than the slice voltage. (And won't work at all if pulse is lower)

In the real world the pulses can have quite rounded corners, probably more pronounced with long extensions, so if the slice voltage is up in the rounded area things can start to get quite jittery.

[Martin_K]

Unlike other electronics components, no servo I have bought came with a datasheet describing the requirements for connecting to the device. It would be nice to know what to expect in advance of testing.