ESC and Battery connections

[Terry Whiting]

Timbo, can you help 

A few years back when I first came into electric,  I was told never extend the leads between battery and the ESC, and never have done,  but. is this a real issue today? as I have heard of some multi models with their  ESC's in the motor nacelles.                                                                                                                      My reason for asking is I would like to go brushless in my Tony Nijhuis Lanc, but my paper conduits in the two root wing bays will not take 6 wires without surgery. 

[Pete Rieden]

Yes, it's still an "issue". Most of the ESC manufacturers suggest no more than 8" (200mm) between the ESC and battery, but the motor wires can be as long as you like. The reason is the inductance of the cable and the way it causes large voltage surges when the input MOSFETs switch. If you want longer cables you can fit extra compensating capacitors, but this should be done with care to avoid creating a "resonant" circuit. There is some guidence on what and how to do this on the Schulze website.

PDR

[Terry Whiting]

Thanks Peter for your help. I would rather be safe than sorry, so in my case it would have to be surgery, 

cheers  Terry.

[Tim Mackey]

Although Peter is correct, there have been instances where due to design limitations the "rule" has been broken, with no ill effects. Usual disclaimers apply

[Terry Whiting]

Timbo thanks to you also.

I think before  I perform a tracheotomy on the paper conduits I will make up a bench test piece. Any idea what sympton I should look for, or would it be 'goodnight ESC?'    I understand your usual disclaimer .....

[Tim Mackey]

Leads ( and of course motors) are inductive. In switching power supplies (e. g. a speed controller! ) these produce high voltage spikes (potentially over 200 Volts) when the switching transistor (MOSFET) is switched from the ON to OFF state. A battery pack is able to "damp down" those high voltage peaks which can destroy your controller, - only when there is no ( or very small ) induction between ESC and battery.
With long leads the ESC does not "see" the battery and the high voltage spikes are therefore not damped - your controller gets more than the allowed voltage. Depending ( to some extent ) on the quality of the ESC and especially its large capacitors, likely "symptons" could in a  worst case scenario a complete goosing of said capacitors -if the battery is a long way away as it were, the capacitors take the full brunt of it instead. By all means try a bench test up first, but this does not of course ensure long term success. I have a large EDF which has a horribly long run of cabling from the battery to ESC ( almost 2 foot! ) but have experienced no problems to date on bench testing. Good luck.

[Terry Whiting]

Thanks.   It would be nice if spikes could be recorded whilest testing. 

[Pete Rieden]

Terry Whiting wrote (see)

Timbo thanks to you also.

I think before  I perform a tracheotomy on the paper conduits I will make up a bench test piece. Any idea what sympton I should look for, or would it be 'goodnight ESC?'    I understand your usual disclaimer .....

The usual symptom is the escape of the magic white smoke after between 10 and 20 seconds of running, but it can occur at any time (several minutes, several flights or even several years) so a single test doesn't really give much assurance. Adding additional capacitance is a fairly simple exercise (see the schulze website) but do remember to keep the capacitor legs as short as possible - once they are more than a few mm long the inductance of the capacitor lead completely negates the benefit.

25 years ago (gulp) when I worked as a trials engineer in the sonar world we had a new 1kW white-noise generator (part of a sonar "jammer") which had MOSFET-driven output stages. When tested the first couple only lasted a couple of minutes at half power, and when we investigated we found that the wireman who assembled it had decided to make a neater job of the installation by standing the MOSFETs on the ends of their legs about 10mm from the PCB. This made a much neater installation for the heat sink, but the inductance of the legs killed the MOSFETs. The next two were build "to drawing" with the MOSFETs snugged right down to the board. These two successfully completed a 10 hour type-test at 1.6kW...

PDR

[Terry Whiting]

Peter  Thank you, I think I will leave well alone, as  my knowledge in electronics is limited to the extreme.   My Lanc will  go under the knife.

Thanks again you guys for your help.

[Alan Brock]

I am building a 63" Blackhorse Mosquito with two AXI 2814/12 s and the well illustrated instruction manual shows the ESCs mounted in the engine nacelles and the Lipo mounted in the nose with long connecting wires (I estimate to be about 350mm). Is this likely to cause a problem?

[Tim Mackey]

Well really, all that was said above applies in your case too.... you will just have to try it .

[Steve Hargreaves - Moderator]

On a similar theme can I ask the panel about the need to "shield" extended motor wires.

I have just built a Tony Nijhuis EE Lightning & the plan shows the ESC at the back with the motor & extended battery leads to the (nose mounted) battery. The Rx lives up front. Ever mindful of the good advise on the forum I swapped the Rx & the ESC so as to keep the ESC to battery wires nice & short. This means that I have to extend the ESC to motor wires & also that said wires now run very close to the Rx (now at the back near the motor). Keen to mitigate the problems of possible EMI I have wrapped the ESC to motor wires in aluminium foil in a very basic attempt to provide some shielding.

Does the panel think this a good idea or a complete waste of time?

I am on 35MHz & the model seems fine in ground tests but is as yet unflown due to the weather  (I think I need to put floats on my Lightning!!!)

[Tim Mackey]

Well I am no engineer, but I doubt it will do much unless the shielding is somehow grounded ? Wrapping receivers in tin foil used to be a popular "fix" to help prevent interference, but from memory I seem to recall the "experts" suggesting it would only be effective against microwave length stuff anyway. I did try it on my EDFs when I was on 35Mhz and it made little if any difference. As rapid as the EMI is likely to be in the motor supply wiring, I very much doubt its that high a frequency! Won't hurt I guess, but whether it will provide the solution you are looking for is up to someone better technically qualified than I to say. I also have a particular model which has very long ( almost 2.5 feet ) cabing from ESC to motor, and is operating at around 1kW+ yet I have experienced no problems...however I do ( perhaps unjustifiably ) put a lot of trust in the immunity to such potential interference of the 2.4Ghz system I use

[Myron Beaumont]

Timbo

Cant be sure that you're not employed by Spectum Or the 2.4 brigade £ 270  & still flown nothing

[Myron Beaumont]

  PS   The weather is a factor as well of course !

Thinking of using all my old fashioned batteries to power 'er indoors's boats (  Relaxing (for a change))

 Now got a lipo for my Tx that doesn't fit   Grhhhr

  You did tell me not to "upgrade " to nimh from nicads . Say no more !

[Steve Hargreaves - Moderator]

Thanks Timbo.....like you I have my doubts it will actually achieve anything but I was working on the Tesco theorum of "Every Little Helps"!!!!

I did have an electrical model with a single conversion 8ch GWS receiver that was very glitchy (is that a word??) & wrapping the Rx in aluminium foil certainly improved matters in that case. In the Lightning I have a Corona dual conversion Rx with DSP (or digital signal processing...wow!!!) & these seem very good in electric flight....no glitching at all in my other models......

[Brian Parker]

Hi! Just back from a few days break, some interesting postings on various threads.

 

If screening is deemed necessary (unlikely in this case) then wrapping cables in aluminium foil will have minimal effect on any radiated interference unless ‘grounded’ (to 0volt potential) at each end of the screening. It’s the grounding that completes the low-impedance path.

Painting the receiver casing, preferably the inside, with conductive paint then grounding the casing may prove more effective, but does require care when removing/replacing the casing, (although now a days many receivers don’t have proper casings). If painting the outside of the casing take care to paint only where it’s required!

 Also if the casing/conductive paint is not grounded to 0V the conductive paint may act as an additional antenna with predictable results.

[Brian Parker]

I did say to 0volt potential.

[Pete Rieden]

Brian Parker wrote (see)

I did say to 0volt potential.

Actually that's rather the point. Whilst the connections to the two ends of the screen may well go to the same potential, they will be of different lengths. This results in the connect at each end of the screen being at a different potential whenever a current flows. Since the screen now has a potential difference between the two ends there will be a current between them, and that produces the "earth loop".

This is why the signal ground used in (fregsample) hi-fi equipment is most explicitly NOT connected to the chassis earth at its mains supply - in fact considerable effort is expended to isolate them. They then provide a single point on the back of each bit of equipment to connect up to an earth, and only ONE of the equipments has the chassis and signal grounds connected (usually the amp). If you connect the mains earth to the signal ground on any of the other bits (the CD player or the turntable) you immediatly get an awful 50Hz "hum" - this is the "ground loop".

This phenomenon is a huge problem in aviation systems engineering. On aircraft we want to save weight and space by generally not providing ground wires - we want to use the aircraft structure as the ground return conductor for everything, in the process typically saving betwenn 1000 and 2000lbs of copper weight in the wiring looms. This does, of course, introduce massive ground loops because the structure is carrying everything from the tiny ground-reference currents from the receiver antennas to the huge gobbetts of current returned from radar transmitter amplifiers and servo mechanisms. But we need the weight reduction so we accept the poor signal environment and develop extremely clever signal formats using balanced line techniques to achive the large common-mode rejection ratios required to clean up the signals. It's a trade-off in conflicting requirements!

PDR

[Brian Parker]

Peter,

I thought we were writing about Toy Planes with very short cables and no ground plane (no pun intended) therefore any potential difference would be negligible (and I didn’t foresee mains hum being a problem), but not wishing to become involved in a technical argument and as you are obviously determined to be correct your views are OK by me.

However, in Steve’s case, a length of un-insulated stranded wire connected from ESC 0volt to Receiver 0volt running tightly alongside the cables (in effect a ‘drain’ strap) then wrapped in the foil may prove effective in ensuring EMC. I doubt however that any screening of cables will be necessary.

 

[Steve Hargreaves - Moderator]

Well whatever the theories & counter theories about screening or not screening, earth loops & the like I DO know I got a whole load of glitching on my Lightning which made the maiden flight very interesting. Basically it climbed away OK (so I know it flies) & then became an absolute pig to fly....like trying to stuff an octopus into a string bag.....!! I had to fight it all the way but then lost on a technical knockout!!!!!

 The damage is slight however & the thing will fly again.......I intend to swap the ESC & motor to the positions shown on the plan. This will mean extending the motor wires significantly but will move the Rx away from the motor & motor wires.......

Fingers crossed!!!!

[Jeremy Butler]

What about the lead length from ESC to the motor. Is it ok to shorten them? Should they be a minimum length?

[Steve Hargreaves - Moderator]

There are many more highly qualified people than me out there Jeremy but I would say no. The leads can be as short as you like........

[Jeremy Butler]

Thanks Steve. I thought that might be the case, but before I get at it with my cutters and soldering iron I will wait and see what the consensus is.

[Tim Mackey]

shorter the better really