Extending leads - battery or ESC?

[Glenn Philbrick]

Having looked at the last post from John and reading the link I think it might be prudent for me to fit these extra capacitors as it looks very easy to do and hopefully will remove the possiblity of loosing the model. So now where can these capacitors be bought from, how manu should I add for sometihng like a 12" increase in lead length.

Thanks,

Glenn

[John Olsen 1]

Hi Guys, Well, I have not played much with el*ctr*c models, but I did do some work on power electronics many years back at university, including playing with some early power MOSFETs. The ones I played with were toys by modern standards, they could switch two amps in 2 nanoseconds, with a maximum rating of 80 volts. The devices (which need not be MOSFETs like I was playing with) in a modern ESC are switching higher currents, quite probably in a similar or maybe even shorter time. So what happens when you switch a device off very fast? If there is inductance in series, the current will try to keep flowing. The voltage induced will be proportional to the rate of change of current with time. So if we take the inductance of 774 nH mentioned above, and switch say ten Amps in 1 nanosecond, we will induce a voltage of 7740 Volts. The switching device would have to withstand this Voltage, unless we do something about it. One solution you sometimes see when a transistor is switching a relay coil, is a diode to clamp the spike, but this is probably not a possible solution for the ESC case, due to lack of anywhere to clamp to, and also due to the cost of suitable diodes. So the capacitor at the input acts as protection, by absorbing the spike of energy. This is actually exactly similar to the effect of the capacitor across an old fashoned set of ignition points. So it is actually quite reasonable that if you increase the inductance, the energy to be absorbed will increase. The result may be to cause the capacitor to fail, or it may be to cause the Voltage excursion to increase to the point where the switching devices fail. Either way, the ESC is probably history at that point.

There will of course be some tolerance built in, but since the only way to find out how much is likely to involve sacrificing a few ESCs, I would suggest erring on the side of caution.

regards

John

[Peter Beeney]

This is beginning to get interesting, and from John’s link to RC Modelflyers, and the work done by PDR, it would certainly seem that if you you extend the battery cables you stand a good chance of eventually zapping the ESC. Unfortunately there is no indication of any inductance values found and one thing struck me when I read about the motor they used as a load, suppose later they decided not to use a motor, purely just a resistive load; would we still get the same effect? Or, in other words, is the much more obvious greater inductance associated with the motor windings also involved here? Reading all through that thread, it does seem as though it’s also the case that sometimes some folks have extended the cables and have not experienced any problems. The lengths, too, maybe unusual, 800 mm is about 31 inches, 1,400 mm is 55 inches. Would these normally be used in a standard model? Or even a large one?

Whilst I agree that if the switching time is 1 nanosecond the voltage spike is 7,740 volts, would it actually be that quick in practise? I’ve no idea of the switching time, and I don’t think this is mentioned, either, but if it’s lengthened to 100 nanoseconds the spike then becomes 77.4 volts; 1 millisecond would equal 7.74 volts, if my calculations are correct. Peter mentions a pulse width of 1 millisecond in the text, with the high voltage apparently holding up for this time. So is there some sort of interaction between the motor coils and the power leads? Definitely confusing, but is it possible that if it’s imperative to lengthen the battery leads is it an idea to place the same amount of capacitance between the leads at the points where the standard length is doubled? This might keep the capacitance balanced along the length?

In the case of the diode across across the relay coil to prevent the voltage spike damaging say a transistor switch, this is connected reversed biased across the relay terminals. When the current stops the coil inductance tries to maintain the current flowing in the same direction, so the negative end now become the positive end, and now the diode forward conducts and shorts out the voltage spike. If a capacitor is used it might have a series resistor to limit the current. A capacitor quench circuit is oscillatory, it ticks to and fro a little bit. This is caused by the inductor charging the capacitor, then the capacitance charging the inductor, then the inductor charging the cap…… a bit like a bungee jumper going up and down, using up all the stretch. This oscillating period doesn’t last very long, though, but I think this is the 'ringing' cables that Peter speaks of in the text, so I think this might point to some sort of slight resonance going on.

With the greatest respect, though, I’ve always understood that the ignition capacitor is slightly different. Let’s assume we have an auto coil, not necessarily to do with cars, this is one with the primary winding as part of the secondary winding, to make one circuit, the primary current outlet is at a tapped point. The capacitor is in parallel with the points, which are in series with the primary winding, so that when the points open the capacitor is then placed in series with the primary winding. The decaying current collapses into the capacitor, and the resultant voltage due to the coil inductance, say around 400 volts, induces a further higher voltage in the many more turns of the secondary winding, which together with the primary 400 volts in series with it causes a spark at the plug. This discharge into the capacitor is also oscillatory, which helps to maintain the spark for a while. When the points close again any voltage remaining in the cap is shorted, ready for the next opening. These oscillations are in the nature of of the beast, it happens when capacitance is linked with inductance, and this might all be going very slightly in the direction of a tuned circuit, the capacitance has to match the inductance, it’s always around 0.2 microfarards. Significantly, though, it’s also a high voltage type, too, in the order 6 to 8 hundred volts, to withstand the high induced voltage on the primary.

I very seldom read other forums, I often find them difficult to understand, but I notice one poster there certainly confirms what I’ve been saying for years, and that is there are no problem with paralleling two or more BEC’s together, or a BEC and a battery. But this situation has also been confirmed by a few others, too, on odd occasions, so now I’m not entirely in isolation.

The Alternative Way is gradually gaining momentum…….

PB

[Steve W-O]

PB, the purpose of the capacitor on a car is simply to prevent arcing on the points. The car will run fine without the capacitor, but the points will burn very quickly.

As for keeping the spark going longer, what for, the fuel is alight already?

This was an old snake oil story by the people selling early electronic ignitions. The only benefit ever found was having two spark plugs firing at the same time, hopefully resulting in the flame front meeting in the middle, but as so many things affect the time a spark plug fires slightly, never any practical advantage, exept if one failed, the reason for twin systems on planes.

I have read so much about the ESC story, but I must say I am swayed by scope traces, they don't lie!

I would like to try it myself one day, and compare signals at the battery end and ESC end of a longer cable, maybe I will see if the batteries in my portable scope are any good, easier to get into the plane area.

[John Olsen 1]

OK, dealing with some points that have arisen...The switching devices in the ESC will certainly be chosen to have a high switching speed, because the time taken to switch is a major part of any loss in the circuit. Think of it this way...when the device is on, it has a low resistance, so very low power dissipation. When it is off, there is minimal current (leakage only) so there is also a very low power dissipation. During the time that it takes to switch, there is both current through the device and voltage across it, so there is power dissipated. So for a hypothetical device switching 10 Amps from a 7.4 Volt battery, the peak dissipation will be when the current is at 5 Amps and the voltage is at 3.7, eg a bit over 15 Watts. But since this power is only dissipated for a very short period of time, eg while the device is actually switching, the mean power dissipated is very low. So we can have very small light ESCs, instead of them needing a big heatsink. (also there will be some power dissipated during the on time, as no device actually has zero resistance.)

So a capacitor actually lets the device itself turn off very rapidly, while providing a path for the current to continue to flow a bit longer, meaning that the voltage across the device will not go so high. Yes, this is potentially an oscillatory circuit, one of the designers problems is to make sure that any ringing is damped well enough. The circuit I was designing at university actually showed a slight ringing at first, about three cycles for about a nanosecond or so, but a tiny bit of capacitance with a resistor in series damped it out, at the expense of slightly slowing the rise time.

As far as the ignition goes...well, this is in fact a very similar situation. The points open, the current tries to keep flowing and so creates an arc across the points, rapidly destroying them. A suitable value of capacitor provides a path for the current to flow while the points get far enough apart, while still interrupting the primary current fast enough to induce a high voltage in the secondary. There will be an oscillatory effect there too. I don't think the mixture in the cylinder cares much about that, provided enough energy gets delivered into the plug gap it will ignite.

John

[Peter Beeney]

Steve, I’m sure you may well be right about a car running fine without the capacitor, but I’m afraid I’m not swayed that easily. It doesn’t always seem to work like that every time. In another early life I bought an old motor bike that had an ignition switch cobbled up under the seat. When I checked to see how how it had been done I found the capacitor wire had been cut in half and extended to the switch. So the capacitor was switched in and out of circuit. This certainly stopped the engine when required, and definitely prevented it starting, too. Why it was done that way, I’ve no idea, but the bikes in those days often didn’t have very much security anyway. I didn’t keep it very long, but the switch and sparks were certainly never any trouble.

Keeping the spark going a little longer is perhaps only beneficial in the sense that the stronger and longer the spark is, the better it will ignite the mixture. One little snippet I’ve gathered somewhere is that a spark of less than one millisecond will not reliably ignite todays lean-burn mixtures, although that’s probably a bit obsolete now, too, I'm sure CD units must be all the vogue nowadays anyway, controlled by engine management systems.

I’m not very good at modern car electrical systems, but I do know that a series LC combination is a very low impedance, thus it will allow maximum current flow through it and develop maximum voltage across it, particularly near or at resonance; which is exactly what we are looking for here.
This was the story that was preached to me well enough, and until I see different I shall stick with it. The faster the magnetic field collapses, the higher will be the voltage produced in the primary winding and thus in the secondary.
A parallel LC circuit becomes a high impedance at resonance.
However, I do appreciate all this applies to AC frequencies, of course, and we are only looking at one shot at the time, so that must make a difference.

I’ve never heard the snake-oil two plug story, but I’d say that would be a major modification (expensive!) to a modern car. In general, I’d consider you’d need to be pretty persuasive to convince many people to do this.

I’ve never seen seen any traces of the input to the ESC, but I’m sure they exist, and, as you say, they undoubtably will be accurate, it’s just of interest as to where exactly where the higher voltages are coming from. I don’t have any means of finding out, unfortunately, so, as always, I shall just have to speculate.

With any further regard to the ignition capacitor, I will have a swan around in the library and see if there are any modern up-to-date textbooks on the subject lurking about; but I expect they are a bit thin on the ground now. If I can find any info. I will repeat whatever it says. Although it’s only of very limited passing interest, I guess, it really isn’t going to be earth shattering stuff. I started out by thinking that about the capacitor and the points arcing, which is perfectly true, of course, and they still tend to burn a tiddly touch even with the capacitor in circuit, I believe, but when I started to look closer I realised there might be more to it.

All good stuff…. ...if you are so inclined!

PB

[Glenn Philbrick]

 

See here

Hi Guys. I now have a pack of capacitors from Farnell as recommended in the post on RC groups and RCMF these are the ones in the link above. In the link on RC groups that has very good description of how to mount the capacitors it says that they have a polarity, mine have no polarity markings on them, one lead is longer than the other so I am stuck. Any help would be good.

 

Thanks,

Glenn

Edited By Steve Hargreaves - Moderator on 19/04/2012 15:44:42

[Steve Hargreaves - Moderator]

Glenn the capacitors in the link show a lighter coloured line with small rectangles printed on it.....a bit like "minus" sign.........this is the negative side. The negative lead is also shorter than the positive lead.....

Please take a moment to read this too....

Edited By Steve Hargreaves - Moderator on 19/04/2012 15:48:41

[Glenn Philbrick]

Thanks Steve. This is just the info I was hoping for. I hope I did not cause a problem by the way I posted the link, I've printed the instructions how to do it now so I will not forget.

Glenn

[Tosh McCaber]

I've just posted a thread re extension leads, on which I was directed to this one. My original query was:

"To those of you who are more experienced with electric power than I am.

Can you tell me - is it in order to extend battery leads between battery and ESC? I need another 75mm, if possible. I could either put an additional lead in , with corresponding plug and socket as an extension cable, or solder on an extension lead to the ESC cable.

I have a 3536/1100 motor, 40a ESC, and 2200mah 3SLiPo battery on my model.

I seem to recall somewhere that this may or may not be practical (which may or may not have referred to shortening cable?)

Any advice welcome!"

Having read the various discussions abovein this thread, it appears that it's in order to extend the motor, rather than the battery leads. If I go with motor extension leads, I'd need an extension of approx 125mm. Does that sound OK?

[Keith Berriman]

I built a Depron Jet shape model with battery at front and motor at rear and needed to extend battery wires by 9" to get things working

I was of the opinion that battery power was straight voltage movement where as the ESC to motor is pulsating current and more likely to cause problems

[Chris Walby]

Before this gets into a raging debate (again) the best bet is to read the manufacturers guidance as they contradict each other. Some say you can, some extend with capacitors and some it it not advised.

Without knowing the make or model of the ESC its just peoples opinions.

Tin hat on and TTFN

[Simon Chaddock]

Tosh

It is the total length of the battery to ESC lead that is the issue. Whether you extend by adding in an extension lead with plug and sockets is up to you.

Yes extend the motor leads is preferable. There is no practical limit other than eventually the wire resistance will reduce power a bit.

On the other hand I would not expect a modest extension of 75 mm on the battery to ESC wires to have much impact either.

Just make sure any long leads are kept as far as possible from the Rx aerial(s).

I have extended the ESC to motor wires to nearly 1000 mm without any problems and that was with a 60 A ESC and on 4s.

[Capt Kremen]

Whilst adhering to the principle of not extending ESC to battery in most models, I think there must be a modest degree of latitude. For example, buy several LiPo, each one from different suppliers/makes. Rarely if ever are the LiPo to LiPo plug wire length exactly the same. Therefore the ESC to LiPo lead length will vary depending on which brand of LiPo is used. Some Lipo wires are quite long, others barely long enough to flex!