Extending leads - battery or ESC?

[Steve Hargreaves - Moderator]

Interesting stuff Frank..... Seems even with the extra cap bank Castle Creations suggest a limit of 8" between battery & ESC....

[ericrw]

It seems thatI have to go back to the drawing board and get the speed controller nearer to the battery, Which will not be easy!! Does anyone know who sells Schulze 9 ESC in the UK ? This type of ESC allows upto 7" of additional extention lead.

[Steve Hargreaves - Moderator]

How much do you need to extend the wires Eric......I've used 3-4" extensions without problems.....I wouldn't think an ESC would work particularly hard in a Cub after all....

IMHO as long as your ESC is not right on the limit current wise I don't think you'll have a problem extending the leads a bit......

[Steve W-O]

Posted by Dave W on 03/04/2012 16:31:53:

I'm going to let another worm out of the can!

If it is considered not a good thing to lengthen the wires between the battery and the ESC because of possible damage to the ESC's input capacitors, what effect will putting a watt meter in this lead have?

This introduces more wire and some circuitry.

A good watttmeter (probably even a cheap one if it is recent) does not intruduce any circuitry, just additional conductor length.

I don't see any reason why you can not lengthen the wires between the battery and the ESC, as long as you do so with a heavy cable to keep the impedance down.

My Twinstar had the ESCs behind each motor, the wires are about 14" long, then another 6" on a Y piece to the battery connector. I didn't expect any problems, and I haven't had any.

[ericrw]

I've decided after reading both Steve's latest posts to go for the easy option. My flying is only gently circuits; so I don't think I should worry too much !!

[Frank Skilbeck]

I think it's also worth bearing mind that these recommendations are for ESCs working at their max capacity, so if used an ESC rated for say 8s and 100A but only used it for 4s and 60A then the capacitors will be undersized and possibly OK, but I must admit I don't know how to verify this.

[Steve Hargreaves - Moderator]

I think thats a good call Eric.....I do think we worry too much sometimes!!!

[Andrew767]

Steve...On a typical Twinstar set up, what size wire would you recommend?

Thanks

Andrew

[PatMc]

Posted by Steve Hargreaves - Moderator on 04/04/2012 09:44:20:

How much do you need to extend the wires Eric......I've used 3-4" extensions without problems.....I wouldn't think an ESC would work particularly hard in a Cub after all....

IMHO as long as your ESC is not right on the limit current wise I don't think you'll have a problem extending the leads a bit......

The ESC will be working harder at prolonged low throttle.

[Steve W-O]

Posted by Andrew767 on 04/04/2012 10:25:18:

Steve...On a typical Twinstar set up, what size wire would you recommend?

Thanks

Andrew

I'm sure I would have used 12g for the Y piece and the extensions.

More than what is needed, but it does work. You would probably be OK with 14g.

I think the actual reason would be that at the switching frequency, the positive and negative rails within the ESC need to be at the same potential (not DC, at the switching frequency) so a high rate ESC would need smaller capacitors to ensure this is so. This is more to do with decoupling than smoothing.

By having heavy wires the battery does the job to some extent, but if any problems were present, some capacitance across the feed of the ESC would help. You don't need overpriced "packs", just correctly rated capacitors from any supplier.

[PatMc]

Posted by Steve W-O on 04/04/2012 09:44:43:

A good watttmeter (probably even a cheap one if it is recent) does not intruduce any circuitry, just additional conductor length.

I don't see any reason why you can not lengthen the wires between the battery and the ESC, as long as you do so with a heavy cable to keep the impedance down.

My Twinstar had the ESCs behind each motor, the wires are about 14" long, then another 6" on a Y piece to the battery connector. I didn't expect any problems, and I haven't had any.

It's not extra circuitry that causes the problem, it's the cable length.

Heavy cable will only reduce resistance & won't affect inductive reactance. It's the inductive reactance part of the impedance that causes the voltage magnification (spikes).

Keeping the battery - ESC cable length down is a precaution against a known potential problem not a safeguard against definite failure. Damage to an ESC caused by voltage spikes can be cumulative & the effects not be apparent for ages, if at all.

IMO the second link in Frank's post gives a comprehensive explanation.

[Myron Beaumont]

Well,I did ask for enlightenment and very informative responses have I read. Just a thought (probably stupid) but would co-ax cables be of any advantageous use instead of ordinary wire if all earthed at any particular single point ?

[PatMc]

Myron, the problem is caused by the inductive reactance of the conductors not inductance from external sources.

[Myron Beaumont]

Of course --I'll get mi coat

[Martin Harris - Moderator]

How much value would there be in reducing the capacitance/induction effects of long conductors by separating them as widely as possible? If you're going to lengthen the conductors this seems an easy precaution to adopt...

[Myron Beaumont]

Martin

In a similar sort of way was the idea of screening preventing the inductive interaction of one wire to another .but I'm proved wrong of course

[ericrw]

I've left the room a while back !!!

[PatMc]

Posted by Martin Harris on 04/04/2012 12:47:56:

How much value would there be in reducing the capacitance/induction effects of long conductors by separating them as widely as possible? If you're going to lengthen the conductors this seems an easy precaution to adopt...

Separating the wires will reduce the capacitance but you don't want that, in fact it needs to be increased to counter the increased inductive reactance. Separation will have no effect on the inductive reactance.

[Steve W-O]

Posted by PatMc on 04/04/2012 14:07:28:

Posted by Martin Harris on 04/04/2012 12:47:56:

How much value would there be in reducing the capacitance/induction effects of long conductors by separating them as widely as possible? If you're going to lengthen the conductors this seems an easy precaution to adopt...

Separating the wires will reduce the capacitance but you don't want that, in fact it needs to be increased to counter the increased inductive reactance. Separation will have no effect on the inductive reactance.

It will in fact increase the problem.

[Peter Beeney]

I think I’ll still continue to along with the ‘bumpy supply lines’ ESC theory on this one, at least for the moment. I’m not sure that I’m totally convinced that it’s simply an self-inductance thing. When I think about it, I wonder just how much inductance a short straight piece of wire can support anyway? In general terms I think the short answer would be ‘Not very much!’ I’ve attempted some calculations, and my efforts say that 30 centimetres of 2.5 mm squared copper cable will yield 345 nano henrys*, that’s 345 by 10 to the minus 9, or 0.000000345 henrys, doubling that length to 60 centimetres equals 774 nano henrys. How big is that? Or conversely perhaps, how small? If we double the cross sectional area to 5 square mm a 30 centimetre long cable is reduced slightly to 325 nH. Hopefully I haven’t got this crossed up too much, but even if I have, this all still strikes me as being on the point of almost negligible.
Surely to make an inductance the wire has to form a coil, generally with a ferrous core. I thought that to suppress a voltage spike a capacitor would be a high(ish) voltage type and lower capacity. I’ve just looked at an ESC that I had to give back to another pilot, and this has two 35V 470 micro farad radial capacitors in parallel across the input. I’d guess that these have to be 35V because I think the next valve down is perhaps 16V, and this ESC, a 60A unit, is suitable for up to 6S, a possible maximum input of 25.2 volts. So why are they not higher voltage types? And 940 micro farads seems to be quite a lot of capacitance, too, more suited to be used for ripple smoothing?

Also as I’ve always seemed to understand it, any spike up in voltage is usually caused by the current switching off, when it stops flowing the fast decaying and therefore fast moving magnetic field thus creating a voltage, and because the magnetic field is moving fast this voltage is quite high; the voltage is proportional to the speed at which this field moves; and the amount of magnetic field; it is however, pretty brief! But, as I say, I’m not quite sure just how how much of a magnetic field we have here anyway, I cannot believe it’s really that great.
Taking a guess at a speculation, if you doubled the length of the leads you might want to add at least half the amount of capacitance again, but I think I’d be getting the old contact thermometer once more, just to see how hot things are getting. I’d be inclined to think that if it wasn’t that warm everything is perhaps in reasonable order; in other words, the capacitors are not being overworked.

I’d also try and keep all the cables close together, currents going in opposite directions cancel each other out, but they need to be next to each other. If I were to make a solenoid, that’s an an electro-magnet, all the turns around the pole piece would go the same way; if I then wanted to turn this into a non-magnetic device I could place an equal number of turns over it and pass a similar current in the opposite direction. A standard wire-wound resistor is simply a long piece of wire wound on a former, in one end and out the other, and the resistance is accurately measured, a non-inductive wire-wound is the same wire on the former doubled in a loop, the current then goes forward round and round and then comes back round and round again, out at the same end. Also twisting the motor wires together, if possible, is a good idea, to reduce the wire acting like an aerial and radiating any RF, but I’m not sure any of this is guaranteed to fully work in practise. Occasionally these unwanted side effects have in the past affected models in various ways, anyway.

Now I’ll go back to the flying, at least here I know by the end result, how many bits and pieces there are to take home, as to whether I’ve got it all wrong again or not.

* The Henry is the SI unit of inductance, named after Joseph Henry, most electrical units are invariably named after their discoverers, which I think is very appropriate, and is that inductance which can produce an emf of 1 volt when the current is changing rate at 1 ampere per second.

PB

[Myron Beaumont]

Peter

You've lost me !

[Biggles' Elder Brother - Moderator]

Posted by Myron Beaumont on 03/04/2012 14:17:59:

It might seem like a stupid point but being as electriciy travels at or nearly the speed of light .How come 3" will make a scrap of difference to either of the connections even though the current directional fllow is very rapid .? Havn't worked out the actual figures ,but at 186,000 miles per second & even allowing a few magnetic poles to reverse direction at extemely high rates ,I can't see that 3" could be a problem . Please enlighten me someone! .Being an IC man ,I have difficulty understanding how a cycle of burning fuel can happen in such a minute fraction of time but we all know that it does happen -About 300 times every second & of course electricity travels so much faster than a fuel mixture

One for Erfolg maybe or BEB ?

I hadn't seen this until now - just for the record its a common misconception that electricity flows at the speed of light - it doesn't, nothing like it in fact.

I remember years ago being taught how to calculate the electron "drift velocity", as it known, in a conductor given a particular resistivity and voltage etc. It quite modest actually. An analogy for the reason that it appears to flow instantaineously is the water in a hosepipe parallel - push water in one end, if the pipe is already full of water then water comes out the other end pretty well immdiately - its not the same water of course. It similar with electron flow in a conductor.

BEB

[Glenn Philbrick]

Now that we are all perfectly confused, about time we had some practical experiance. I have built and flown a Mossie and Lancaster both of which had the leads to the ESC extended. In the case of the Lanc the outer motors by about 10". I have experianced no problems whatso ever. Maybe I'm just lucky but I doubt it.

Good luck.

Sometimes too much knowledge is dangerous!!

[Brian Parker]

Is not the problem one of impedance?

We have a lot going on in the input side of the circuit.

Capacitive Reactance reduces with frequency whilst the equivalent series resistance (ESR) of the capacitor increases with frequency. By increasing the cable length we are now altering the inductive reactance.

Our circuit contains capacitance, inductance and resistance (ie impedance) all opposing current flow.

The capacitor is the weakest link.

Glen your luck could run out.

[John Privett]

Frank posted a link on the previous page to an RCGroups article by Ron van Sommeren. That article had a link to a post on another UK forum detailing the results of some testing - on behalf of an importer - to demonstrate the results of increasing the length of the battery leads. In this instance the importer was being sued by a customer who had ignored the maximum battery lead length statement in the instructions. The importer was seeking to demonstrate that extending the leads does cause problems - hence the instructions should not be ignored!

The posting is here and well worth reading. It shows that the back-emf 'spikes' do indeed increase significantly with increasing battery lead length. With the particular setup used in the tests, spikes remained less than 100v until the battery leads were extended to around 250-270mm. With 800mm leads, the spike was 430v and the ESC survived for a little over a minute before expiring in a puff of smoke. Another test with 1400mm leads (the original user who was suing the importer had used leads between 800 and 1400mm on a 4-motor plane) demonstrated spikes of over 700v and a dead ESC in less than 20 seconds.