Silicone Wire Size relative to Volts / Amps / Watts ?

[Old Geezer]

I've been flying electric since the days of the NiMH powered Easy Pigeon [Ugh!] and Nigel Hawes' PRATT [a revelation!]. But my choice of wire from ESC to motor has always been based on the TLAR principal, poor old Acker Crist who taught me A Level Physics may have gone to his grave unsure how ever I conned myself a pass [2nd time round ]. I never did get the hang of conductivity:cross section of a conductor and other such mysteries.             So, is there a rule of thumb that allows you to, for example, decide the size of wire from a 3S 2200 / 100 Watts ESC combo to a 2836/08 spinning say a 10x4? I can't be too far outside the ball park 'cos I've never melted any wires, ESC:Magic Smoke - Yes, Over-propped a motor:Magic Smoke - Yes, but no connectors or wires cooked.           Advice anybody re Wire Dia to Watts and Volts [ and Amps ] too?   

 

 

Edited By Old Geezer on 03/09/2018 16:20:34

[Gary Manuel]

Wire diameter or more accurately the cross sectional area is what defines the maximum current that any particular conductor can carry.

There are plenty of tables available on the internet,

[Dave Hess]

There's a table in the link below. It looks about right for short lengths, like we use in aircraft, but for lengths of say a foot (300mm) or more, I'd go up a size. Wire size is based on amps. Volts, watts and cells don't matter. I would choose the wire based on the ESC current rating, so 20 amp ESC = 18swg, 40 amp ESC = 14swg.

**LINK**

I like 14swg silicone wire because its resistance is 1 mOhm per centimeter. That means that you can measure the current with a voltmeter. You add a couple of thin flying leads or other attachment points to one of the battery wires 10cm apart, then set the voltmeter to measure milivolts. The reading on the meter (say 30mV) is the current in amps (30A). That's really handy if you don't have a wattmeter. You can do the same with any size and length of wire, but then you have to calculate the current using the wire's resistivity x length, then divide the voltage measured by that.

Edited By Dave Hess on 03/09/2018 17:42:17

[Former Member]

[This posting has been removed]

[DaveyP]

Only at radio frequencies when the "Skin effect" comes into play

[Attilio Rausse]

Don't know if this helps, but I just match the wire that's already on the ESC.

[Former Member]

[This posting has been removed]

[Geoff S]

Posted by Dave Hess on 03/09/2018 17:31:44:

There's a table in the link below. It looks about right for short lengths, like we use in aircraft, but for lengths of say a foot (300mm) or more, I'd go up a size. Wire size is based on amps. Volts, watts and cells don't matter. I would choose the wire based on the ESC current rating, so 20 amp ESC = 18swg, 40 amp ESC = 14swg.

**LINK**

I like 14swg silicone wire because its resistance is 1 mOhm per centimeter. That means that you can measure the current with a voltmeter. You add a couple of thin flying leads or other attachment points to one of the battery wires 10cm apart, then set the voltmeter to measure milivolts. The reading on the meter (say 30mV) is the current in amps (30A). That's really handy if you don't have a wattmeter. You can do the same with any size and length of wire, but then you have to calculate the current using the wire's resistivity x length, then divide the voltage measured by that.

Edited By Dave Hess on 03/09/2018 17:42:17

That's a bit high. The table here gives a value of 5 milliohms/metre ie .05 milliohms for 1 cm. When you consider that a good LiPo cell has an effective series resistance (esr) of around 2 to 3 milliohms a connecting wire of just 3 cm would be doubling it.

For our purposes the only factor is current (The voltages we're concerned with a low) and temperature for the insulation (silicon is OK for soldering temperatures if it gets as high as that in use you've got serious problems!).

I used to wind RF coils when I was a teenager and learning the trade.  The wire used was multistrand Litz enamel coated wire because of the so-called skin effect.  Skin effect doesn't affect leads from the battery but may possibly the ones to the motor (not sure but not much).  The reason for multistrand wire is simply for flexibility.  In the sizes we need for 40 to 100 amps solid conductors would be like copper bars

Geoff

Edited By Geoff Sleath on 03/09/2018 22:58:37

[PatMc]

The current from ESC to motor is constantly switching between the wires. At any given time only two of the three are conducting, therefore the average current they are carrying is 2/3 of the amps measured in the battery - ESC leads.

[Old Geezer]

Thankyou Gentlemen - I've bookmarked all your suggestions - best of all, downloaded the page on the 4Max site and printed off a hard copy for the office wall. ( It'll be nice to go into the whys and wherefores when I have the time, and understand what I'm using, but the 4Max table is just soooo handy! )

[Nigel R]

"into the whys and wherefores "

for our purposes;

* the wire needs to be big enough to not cause a "significant" voltage drop

* the wire needs to be big enough to not get hot

For other uses, other factors come into play as mentioned above.

As Geoff says, we actually need to bend our wire, too, so it is multistrand, not a solid bus bar.

If you thought model wiring was bad, you should see the stuff that goes into a full size!

[The Wright Stuff]

Absolutely, and on the other hand, if the wire is too big, then it'll add unnecessary weight, cost and bulk, be difficult to solder, and hard to bend or route neatly...

In practice, one chooses the smallest cross section that comfortably accommodates (with a margin) the maximum current foreseen in the application.

My impression if I'm honest, is that there is very frequent overkill in that the wire dimensions chosen are considerably larger than they really need to be. The headroom could be made lower if there was a fuse in the circuit. The use of fuses in small electric models doesn't seem to be common - but that's an aside...

Edited By The Wright Stuff on 04/09/2018 10:50:32

[Peter Beeney]

O G,

My very simple take on this would be to suggest that if we double the length of a given piece of wire we double it’s resistance; if we double it’s cross sectional area we halve it’s resistance. So my equally simple rule of thumb might be - keep the wires as short and as fat as possible! But all within reason of course.

It’s always struck me that the model world tends to be fairly economical with the wire sizes, understandably so, flight battery and motor wiring seems to be generally ok; and probably you can considerably overrate short lengths for short times and get away with it too, the one that I’ve never been really convinced about is the long extension servo lead. For instance, if I move the rudder and elevator servos to the back of the fuz I cut the servo lead in half and solder in a homemade extension with heavier wire cut to the right length. It’s maybe the manufacturer’s lead is ok anyway, but from a safety point of view I’m willing to accept a small amount of extra ballast just to be reasonably sure the servos are going get enough volts and therefore enough amps to faultlessly carry on performing when I really really need them to.

Once upon a time a colleague took a very nice large scale model to a comp. The receiver supply was two 5 cell NiMH packs via an expensive all singing and dancing backup unit. Apparently he did a fair bit of flying during the day and then finally was asked if he would do a few circuits for the benefit of a local television crew that were there. This flight also became a bit prolonged and unfortunately before he could land the model crashed…

He first thought the batteries were at fault but when I checked them they proved ok; however, when I tried the volts drop across the whole supply circuit with the two packs fairly well discharged, as would have been his case, it was ranging down to nearly a volt when moving the servos, taking the voltage at the rx momentarily down below 4. In the air the control surface loads on the servo motors would have only made this situation worse. He was instantly convinced enough to remove all the paraphernalia and just revert to a single battery saying that he would change the pack halfway through a flying day.

This seems to me to be a classic case of two faults getting together at the same time, the excessive volts drop and maybe not quite enough battery capacity for unusually long flying times. But I would have also employed my very best favourite safety belt too, the on board rx voltage checker. I’m sure this would have prevented that last take off so that at least he could get a bit of charge back in the batteries first.

To my mind taking off without know how much charge is in the batteries is akin to taking off without knowing how much fuel is in the tank. But at least in the fuel tank case you would still have some gliding control…

Sorry for going a bit off topic there but I’m sure we can at least say that it’s just a little bit related…

PB

[Nigel R]

"the one that I’ve never been really convinced about is the long extension servo lead"

How big are your servos?

[The Wright Stuff]

Posted by Nigel R on 04/09/2018 14:10:37:

"the one that I’ve never been really convinced about is the long extension servo lead"

How big are your servos?

Indeed, and how long is your plane?

[Dave Hess]

Posted by Geoff Sleath on 03/09/2018 22:49:39:

That's a bit high. The table here gives a value of 5 milliohms/metre ie .05 milliohms for 1 cm. When you consider that a good LiPo cell has an effective series resistance (esr) of around 2 to 3 milliohms a connecting wire of just 3 cm would be doubling it.

For our purposes the only factor is current (The voltages we're concerned with a low) and temperature for the insulation (silicon is OK for soldering temperatures if it gets as high as that in use you've got serious problems!).

Is your table for solid wire rather than stranded? I'm going by standard Chinese silicone wire, where they probably cheat regarding the amount of copper in it. From memory, the official value is about 0.85 mOhms/cm, but it's always higher in practice, so 10cm is about right. It'll give an an accuracy within about 10%, which is as close as your average wattmeter.

Here's a comparison I did a few years ago:

**LINK**

[Peter Beeney]

Nigel R @ 04/09/2018 14:10:37 and The Wright Stuff @ 04/09/2018 14:17:21

To answer your questions both the models and the servos are pretty much standard size, although I’d most likely upgrade the servos to digital metal geared types. For the very same reason, really, it gives me more overall confidence in the control system.

As I said, I’m sure the standard type long extension lead is perfectly ok, it’s simply that I regard it as being a bit flimsy. Maybe there are heavy duty types around which would suit me be better, but by the time I’ve ordered and the postman’s delivered I can have the servos already working with my cobbled together bits and pieces.

I certainly don’t have any issues with anyone that uses these either; and I guess that’s going to be a very large percent of the modellers that needs to do so; it’s just my personal preference to fix my own…

PB

[The Wright Stuff]

Peter, for standard servos, with a max current of 1 or 2 amps, the voltage drop along a modest (let's say 1 metre) cable is probably negligible compared to the drop in voltage due to the battery becoming more discharged. So while it is true to say that the two faults compounded each other, the discharge dominates the resistance of the cables.

It's different for the motor cables, because the currents involved are 1-2 orders of magnitude greater...

[Nigel R]

Standard extensions won't hinder anything then.

If I remember right, the standard servo lead cable size is good for about 3A and the average standard (40g ish) analogue servo has a stall current of under 0.5A - way within. A whole airframe's worth of servos can happily dangle off a standard lead coming from the battery.

Digital servos need no different wires (stall current is the same).

Even the jumbo servos for monster planes only need to go up a couple of wire gauges to be well covered.

More of an issue is making sure the plugs don't become detached!

I found this old(ish) thread on this site, where some tests were done on stall current:

https://www.modelflying.co.uk/forums/postings.asp?th=44193&p=2

[Plummet]

If the wires warm up they are probably too thin.

If connections warm up they may need cleaning or replacing.

As the wires warm up their resistance will increase.

If the wires are well cooled then they can carry more current.

Thick insulation may help them to warm up.

Plummet

[Peter Beeney]

T W S, I have no issues at all with what you are saying, as I said, this is just my personal view and preference. I’m quite sure that I’ve not even advised anyone to consider replacing these leads either, and I’ve never heard of any related problems. If you were referring to the batteries in the scale model then although there was quite a lot wiring associated with the backup unit I’d have thought that this might have been the main culprit in itself.

If there were any standard diodes in there for instance, they alone are worth around a 0.6 - 07V drop, transistors the same. Schottky diode types are better, the forward voltage would be around 0.15 - 0.3V. The volts drop would have been there all the time, when the 5 cell battery was at 6 volts, the rx = 5 volts, (on load), when the battery voltage was at 5 volts, it would be virtually flat, the rx = 4 volts. At the end of the day, though, I suspect that his problem may well have been good old fashioned flat batteries anyway, no way of knowing. One of today’s telemetry systems would have been a godsend there, methinks!

If anyone has one of these systems in use it may well be doing a similar check just to see what the drop actually is nowadays.

Soldering the leads together fixes any connector issues. Years ago I was told by an ‘expert’ that my soldered joints were suspect and could fall apart at any moment. So I formulated my own test on said joints and I soon discovered that this was definitely not so. If fact, quite the opposite was the result. I’ve done this solder stuff a fair amount in the past, when I made my own 5 cell NiMH packs I soldered them to the switch harness. This meant they were never going to become unplugged in flight and with the added little bonus that the wires could be cut to length and therefore would help to keep the fuz tidier

I also made Nicad and NiMH flight packs and I used silver loaded solder on all the new battery work; the conductivity is a little better than standard lead/tin solder. Although I’m sure in the end it didn’t really make much difference anyway, I was just a glutton for punishment.

There will also be some volts drop across the motor cables too, but maybe insignificant amongst the rest of it. However, as I remarked in another post elsewhere recently, our power measurement is only a measure of the heat generated within the circuit so it’s very important to keep the resistance as low as possible anyway.

PB

[Nigel R]

"Years ago I was told by an ‘expert’ that my soldered joints were suspect and could fall apart at any moment."

Belt and braces:

https://en.wikipedia.org/wiki/Western_Union_splice

Bit fiddly on the small multi strand wires in our servo leads. But the best way.

Gets a bit bulky on an 8AWG motor wire!

[The Wright Stuff]

+1 for soldering, rather than replying on connectors.

[Peter Beeney]

I’d not seen that Splice before, Nigel, it’s very tasty, but as it happened my technique was \ is rather more simple and straight forward. My ‘test piece’ was done exactly as per a normal joint so for that I took a piece of 35 MHz rx aerial sized wire, cut it, trimmed about three sixteenths of an inch of insulation off each cut end, tinned the bare ends, laid them together overlapping for the length of the trim and just touched them with the hot iron. I very much suspect that many folks also do it exactly this way too…

To give it a serious workout I clamped it in the bench vice with the join hanging down underneath. Then I hung my power drill box on the end and gradually placed weights on top of that until something broke. I suppose I did this trick in all about 10 times to get a fail average, (I only think in belt and braces terms), and in each case the wire broke just where the solder stopped and the bright copper wire was just under the end of the insulation. The break was always close to the solder joint but it never was the actual soldered joint. After that I decided that I’d just keep soldering on regardless… No idea of the total weight involved unfortunately, surely a few lbs though, I was just interested to see the joint part; and indeed, in all honesty I thought that was exactly what was going to happen!

Maybe the ‘expert’ was a serious practitioner of dry soldered joints in his day…

Another type of join might be to form a U in each wire end and then hook squeeze and solder them together. Break that if you can…

Battery connections done in just the same way. Staggered cut by about half an inch, followed by the same little overlap solder jobby and finally a piece of silicone fuel tube slid over the whole caboodle. Only takes about half a day, or perhaps even a lot less than half an hour, to do. I thought it might make a convenient point to cut an old pack out or test the pack voltage for whatever reason but in the event I hardly ever needed to do either.

I’ve also soldered a couple of big motors to their ESCs in the past. Same way as I did the flight pack cells. A big flat sided chunk of copper iron heated hotly with a blow torch, then the two well tinned ends held against each side, the iron then slid out and the ends instantly butted against each other. That was about as elegant as I could get and I never had any problems. At least, not with that arrangement anyway.

Unlikely to do it again though, connectors are much easier…

PB

[Nigel R]

"Another type of join might be to form a U in each wire end and then hook squeeze and solder them together."

If I'm honest, that's as far as I usually go with stranded cable - plenty strong enough when soldered and a bit less bulky. It's also self-supporting while you solder (whereas two bits laid side by side are not). If it's a joint liable to be stressed, I would try and get in a twist or two, in the style of the lineman splice.

The lineman splice was originally intended to have mechanical integrity without solder on single core wire, I believe. Adding solder is/was icing the cake.