2

I am delivering 1 kW power to a tethered drone. To minimize the wire weight and maximize maneuverability, I would like to keep the wire as thin and lightweight as possible.

The motors on my drone have a peak current of 70 A. However, a wire capable of supporting 70 A will be pretty thick.

The way around this is to step up the voltage on the ground to say 10 kV, thereby limiting the current to 0.1 A (to deliver 1 kW power). The drone will then have a transformer to step down the voltage along with an AC-DC converter.

From what I understand, the issue with high voltage is the possibility of a dielectric breakdown and arcing. Since I will be running two wires (live/ground) to the drone in a single cable, I will need to insulate the wires sufficiently to prevent breakdown.

How much insulation (say PVC) will I need to handle 10 kV?

Will it make a difference if I transmit DC instead of AC? Which one would be better?

Original question: Calculate the amount of insulation thickness needed for handling V volts across 2 copper wires.

SamGibson
  • 17,231
  • 5
  • 37
  • 58
Pragy Agarwal
  • 121
  • 2
  • 1
    Have you considered using standard 240v 3phase cabling? 2amp capable 240v cable should be available off the shelf. – user1937198 Oct 09 '22 at 15:24
  • 1
    Have you considered the weight of the step-down transformer aboard the drone? – Phil Freedenberg Oct 09 '22 at 15:30
  • @user1937198 If I understand correctly, 240 volts * 2 amps will only be 480 watt, which doesn't meet my requirements. – Pragy Agarwal Oct 09 '22 at 15:36
  • @PhilFreedenberg Not yet. I would like to get an estimate of the wire thickness and weight first. – Pragy Agarwal Oct 09 '22 at 15:36
  • 2
    I would suggest 300-400 V DC, then you can use standard twin mains-rated 3 A flexible cable, and a more or less standard SMPS down to your required 14 V, which will be the lightest way to transform 1 kW. You can't lighten the cable down to 100 mA, it would not be robust enough to survive handling, and the 10 kV would be hard work to transform. – Neil_UK Oct 09 '22 at 15:39
  • 3
    Even ignoring the wires, a power supply capable of stepping down 10 kV to something reasonable for your application will be *far* too bulky and heavy for a drone. Not to mention that things rated for handling 10 kV are *expensive*. – Hearth Oct 09 '22 at 15:43
  • @Neil_UK thanks for the suggestions. I will consider using a lower voltage. Is my question about the insulation thickness needed to handle 10KV answerable, or is it lacking some key information required to answer it? – Pragy Agarwal Oct 09 '22 at 15:47
  • @Hearth okay, thanks. Is the reason for high cost of 10KV rated things because of the insulation requirements? How much insulation thickness is needed to handle 10KV? – Pragy Agarwal Oct 09 '22 at 15:48
  • 2
    @PragyAgarwal "How much insulation thickness" is not really an answerable question. What are your standards? What goes wrong if the insulation fails (i.e., is the equipment destroyed, or could someone be killed)? Also, you suggest PVC, which isn't really an insulation used for high-voltage cables to my knowledge. Most high-voltage things seem to use PTFE, ETFE, FEP, or various silicone rubbers for insulation. – Hearth Oct 09 '22 at 15:51
  • @Hearth I have a cheap, lightweight transformer that steps up 12V DC to 10KV AC and produces arcs of up to 1/2 inch. Wouldn't the same circuit in reverse act as a step-down transformer? – Pragy Agarwal Oct 09 '22 at 15:51
  • 3
    The reason for the high cost is in large part due to insulation and creepage requirements, but also due to the fact that very few people need it, because the average engineer would avoid working with voltages that high whenever possible. And with good reason; not only is it expensive and difficult, it's *dangerous*. – Hearth Oct 09 '22 at 15:52
  • @Hearth If the insulation fails, the equipment is destroyed. No harm to any person. How does one define "insulation failure" though? Wouldn't the same definition be applicable to the "insulation thickness needed"? – Pragy Agarwal Oct 09 '22 at 15:53
  • 1
    @PragyAgarwal If it takes DC as input, it's not a transformer. What you have there is probably a flyback circuit for generating high voltages. They do not function in reverse and attempting to use it in reverse will destroy it. – Hearth Oct 09 '22 at 15:53
  • @Hearth oh, thanks for that info. What resources should I go through to be able to learn things like this? – Pragy Agarwal Oct 09 '22 at 15:54
  • 4
    @PragyAgarwal A bachelor's program in electrical engineering? – Hearth Oct 09 '22 at 15:56
  • 2
    @PragyAgarwal 240V single phase would only be 480W for 2 Amps. 240V 3 phase delta would be 1.4kW over 3 wires. – user1937198 Oct 09 '22 at 16:21
  • 4
    Switching from 2 to 3 wires to use a 3 phase supply doubles the amount of power you can transmit over the same weight of metal. In a scenario like this where cable weight is critical, you should seriously consider this option. And 240v 3 phase AC is a standard off the shelf option. – user1937198 Oct 09 '22 at 16:28
  • 2
    You say the peak current is 70A but if the peak durations are short so the average current is, for example, 10A then you can build a capacitor/super cap storage on the drone to handle the spikes and lighten the cable. Perhaps you can give an estimate of the joules in a peak spike? – ZSG Oct 09 '22 at 22:18
  • 1
    What length will the cable be? – Bruce Abbott Oct 10 '22 at 05:35
  • @BruceAbbott the drone will be at a 400m altitude, straight up from the ground station. – Pragy Agarwal Oct 10 '22 at 08:28

2 Answers2

3

The idea is not to touch high voltage with semiconductors.

Instead, generate three phases of low-voltage high current 20kHz square wave, using H bridges. Use at least 120VDC as the input supply voltage. That feeds three step-up high frequency transformers loaded for resonance tuned at 20kHz. You want to suppress the harmonics since they'll suffer heavy losses in the long wires.

Connect the secondary neutrals together, and feed the three phases only (no neutral) up the tether in a delta configuration. Those can be three separate hook-up wires twisted together - or even hanging free, for better air cooling. Target about 1A current per phase maximum, AWG 20 wire, 1kV or less AC voltage (that's 2.8kVp-p and it's what the insulation in a cable would have to withstand).

For AWG 20 or 0.5mm2 wire size, 20kHz is about the maximum to keep the conduction across the entire cross section of the wire and not suffer from skin effect losses.

On the top, use three single-phase transformers in a delta-star arrangement to step down the voltage, and three half H-bridge synchronous rectifiers to feed the DC link capacitor. The DC link is the DC source for the BLDC inverters etc.

You'll be wasting some energy in the wire, but that can be acceptable since the wire is hanging in free air. To keep the wire cool, the spool on the ground has to be a very long drum with forced air cooling through the inside of the drum, and only a single wire layer. You'll most likely need three drums in parallel, one for each phase.

AWG 20 hook up wire is less than 1kg per 100m, so for three phases you'll need say 2kg drone load capacity for each 100m of tether length.

Kuba hasn't forgotten Monica
  • 32,734
  • 1
  • 38
  • 103
2

I'll try to provide an answer, although it consists of questions and suggestions. The title of this question should probably be "What is the most efficient and practical means for delivering 1 kW power to a tethered drone". There should be a "sweet spot" combination of wire gauge, insulation, and voltage. Perhaps look at test leads typically rated 5 kV, and rather flexible. They often use silicone insulation, which is very flexible.

I agree with comments suggesting 3 phase power at high frequency, and it should be possible to use commonly available ferrite cores and bobbins to make a small three phase transformer. It might be easier to use two separate transformers.

If efficiency is not a major issue, it will probably be best to match the impedance of the cable to the load, so that a 2 kW ground station could deliver 1 kW to the drone, "wasting" 1 kW in the tether. In free air, with the heat distributed along a considerable length, heating should not cause a problem. I would suggest using 1 kV three phase which would result in 500 V phase to phase at the drone, with a current of about 1.2 amps per phase. Use the length of the tether to determine the wire size that will drop 500 V over its length. It might even be possible to use aluminum conductors, although perhaps with some steel strands (or maybe Kevlar) for strength.

You can probably find off-the-shelf 480 VAC (720 VDC) switching supplies or DC-DC converters rated 500 watts, and bypass the input rectifiers and H-bridge, using one phase of the three phase supply.

A quick search for "flexible thin high voltage silicone wire" came up with this supplier, and the smallest available is #30 AWG rated 3 kV, and 1 mm diameter. It has about 0.1 ohm/foot resistance, so for 1000 ft, it's 100 ohms, which is a drop of about 120 volts at 1.2 amps. That's just about 10% for a 1200 volt supply. Weight of the copper is about 0.0003 lb/ft, so for three wires a 1000 ft tether would weigh about 1 pound. You can figure the weight of the silicone insulation knowing the total volume and density of silicone.

Silicone has a density of about 100 lb/ft^3, or 2 Mg/m^3 and a 1 mm diameter wire 1000 m long (for 3 strands) is 78 cm^3 or 0.00275 cubic feet. So weight would be about 0.275 lb or 124 grams.

PStechPaul
  • 7,195
  • 1
  • 7
  • 23
  • Why would you want to waste a kW in the transmission? You can probably do a lot better than that without losing too much power. – Hearth Oct 09 '22 at 22:13
  • 2
    The primary advantage to inefficiency in transfer is allow you to use smaller conductors, with a higher voltage drop/unit length. Exactly where that sweet spot will lie will depend on the length of the tether, temperature rating of the insulation, and current required, vs voltage used. It may potentially make sense to operate the tether at up to 70C at peak load. – user1937198 Oct 09 '22 at 22:26