Pulse Current Through 22 Gauge Copper Wire 10us to 100us Pulse 10Hz Repeat @25C

It has a resistance of 52.8363Ω/km according to this. So here are two options:

• <...22 gauge...> has a cross section area of 0.326mm². As with all cabling, it depends upon the method of installation. One can use the principles embodied in BS7671 to determine the continuous withstand capability of the desired installation.
• "Nouse" suggests <=1A. So a wire 10^-4km long passing 1A would dissipate 10^-4W. Actual dissipation will be proportional to the mark-space ratio of the pulses.One might try a continuous current of 1A through a trial length to see if it can withstand it. One might then increase the current until it fails, and then turn it down a bit to arrive at an acceptable value.

BSEN7671 gives the standard adiabatic formula, units seconds t, amps I, square millimetres s. Times of 100microsec will be adiabatic. That equation is used for up to 5 seconds. Re-arranged..

sk = I√t

Table 54.6 gives k =228 for 30'C to 500'C rise, copper.

My STC "Reference Data for Radio Engineers" gives fusing current of 22 AWG copper 0.0253 inch diameter as 41.2 amps and formula I = kd^3/2 k = 0.244 copper, d inches I amps.

It's approximate - if you look at fusing current of wire in glass commercial fuses you will find fusing time has wide tolerance e.g. 10 to 300 ms @ 10 x rated.

The repeat rate depends on the cooling.

67model

would dissipate 10^-4W would dissipate 53mW

..."They discovered that electrodes covered with a molybedenum telluride catalyst showed an increase in the amount of hydrogen gas produced during the electrolysis when a specific pattern of high-current pulses was applied. By optimizing the pulses of current through the acidic electrolyte, they could reduce the amount of energy needed to make a given amount of hydrogen by nearly 50%."...

https://phys.org/news/2019-10-method-hydrogen-efficiently-capture-renewable.html

Skin effect for 0.025" diameter round copper is given as 11% increase over DC for 100 kHz sine & 2.6 times for 1 MHz.

Note that for square wave f, harmonics are 1/3 times amplitude 3f, 1/5 for 5f etc, so I² is 10% fundamental [3f] or less. For isolated 10 μs rectangular pulse, 10/second , most of energy is below 100 kHz.

With 0.01 to 0.1 % duty cycle, heating depends on your unstated current N.B. for copper, initial 25'C final 100'C, k = 114; 25 - 70 'C, k= 90.4

With k = 90.4, a single pulse of 2.9 kA/100 μs will heat to 70'C typical PVC continuous operation. A rate of 10/sec would be limited to about 100 amp for 70'C copper temperature.

Hi

Yeah, that calculation sounds great @100A even for single pulse.

I am using transformer wire with coating for 200C tolerance.

If single pulse raises temperature to 70C up at 30C room temperature then it may need seconds to come to 35C. This part requires Coulomb cooling in free air.

About frequency spectrum, I can use few uH series inductor to keep spectrum within 100kHz. LCR pulse shaping will be great idea. Keeping charge on capacitor and discharge through series inductor and resistor and finally into a copper wire in 10us or 100us pulse.

Frequency effective is under 1MHz or can be considered 100kHz for calculation purposes here.

Experimental wire is 22gauge transformer wire rated for 200C.

I will be switching to RG213 subsequently due to high voltages are involved and experiment to be repeated for 10kV discharge pulses. I can use unshielded free wires too with silicon insulation.

Very interesting information. Thanks.

My interest may go to Hydrogen generation as well as Pathogen killing by within fluid ark discharges.

Hi again,

This is the essential circuit of pulse drive for gas turbine ignitor [2 sparks/sec].

Capacitor C ~8μF with resistance Rc & inductance Lc is charged to 1.8 - 2 kV DC by a current limited source [it has to withstand near short-circuit, ]. A gas discharge tube GDT breaks down at 1.9 kV & connects to plug via inductor L with Resistance R_L.

With peak current near 1 kA & 100 microsec duration, it is not efficient. Without D, about 25% of energy in capacitor gets to ignitor, which is 10 ohm approx, once HV has got it conducting.

Once connected in parallel, L & C behave as a resonant circuit. With no losses, 0.5CV² initial energy in capacitor transfers to 0.5LI² peak energy in inductor cyclically. In real circuit, there is damped oscillation without D. Ideally D conducts after C reaches zero volts, the inductance of C means it is actually negative, taking possible energy from load. The inductance of C can be quite significant just because of its dimensions, current going in a loop round far end.

A "flywheel" diode D increases efficiency ~50% by circulating current in L & ignitor, with L/R_. exponential time constant. Replacing GDT with SCR reduces losses & overcomes limited life of GDT. An elaboration was to break down igniter plug with short HV pulse then connect plug to a lower volt, high current, source.

If you are basically blasting current into a piece of wire, which is only ~ 50 milliohm/metre for 22 AWG, it may be better to feed pulse into a transformer. The circuit below was used for capacitor discharge motor car ignition. The capacitor charges from an inverter source via the inductor, which was the ignition coil primary. That enables the SCR switch gate to be at near chassis potential and easier to drive.

N.B. the discharge capacitor C has to have low resistance & be suitable for high pulse currents to be reliable, foil & film types with extended foil are used. These circuits are nasty EMI sources, due to high di/dt, dV/dt & need care over enclosures, bonding, cabling & filtering of supply.

RG213 cable has 3 sq.mm centre core (7/.029"), 4.9 mm² braid [192/0.18 mm] & 60'C max polyethylene insulation. Comparing data for 2.5 sq.mm 3 core flexible cords, which have 25 amp rating & 19 mV/amp/m [loop] volt drop & diameter 10.0mm almost same, I conclude it has watts loss/metre of 12 watts continuous at 60'C insulation temperature & 30'C ambient.