Help With AC Connector for Induction Heating Unit

[RyanRDB]

Hey everyone,

I'm working on an application for an induction heating circuit. I need to find a suitable three-connection plug / jack for a connection between the capacitors and the working coil.

There is one positive and two ground connections & the resonating frequency is about 225 khz. It also draws 5 to 6 amps, max. I was wondering if someone could tell me what the best-suited (& most compact) connector would be for this purpose. Here are a few preferences:
- Compact shape (1/2" diameter or less)
- $3 each or less (planning on ordering 50-100)
- 6 amps (ac) rating

I tried a 3.5mm stereo jack (3 connections) just for the hell of it, but I’m assuming the connections aren’t robust enough to handle the current. My next thought was possibly a mini xlr plug, but the ones I’ve seen are a little bulkier than I'd like (larger than .5" diameter & a little long), and somewhat expensive (more than $3). Any other ideas?

Thanks in advance.

[T3sl4co1l]

That doesn't sound like any induction heater I know... !?

[RyanRDB]

Lol, it's a zvs induction heater with a center-tapped work coil. The application I'm using it for requires a connector between the capacitors & the coil, but I'm having a hard time finding a suitable one.

The closest thing I've found so far (that could potentially work) is a mini xlr connector, as it has 3 pins. The only other 3-connection connector I can think of is a stereo plug/jack, but I don't think it can handle anywhere near the 5-6 amps that the unit draws.

This is my "pluggable" version that I experiment with:

[Richard Crowley]

Check out "Anderson PowerPoles".  Modular, available in 10 colors so you can color-code your pins, and available in a wide range of current capacity.
http://www.powerwerx.com/anderson-powerpoles/
This is just one of several convenient vendors who regularly deal with hobby/amateur radio people, etc.
You can join the poles together to make connectors of any number (and physical configuration) of pins.

[T3sl4co1l]

Oh, a toy homemade one... gotcha.

Probably want something bigger than the screw terminals shown, but Eurostyle pluggable terminal blocks come in a size or two above that.  Examples; http://www.digikey.com/product-search/en?pv89=86&pv89=196&pv89=56&pv89=94&FV=fff40016%2Cfff803be%2C1600005%2C478004d%2C4780056%2C4780057%2C478005a%2C478005b%2C4780060&mnonly=0&newproducts=0&ColumnSort=0&page=1&stock=1&quantity=0&ptm=0&fid=0&pageSize=25

There's also Molex 'bullet' type connectors (board or panel mount or inline), blade style (should be pretty good for high frequencies, using an alternate polarity pinout) and many more.

Tim

[bktemp]

6A sounds really low for induction heating. Did you measure the current thrue the coil or at the DC input?
Since most of the current circulates between the working coil and the resonant capacitors, the current at the driving circuit is much lower. If you mount the capacitors near the working coil, you can use much smaller connectors.

[RyanRDB]

Thanks for the input everyone -- I appreciate any suggestions / help I can get for this.

Check out "Anderson PowerPoles".  Modular, available in 10 colors so you can color-code your pins, and available in a wide range of current capacity.
http://www.powerwerx.com/anderson-powerpoles/
This is just one of several convenient vendors who regularly deal with hobby/amateur radio people, etc.
You can join the poles together to make connectors of any number (and physical configuration) of pins.

That's a really good suggestion -- seems like a super versatile connector. The only thing (that I should've also mentioned in my OP) is that the connector needs to be aesthetically pleasing, as I'm working on a prototype for a consumer appliance. Your suggestion would've been perfect if it was just for personal use.

Oh, a toy homemade one... gotcha.

Probably want something bigger than the screw terminals shown, but Eurostyle pluggable terminal blocks come in a size or two above that.  Examples; http://www.digikey.com/product-search/en?pv89=86&pv89=196&pv89=56&pv89=94&FV=fff40016%2Cfff803be%2C1600005%2C478004d%2C4780056%2C4780057%2C478005a%2C478005b%2C4780060&mnonly=0&newproducts=0&ColumnSort=0&page=1&stock=1&quantity=0&ptm=0&fid=0&pageSize=25

There's also Molex 'bullet' type connectors (board or panel mount or inline), blade style (should be pretty good for high frequencies, using an alternate polarity pinout) and many more.

Tim

Lol, yea the unit’s nice and simple/compact – I’m heating a relatively small work piece so it doesn’t need to be too powerful.

Thanks for the suggestion on the terminal blocks – the ones I’m using are rated for 10A. Does that seem inadequate to you? That link looks like an awesome resource btw.

When you say “should be pretty good for high frequencies,” do you mean that certain connector shapes/styles are better for higher frequencies? Sorry if these are stupid questions – I knew absolutely nothing about electricity/electronics before I started this project two months ago.

6A sounds really low for induction heating. Did you measure the current thrue the coil or at the DC input?
Since most of the current circulates between the working coil and the resonant capacitors, the current at the driving circuit is much lower.

Wow, I feel stupid -- I've just been going by the measurement I took, going from the PSU to the circuit. That's where I got 6 amps from.. Also, the PSU only supplies 6A, max. I haven't read anywhere that the amperage could be elevated, between the capacitors and the work coil. Probably should've tested that as well.

I just did two measurements for the amperage between the capacitors & coil:
1. The positive center-tap --- Got about the same result as my first test, between the PSU & the circuit

2. One end of the coil (ground) --- The multimeter flashed 15A & then went blank (the multimeter only goes up to 20A)... Do you suppose that means the current IS increased between the capacitors & coil (no idea why that would happen though..)

If you mount the capacitors near the working coil, you can use much smaller connectors.

By that do you mean: create a fixed connection between the capacitors & work coil, and use a connector between the capacitors & the rest of the circuit? I guess that would mean I could run smaller gauge wire also (no skin effect with DC)..

If I wasn't to mount the capacitors near the coil, and instead use a connector between the capacitors and coil, what amp rating would you suggest for the connectors?

Thanks again for all the suggestions / input.

[T3sl4co1l]

What coil dimensions (wire size, coil i.d., center-to-center length, number of turns)?  What capacitance?

A DMM will not even respond to anything over a few kHz; you're probably lucky it still works (or... maybe it doesn't).

AC current takes the path of least impedance, which is generally the closest path between two conductors carrying the current out and back.  An infinite parallel plate is closest at all points, so the current distributes evenly on both faces.  But finite parallel plates (like you get from a mirrored set of nice wide traces on either side of a PCB) have edges.  The magnetic field necessarily and implicitly generated by that current, shapes where that current flows.  The field is 'pinched' between the plates (because the plates are conductive, which forces magnetic field lines to the outer surface -- skin effect), and where it is able to spread out freely (at the edges), current is concentrated.  So, if you have parallel plates, current predominantly flows along the edges, with some fraction of the total flowing down the middle, face to face.

In general, you can observe that current prefers to flow along sharp edges, and facing surfaces.  So a square conductor carries more current in the corners than the surface (let alone in the body -- skin effect), whereas a round conductor carries most of the current in the facing side.

Another way of thinking of it: like currents repel (hence, pushing current out into the corners of a square conductor), opposite currents attract (drawing currents closer to the surfaces closer to the return conductor).

So, if you have connectors with round pins, if you use every other connection (per row, or in a checkerboard pattern if it's a multi-row connector), you'll get a reasonable distribution.  If you use all the connections on the one side for one polarity and all the others on the right for the return, most of the current will flow through the center pins (where the currents are closest -- path of least impedance).

If you have connectors with flat (blade) pins, a lot of current will flow along the edges, but also along the flat surfaces, so in addition to the generally larger build, the current capacity isn't too bad.  As long as the polarity alternates, of course.

Tim

[ConKbot]

Aesthetically pleasing, cheap, and relatively dumb end user proof?

http://www.parts-express.com/neutrik-nl4fx-speakon-spx-cable-connector-4p--092-190  Cable mount plug $3.45 each if you buy 100, good for 30A RMS. (Well the plugs are listed as 40 amp RMA, 50A peak, but the sockets are only good for 30)
http://www.parts-express.com/neutrik-nl4mp-speakon-connector-4-pole-panel-mount--092-052  Panel Mount Socket, $1.96 each if you buy 100.

Parts-Express is the cheapest place Ive found them, though I'd avoid ebay etc as there is lots of counterfeit ones. 

They are somewhat commonly used for high power amplifiers/audio equipment, but its not common except in that field, so you shouldnt feel too bad for re-purposing it.

[bktemp]

Measuring the current at the coil is not easy due to the high frequency and high current without affecting the circuit. But you can easily calculate the peak current: I=sqrt(C*U²/L)
I can't read the value on your picture, how many uF do you have put in parallel with the working coil?

[Richard Head]

Measuring the coil current is easy. Just wind yourself a CT using a small 10mm OD ferrite toroid. Place an appropriate burden resistor on the secondary and connect it to a scope.
The turns ratio and burden resistor value will determine the secondary voltage, but keep it below about 2v so the flux density in the core isn't too high. The galvanic isolation it provides will also ensure that the measured current waveform is really clean. The CT shouldn't decrease the Q of the resonant circuit at all.