Inductive power transfer + sensing resistance via inductance ?

[Liquidics]

I am attempting to pass current from a static part to a moving part via induction across a 1 mm air gap. Once that works I would like to use the current on the moving part to power an IR LED pointed at a mated IR sensor and then receive the resistance value from the IR sensor and pass the 'value' back to the static part via inductance over the 1 mm air gap.

The purpose of this is to be able to optically sense a physical interference from sensors mounted on a sliding shaft.

The current circuit has ~9v DC and regulated 5v DC source and runs off of a PIC microcontroller.

How do I generate an AC signal from a battery with a PIC to create the inductance? Or is there a way to create an inductance from a DC source?

How many windings of what thickness magnet wire would I need for the current to be able to jump the air gap and provide enough current for the LED, the IR sensor and the re-inductive 'output'?

 

[berkeman]

Sounds like a fun project. Could you explain in a bit more detail exactly what you want to make? There may be a simpler way to do it. Do you have an initial concept sketch of your project that you an post?

 

[Averagesupernova]

VCRs have a rotary transformer that transfers the signal out of the flying heads, which are mounted on a rotating drum, down onto the circuit board. There is NO amplification before the signal goes from the heads to the rotary transformer. I don't know how much loss there is but there isn't enough to lose the signal in the noise floor. So, knowing this, what you are attempting to do should be quite feasible.

You are wondering how to generate AC from the PIC? Just toggle an output X number of times through the program. Run that signal through a capacitor to block the DC and a buffer amplifier and into the transformer. Do a little research into how to wind the transformer to get the required inductance so the transformer does not end up using too much magnetizing current. Frequency plays a part here too. Typically the higher the frequency the more efficient the transformer is.

 

[Liquidics]

berkeman, aside from induction what other [efficient] methods could you transfer an electric signal to and from a fast moving non-conductive part without delicate flexing wires or wear-prone contact pads?

AvgSupernova, we are definitely on the same page- when I came up with this concept I actually re-disected a VCR head in an attempt to get ideas on how to make it work.

What exactly is the purpose of the buffer amplifier? Do you have schematics for it?

My PIC has a 20mhz XTAL, should I just juice the thing to get the highest possible frequency?

How could I power the LED/sensor with AC? Will I need a bridge rectifier on the moving part? Could I transmit back the resistance signal via inductance if I pass the AC through the collector of a transistor using the DC resistance as the base? Would that work?

 

[berkeman]

berkeman, aside from induction what other [efficient] methods could you transfer an electric signal to and from a fast moving non-conductive part without delicate flexing wires or wear-prone contact pads?

Depends on what you want to transfer, and if a data transfer is needed at all in the first place. That's why I asked for more details. It almost sounds from your OP like you just want to sense the position of the moving part. That is easiest to do just optically, with a pattern on the moving part, and a reflective sensor or two on the stationary part. That's how the nice battery-powered hand calipers work, for example.

If on the other hand you have a sensor on the moving part that you want to transfer data back from to the stationary part where your PIC is, then you are sort of on the right track that you need to transfer power across inductively, and then transfer data back across this air-core power transformer interface.

The VCR spinning head thing is very cool, but that is for a very specialized application.

 

[Liquidics]

Perhaps you are right Berkeman, maybe there might be an easier way to do this. The design function and need are as follows:

There is an actuating cylinder that repeatedly slides forward and back a basket-like mechanism which catches falling spherical products and then shifts them forward to another process.

In order to only actuate the cylinder when the sphere has landed in the basket, not before, there needs to be some way to sense their presence and it must be without physically contacting the sphere.

Reflective optics is an incredibly good solution, except that the environment could get wet and/or dusty, so that might make using reflective optics less reliable than a straight path with the IR LED and sensor mounted in a recess in the basket (the planned design if I can get the inductance working.)

 

[berkeman]

Yeah, if you can keep stuff off of the moving piece, that will help the overall design, IMO. If you can do the mechanical design of the basket to allow a through-shot of your IR sensing stuff, then you'll be able to tell when the ball has landed in the basket from outside the basket, and be able to acutuate the slide. Like, if the basket can have holes in opposite sides, and the base of the mechanism that holds the basket and slide can have the IR transmitter on one side of the basket, and the IR receiver on the other side, then the sphere falling into the basket will cut off the IR beam.

Other things that will increase the relibility of your IR sensor are to use an IR window at the receiver, and modulate the IR beam with some coding, and look for that coding in the receiver circuit. (Kind of like remote control IR beams are modulated, to help distinguish real signals from IR ambient noise.)