switching an inductor

I would like to switch a tapped inductor in a circuit carrying a 40 v pk 100hz rectified ac signal.

The tapped end will be at the output side so will be smoothed to a degree.

It needs to be electronic not relay.

Any ideas please , I am not sure if I can work out a suitable biasing method for this.

unnamed.jpg

Scr? Or fet if it is already rectified

You've drawn an air-cored inductor, do you really mean this?

What value, what current?

MarkT:
You've drawn an air-cored inductor, do you really mean this?

What value, what current?

No
Iron core 2 to 20 henries.

Current from 2 to 10 amps.

No idea where to find one these days as transformer kits have disappeared.

40 Volts, ten Amps.

  • Use the appropriate number of relays.

  • XY problem.

You could switch with properly rated and properly protected FETs. (flyback diodes, snubber networks)

So, what is the problem you are trying to solve?

Need a variable voltage that is electronically selectable?

This thingamabob puts out 40Vac and I need to reduce it to read different scales?

What?

-fab

I'm not getting this. Why are all three taps shorted together at the output ?
The second two have both ends shorted together. What is the point of that ?

raschemmel:
I'm not getting this. Why are all three taps shorted together at the output ?
The second two have both ends shorted together. What is the point of that ?

Perhaps to demonstrate that even inductors are capable of outputting the magic smoke when you shorten together their turns :astonished:

Whatever form of switch is used it must be "break-before-make" and as has already been suggested, relays really are the way to go

raschemmel:
I'm not getting this. Why are all three taps shorted together at the output ?
The second two have both ends shorted together. What is the point of that ?

It's badly drawn

Only one box is on at a time

jackrae:
Perhaps to demonstrate that even inductors are capable of outputting the magic smoke when you shorten together their turns :astonished:

Whatever form of switch is used it must be "break-before-make" and as has already been suggested, relays really are the way to go

Current method uses large lumps of iron and copper.
Also working voltage will be 10 X 40 v and large expensive caps are also required.
If i used relays which would be expensive to get a useful life they would need to be mercury whetted.
Main point of the exercise is to get weight and cost down.

fabelizer:
You could switch with properly rated and properly protected FETs. (flyback diodes, snubber networks)

So, what is the problem you are trying to solve?

Need a variable voltage that is electronically selectable?

This thingamabob puts out 40Vac and I need to reduce it to read different scales?

What?

-fab

Its a psu smoothing circuit.

It's being fed from an IMAG generator and I cannot use capacitive filtering as it kills the excitation current.

In one variant (wind) the input will be varying frequency (and voltage )and the circuit becomes very inefficient hence the desire to switch inductors.

Last time I looked at this it was unfeasibly expensive but I think that using an arduino its do able now

Then use heavy duty relays (aka contactors) with suitably rated contacts for DC. 10amps is chicken feed. You are looking for problems where they really don't exist.

Last time I looked relays were the only way.
If I have to use them its not worth doing as it increases cost and complexity.

fabelizer:
You could switch with properly rated and properly protected FETs. (flyback diodes, snubber networks)

So, what is the problem you are trying to solve?

-fab

The capacitor end voltage will vary by something like 20 %.

I cannot work out how to drive the gate voltage without exceeding the gate voltage at some point.

May be an easy solution but i'm just not seeing it atm.

I had hoped that a simple optocoupler would do it.

Automation Direct has 40A relays. (open style, no case)

Hmm.

All this mention of relays makes me think that folks around here are thinking
that the op should stop playing with matches and go boil an egg.

Boardburner2:
It's being fed from an IMAG generator and I cannot use capacitive filtering as it kills the excitation current.

In one variant (wind) the input will be varying frequency (and voltage )and the circuit becomes very inefficient hence the desire to switch inductors.

It took (just) a little research to figure out what on earth you were talking about.

I think you need to justify the part about "becoming very inefficient". In general, a choke input filter should be just fine with the maximum inductance, so please explain that. (Back to the XY problem.)

Clearly, the usual way to approach this is to rectify the generator output and feed it to a switchmode converter to deliver the desired output voltage I am not quite sure how you optimise the generator V/I curve however - is this what you are looking at?

Boardburner2:
All this mention of relays makes me think that folks around here are thinking that the op should stop playing with matches and go boil an egg.

Something like that.

Boardburner2:
No
Iron core 2 to 20 henries.

Current from 2 to 10 amps.

No idea where to find one these days as transformer kits have disappeared.

20H ??!!! That could be storing up to 1000 joules, so any naive switching scheme will
be toast. You need some sort of path for the 10A at all times or you'll cause
an explosion - that's enough stored energy to vaporize a relay's entire contacts I think.

However you've probably got 10's of volts AC between taps - perhaps massive MOV
or TVS diodes across each relay/MOSFET? They will have to take 100's of watts
during switchover, so MOSFET wins for that by fast switching time.

jackrae:
Perhaps to demonstrate that even inductors are capable of outputting the magic smoke when you shorten together their turns :astonished:

Whatever form of switch is used it must be "break-before-make" and as has already been suggested, relays really are the way to go

Explode before make with a 20H load I should think!

Seriously this size of inductor is extremely dangerous if mishanded, any kind of interruption
of 10A through 20H could lead to voltages far beyond what mains-rated components
can stand and at instantly lethal currents.

Has anyone ever seen a 20 H inductor?

Yes.
its the only reply I have been able to make as the previous 3 attempts keep getting timed out whil writing.

EDIT

I think they changed the laws of physics to make it though.

How big is a 20 H inductor (10A) ?