Using Commom-Mode Chokes as (non-mains) Isolation Transformers?

[Jebnor]

Hello all, I have a thought  .  It might not be very good but here we go...

I have noticed that all the projects for my bench need (yet another) wall wart, or built in power supply.  I have a monster 12VAC transformer from a car battery charger. It puts out >30A @ 12VAC.

If I used that 12V as a power rail, could I use common mode chokes (such as [1]) as small isolation transformers for all my bench-only gadgets? My understanding is that it should be theoretically possible.  They will likely not be as efficient as proper mains transformers, but they are significantly cheaper.

[1] http://www.digikey.ca/product-detail/en/ELF-17N022A/PLK1180-ND/523281


Related theory questions:  How coupled are common mode chokes?  I'm aware that it varies according to the core type, but would they be as coupled as standard mains transformers?  How would inductance and/or impedance affect their coupling?

If it is one of  or could you explain a  little please.

[c4757p]

If you apply 12VAC at 60Hz (sorry, I'm American) to a 1.3mH "transformer primary", it will draw 12/(2*pi*1.3mH*60Hz) = 24.5 A (yep, 25 AMPS).

[Jebnor]

So... 

[cellularmitosis]

Just to clarify, 24 amps would be the current drawn if you shorted the secondary?  How would you calculate the "no-load" current?

[magic]

No, the current when you simply apply AC to the primary. AKA magnetizing current.

dI/dt = V / L
Integrate over one cycle half cycle, lol, of a 50Hz sinewave and you know why mains frequency transformers are huge.

[Kleinstein]

How good the coupling is depends on the design. Some CM chokes intentionally have reduced coupling to also provide some differential mode filtering - a little like the 50/60Hz transformers for microwave ovens.

The usual CM chokes are ferrite type cores and would not work well (only very low voltage before they saturate) with low frequency.  They may be still useful for a small DCDC converter running at some 50 kHz.

There are high performance CM chokes with noncrystalline core material that is actually also used for special mains transformers. But is would still be something like a 0.5 V / 0.5 V transformer for low power (e.g. 0.5 VA).

[TimNJ]

A few notes:

1. Although ferrite cores are usually used in power/RF applications above ~10KHz, they can be used at 50/60Hz, but in general does not make much sense.
2. Laminated silicon steel transformers/chokes can operate at flux densities ("Bmax") between 1.5-2T. Meanwhile, ferrite cores max out at about 0.3-0.5T.
3. High Bmax is preferred for low frequency applications because Bmax is related to the "energy" stored in the core's magnetic field, per cycle...or rather per half cycle. If you want to deliver 300W @ 60Hz, each cycle needs to contain a lot more energy when compared to a typical switch-mode converter's frequency of 100KHz. As flux density is related to the cross-sectional size of the core, a higher Bmax allows a smaller core without saturating.

If you had two identically sized cores, one ferrite, one laminated steel... the power throughput will be lower for the ferrite core (without saturation) @60Hz. So, you can get power through a common-mode choke, but you'd probably reach saturation earlier.

[Zero999]

If you apply 12VAC at 60Hz (sorry, I'm American) to a 1.3mH "transformer primary", it will draw 12/(2*pi*1.3mH*60Hz) = 24.5 A (yep, 25 AMPS).

So... 

The result will be an arc, fire, magic smoke and hopefully the circuit breaker tripping, before it gets too bad.

[T3sl4co1l]

Now this is some necromancy...

For those still curious, they are fine to use, at frequencies where they are suitable.  10-100kHz typically.  The downside is the leakage inductance is typically very high, so only a few watts can be transmitted, within limitations of typical output capacity (say 12V 100mA) and regulation (voltage drop with load).

A resonant control with feedback could deliver much more power (10s, 100s W..?) but is more complicated.  At this point you're more likely to shop for a proper pulse transformer (same sort of thing but with low leakage), SMPS transformer (flyback or forward type) or a fully integrated DC-DC module.

At 60Hz, not only will large currents be drawn, but the core will be in saturation much of the time, so DCR dominates -- it would smoke and burn up nice and fast, while transmitting very little power.

Tim

[TimNJ]

Holy... I did not even see the thread’s original date.