Invention of the day: The DC Transformer

[RoGeorge]

Tadaaamm!!!  Proudly announcing today the mirabulous invention of the DC Transformer!!! 



Made out of interleaved primary/secondary Peltier elements (TEC). 

One primary Peltier stacked with one secondary Peltier and so on, such that they alternate hot/cold surfaces, all in thermal contact.
Pros:
        - No mains ripple, so switching power supply noises, no singing coils or capacitors, all quiet, linear and analog, no RF noises
        - Can raise the DC input voltage entirely passive, by connecting the primary side elements in parallel and the secondary side elements in series
        - The Peltier elements can be connected in series/parallel, can supply negative DC, multiple isolated secondaries, etc.
        - The DC transforming ratio can be changed by rewiring the series/parallel connection of the Peltier elements
        - Eventually can transfer RF signals, too, to bidirectionally transfer data between the primary and the secondary sides of the DC transformer, for example when the DC transformer is used to power a galvanically DC isolated/floating/HV sensor
        - Same DC transformer may also be used without any DC power, as a thermal energy harvesting device
        - Smooth powering up response (might be either a pro or a con feature)
        - The Peltier elements doesn't necessarily need to be interleaved 1 to 1
   
Cons:
        - Transforming ratio might vary with temperature and/or with the loading
        - Power efficiency is very bad, might not matter for applications where only mW/uW/nW transfer is needed
        - Mostly for low power transfers, but it might withstand more power transfer if external radiators are added to extract the extra heat
        - Might require hydro-isolation against water condensation at the cold areas
        - Might require radiator fins to evacuate the heat excess

[Shay]

Cool (no pun intended)

But I think a powerful LED and a small solar panel/photodiode is better? You can even linearize the LED with some negative feedback

[RoGeorge]

The funny thing is, in theory a normal transformer can pass DC, too.  The following is a simulation showing a normal transformer when suddenly powered with DC in the primary.  Note how in the last plot the V(out) in the secondary is practically a constant DC voltage for minutes, though the required components values will be highly impractical. 


LTspice simulation from https://www.eevblog.com/forum/beginners/can-you-pass-dc-through-a-transformer-together-with-0-resistance/msg4955077/#msg4955077

[Phil1977]

Cool (no pun intended)

But I think a powerful LED and a small solar panel/photodiode is better? You can even linearize the LED with some negative feedback

Nitpicking: Light is just very high frequency AC...

[julian1]

How much does stacking raise the efficiency?

[TimFox]

In my experience with Peltier devices to cool temperature-sensitive circuits, the main engineering problem is removing the waste heat at the hot end.
Cooling Peltiers use more electrical power than what they transfer, and both the transferred and electrical power must go out the heat sink.
As the temperature of the heat sink increases, the efficiency goes down.
(They do work:  this is just the main engineering problem for many applications.)

[julian1]

Perhaps they could stack/interleave as a toroid assembly (just like a AC transformer) to reduce -area that creates the dissipation/losses at the side-ends.
eg. with heat conductive wedges to close the arc. half serious.

 

[Smokey]

Hmmm .  If you had an application that ran underwater or something, I bet you could get a fair bit of power output. 

[Benta]

Perhaps if you stacked Peltier- and Seebeck elements*, you might get up to 1% efficiency?

*: yes, they're almost the same, but optimized differently.

[porter]

The funny thing is, in theory a normal transformer can pass DC, too.  The following is a simulation showing a normal transformer when suddenly powered with DC in the primary.  Note how in the last plot the V(out) in the secondary is practically a constant DC voltage for minutes, though the required components values will be highly impractical. 


LTspice simulation from https://www.eevblog.com/forum/beginners/can-you-pass-dc-through-a-transformer-together-with-0-resistance/msg4955077/#msg4955077

kudos to Michael Faraday in 1831
"Have an iron ring made, iron round and 7/8 inches thick and the ring 6 inches in external diameter. Wound many coils of copper
wire round one half, the coils being separated by twine and calico. There were 3 lengths of wire each about 24
feet long and they could be connected as one length or used as separate lengths .... Will call this side of
the ring A. On the other side but separated by an interval was wound wire in two pieces together amounting to about 60 feet in
length, the direction being as with the former coils; this side call B ...
Connected the ends of one of the pieces on A side with battery; immediately a sensible effect on the
needle. It oscillated and settled at last in original position.."

[Kim Christensen]

Hmmm... A resistive element on one side of a peltier (Or thermopile) and a heatsink on the other would convert either AC or DC into DC.
(Some RF power meters work on this principle)

[smaultre]

Two DC motors?
There was another trick how to increase DC Voltage on battery powered, vacuum tube radio..

[5U4GB]

Nitpicking: Light is just very high frequency AC...

It shouldn't take an Einstein to see that it's a bit more complex than that.

[5U4GB]

Two DC motors?

Yeah, I think a DC-only rotary converter would be quite a bit more efficient.  However it doesn't have the coolness factor of this setup :-).

[Phil1977]

And a DC rotary converter would create lots of noise with its brushing.

This could be one of the few real low-noise galvanic isolators for DC...

[magic]

Not sure if there are any motors which produce ready to use smooth DC when used as a generator?

But I think a powerful LED and a small solar panel/photodiode is better? You can even linearize the LED with some negative feedback

Actually, does anyone know what sort of solar cell(s) it would take to get 0.5~1mA at a volt or two from an LED?

[Siwastaja]

And a DC rotary converter would create lots of noise with its brushing.

So let's do a brushless design - the typical semiconductor design driving the coils. Oh wait, now we can remove the mechanical coupling and just put the motor windings on the same magnetic core. We are getting there! 

[5U4GB]

And a DC rotary converter would create lots of noise with its brushing.

This could be one of the few real low-noise galvanic isolators for DC...

You could go the steampunk way and use the DC input to heat a boiler for closed-loop steam generation.  The traditional turbine as generator is going to be quite noisy, but there's all sorts of other ways to run a prime mover with steam.

[Phil1977]

But indeed there are not many ways to convert a rotation to DC without using any switches, diodes or sliding contacts.

[Smokey]

But indeed there are not many ways to convert a rotation to DC without using any switches, diodes or sliding contacts.

use the motor to drive a peltier junction!  Then hook that peltier up to another peltier then up to another motor!

[ArdWar]

Use the motor to drive a refrigeration cycle compressor. If you're lucky you may even get the whole peltier system "efficiency" to "double".

[tom66]

This is all silly.  Just use a DC antenna. 

(Patent pending)

[tszaboo]

Neat idea. If electricity requires changing magnetic field in a transformer, and DC doesn't have that, then let's go ahead and use a different phenomenon to generate electricity.

Cool (no pun intended)

But I think a powerful LED and a small solar panel/photodiode is better? You can even linearize the LED with some negative feedback

Nitpicking: Light is just very high frequency AC...

Good question, if LEDs are just small antennas that transmit radio waves on the frequency of light, than where is the oscillator? 

[tom66]

Good question, if LEDs are just small antennas that transmit radio waves on the frequency of light, than where is the oscillator? 

Electrons precisely falling through the band-gap in the semiconductor material, each producing a photon as they do so?

[coppice]

The first use we made of the low power benefits of CMOS was to power an 1802 MPU and supporting circuitry monitoring something at 55kV, powered from a photosensor attached to an optical fibre. A second fibre passed back the data stream of the results of the monitoring. I think the other end of the power fibre used a conventional lamp. This was in the late 70s, and LEDs were not that bright back then.