Ideas for shunting current with minimal heat generation

[Needmorecoffee]

I would like some ideas/suggestions on how to shunt current in a PWM circuit. The load on the circuit is not sufficient to shunt the amount of current necessary.

I need to shunt about 1.0 additional amps during the "on" cycle of the pulse. The duty cycle could be 40 - 60% and the frequency is approx 5 to 20 Hz. I could use a five ohm, 10 watt resistor but I don't have the space or the heat capacity.

It is hard to get around the physics of heat disappation, but I would like to know your suggestions. I need to fool the driver circuit into thinking that the load is drawing the current.

I am considering a peltier heater, but I'm not sure how reliable these are.

Criteria:
-Minimal heat gain (<1 watt)
-Minimal size (e.g. I can't use a 10 watt resistor due to size and heat constraints)

 

[VEBill]

"I am considering a peltier heater..."

I hope you're not considering a Peltier cooler as a load (as an alternative to the power resistor) because 'it makes cold...' In the context of your post, that's what it seems like...

Back to your question. Can you find the current measuring device (such as the 0.1 ohm resistor), and insert the false signal there?

...or do you need to really sink the actual current? Why?

 

[Needmorecoffee]

Why not a peltier cooler? One of my constraints is that I don't want to generate any more heat.

I don't have access to the measuring device, so I really do need to sink actual current.

 

[waross]

Well a self cooling circuit using otherwise wasted energy is an elegant solution. As a solid state device the dependability should be good.
If you build it, please report back to us.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

Except Peltier units are about the most failure prone devices ever conceived by man. So the dependability is abysmal.

So how does this work? A Peltier cooler doesn't dispose of heat it just moves it an inch or two. They cost a lot. They are dimensionally constrained and they need a specific voltage. How's this help the project?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[IRstuff]

Duh, conservation of energy? A Peltier cooler creates "cool" by moving heat around, It cannot self cool its own entirety and must be heatsunk to achieve its design cooling point. (actually, the distance is usually less than about 1/8th of an inch, unless you're talking about a 2 or 3 stage TEC, or unless you include the heatsink.

You say you need to dump 1 amp at, apparently, 5V ("I could use a five ohm, 10 watt resistor"). That means you need to dump 5W, period, end of statement.

Is this for school?

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[Skogsgurra]

Not sure I understand what you mean.

To 'shunt' current means to divert, create an alternate/parallel path and that sort of things. Does that mean that you want to connect something that loads the output with an extra ampere?

In power supplies, 'shunt' immediately makes you think current measurement device. And that is where I get lost. Which one are you talking about?

You say that "I need to fool the driver circuit into thinking that the load is drawing the current." That could be done without any heat generated simply by adding an offset to the current actual value. That is what VBL says when mentioning a 0.1 ohm resistor (the shunt).

Re the Peltier cooler. Do you plan to use the cooler as the extra load needed and at the same time use the cooler to move heat from circuit to ambient? Killing to birds (or is it flies?). Interesting thought.

Not knowing what your circuit looks like and still not sure what you mean by 'shunt', it is difficult to have any ideas on how to do this.



Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[VEBill]

To the OP: If you're looking for a device that actually 'makes cold' (as opposed to moving heat), it'll be on the shelf next to unipolar magnets and other Star Trek items.

Peltier coolers can be used to provide local cooling (like on a CPU), but they simply pumped the heat from one side to the other. And the power dissipated on the hot side is precisely equal (in the steady state) to the heat removed from the cold side PLUS the electrical power used in the process.

They're certainly not a load that doesn't "generate any more heat". That should be obviouse due to the fundamental rules learned in high school Physics.

 

[Needmorecoffee]

Thanks.

 

[OperaHouse]

"The load on the circuit is not sufficient to shunt the amount of current necessary."

Why do you need to shunt excess current with a PWM? Is the load needed to provide IC power from a bias winding? There are ways around this without having to put a load on the output.