Power of transformers, frequency + constant ferrite volume / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Power of transformers, frequency + constant ferrite volume

1 2 next

Move Thread

LAN_403

Patrick

Mon May 02 2011, 09:55PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Im looking at the math, and ferrite specs, also the TI/Unitrode PDF's. I need to resolve the confusion which swirls in my head.

If I have a H-bridge driven ferrite transformer (no gap) at fixed B and H (assume saturation limited, not thermal limited.) and constant ferrite volume, how would power vary as frequency varies? Am I even asking this question in the right way?

Stated concisely: What factors effect power throughput? Im after maximum power from minimum ferrite volume.

Ive tossed some ideas around, used the thought experiment method, and other math models, but I hope im not out-smarting myself.

James

Mon May 02 2011, 09:59PM

Registered Member #3610 Joined: Thu Jan 13 2011, 03:29AM
Location: Seattle, WA
Posts: 506

I can't claim to be an expert on this, but I know a bit and have gained more through experimentation.

The power transferred per cycle is limited by the amount of magnetic flux the core can hold, and increasing the frequency increases the amount of energy that can be transferred by increasing the number of cycles. Core losses also increase with frequency, as do the effects of parasitic reactance, usually manifesting as core heating. This means that the power a core can handle depends on duty cycle. A core of a given volume may be able to carry far more power for a few seconds/minutes than it can for hours on end.

Hopefully someone can elaborate further on this.

magnet18

Mon May 02 2011, 10:12PM

Registered Member #3766 Joined: Sun Mar 20 2011, 05:39AM
Location:

Posts: 624

these deal with core saturation, hope they help

Patrick

Mon May 02 2011, 10:20PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

James wrote ...

The power transferred per cycle is limited by the amount of magnetic flux the core can hold, and increasing the frequency increases the amount of energy that can be transferred by increasing the number of cycles.

My thoughts have led me to the same conclusion. Im glad to here this from someone else, so I know im not crazy.

James wrote ...

Core losses also increase with frequency, as do the effects of parasitic reactance, usually manifesting as core heating. This means that the power a core can handle depends on duty cycle. A core of a given volume may be able to carry far more power for a few seconds/minutes than it can for hours on end.

My cores will be under oil. But, thermal dissipation will be my next priority after power per volume.

magnet18 wrote ...

these deal with core saturation, hope they help

I have read these and the others many times, but my understanding of them is something I have to verify with others (hence the usefulness of this forum). Ive looked at that integral graph one many times.

magnet18

Mon May 02 2011, 10:45PM

Registered Member #3766 Joined: Sun Mar 20 2011, 05:39AM
Location:

Posts: 624

Well, when doing my experiment in the projects section, I collected more data and the results showed that if you decreased the duty cycle you could increase the frequency before the core went into saturation, but the maximum voltage would be the same.

You have a gun, it can shoot 1000 BB's a second, or 100 bullets a second, or 10 bricks a second. All give the same kinetic energy output. (metaphor courtesy of my aero-engineer father, who thinks electrons are freaky)(I know voltage isn't metaphorically equal to kinetic energy)

This would lead me to believe that if you're core isn't saturating yet, then you can increase the frequency to get more power, but once it goes into saturation, you'll need to modify something other than the frequency and duty cycle, be it volt turns, amp turns, core saturation flux density, or something.

But that's just a junior in high school playing with a flyback, and only 2 sets of data behind my thinking for that whole post, I could easily have gotten bad numbers.

James

Mon May 02 2011, 10:52PM

Registered Member #3610 Joined: Thu Jan 13 2011, 03:29AM
Location: Seattle, WA
Posts: 506

Think of a core as a bucket. You can only fill the bucket just so far before it overflows, but if you fill, carry, dump, repeat at a faster rate you will move more water. The faster you go though, the more water you will spill out in each step of the process, so there is a balance where a given size bucket will transfer the largest amount of water.

That's not a perfect analogy either but none really are.

Antonio

Mon May 02 2011, 11:51PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

For a core used in a true transformer, the power that can be processed depends essentially on the resistances and leakage inductances of the windings. The core shall be operated below its saturation limit, but this limit depends on the magnetization current (the primary current with the secondary winding open) only, and so on the voltage applied to the primary winding, on its inductance, and on the frequency of operation. If this limit is respected, the core does not limit the power, not mattering the extra primary current when the transformer is loaded since the added magnetic field is cancelled by the field generated by the secondary current. Larger cores are used for higher power because a larger magnetizing current can then be tolerated, with less turns in the windings allowing the use of heavier wire.

Patrick

Mon May 02 2011, 11:59PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Alot of what people are saying is affirming what I already believed, so this makes me feel more confident for my own up coming attempts.

Ash Small

Tue May 03 2011, 09:02AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Core material is a major factor here.

The limiting factor is efficiency (heat produced).

The cores I have (N27) are most efficient at 25 kHz, but will transform more power at 100 kHz, but with greater losses.

I'm planning to operate at 25 kHz, adding more cores as required, rather than increasing frequency.

This should both result in less core losses and less cooling required thereby obtaining greater efficiency than running at higher frequency.

Switching losses also increase with frequency.

Steve Conner

Wed May 04 2011, 11:42AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Yes, the important point is that all SMPS transformers are in fact thermal limited, to about one-third of the saturation flux density, or less as the frequency increases. The saturation limited case is of practically no commercial use, so there's no point in studying it too deeply.

Bigger SMPS always run at lower frequencies, because the transformer is just one part of a system, and a good design optimizes the whole system. So, they take into account the labour expense of assembling hundreds of little 100kHz MOSFET modules, as opposed to a few slow pizza box-sized IGBTs. The traction electronics in electric locomotives run at a few kHz and process about 1MW.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob