Thoughts on running 50/60 Hz SMPS on 400 Hz

[Dave]

I need to use some equipment in an aircraft which only has a 400 Hz
supply. My believe is that a switch mode power supply should not really
care too much if the supply frequency is higher. Anyone any experience
of using normal 50/60 Hz SMPS's on 400 Hz?

 

[Joerg]

I don't have any experience at 400Hz, but I'd be wary of any SMPS that
had power factor correction or automatic-voltage-switching built-in.

That, and the y-caps in there might throw you a curve. Bottomline the
manufacturer of the power supply would have to be asked.

Most likely they would not endorse use in an aircraft anyhow. Possibly
it would also require an STC blessed by the FAA -> $$$.

 

[Phil Allison]

"Dave"

I need to use some equipment in an aircraft which only has a 400 Hz supply.
My believe is that a switch mode power supply should not really care too
much if the supply frequency is higher. Anyone any experience of using
normal 50/60 Hz SMPS's on 400 Hz?

** So your idea of a " normal " SMPS is what ??

One that has no PFC circuit ?

That could become a big problem if not designed for a frequency 8 times
higher than usual.

Also, 400 Hz and its harmonics are far more likely to become injected into
places you don't want them by induction and capacitive coupling than 50 Hz.

Maybe you can rig up a temporary 400Hz (at 230 volt ?) supply - using an
audio power amp and a mains transformer (operating in reverse) to see what
happens with your particular gear.


....... Phil

 

[Phil Allison]

"Paul Mathews"

A method that works with many switchmode power supplies: Build a
rectifier/filter box to insert between the 400 Hz mains and the switcher.
In other words, supply the switcher with DC. This generally works even
with
switchers that include PFC. It doesn't require much capacitance to filter
400 Hz.

** DC mains supply was abandoned about 100 years ago - for darn good
reasons.

Even then it was only 240V DC not 350 V DC !!!

IMO - the fire safety hazards and risk of accidental misconnection of
transformer loads are too great to even contemplate this option.



...... Phil

 

[Dave]

Yes, there are several important safety considerations I didn't mention. I
had the idea that the OP was wanting to temporarily power some gear aboard
an aircraft, as opposed to providing some kind of product or fixed
installation.
Paul Mathews

You are correct. Also, the aircraft will not be flying.

 

[Phil Allison]

"Paul Mathews"

Yes, there are several important safety considerations I didn't mention. I
had the idea that the OP was wanting to temporarily power some gear aboard
an aircraft, as opposed to providing some kind of product or fixed
installation.

** That is a total red-herring.

Even building one 230AC to 350 volt DC supply adaptor is very dangerous -
since the only practical way to do it is to use domestic 230 volts AC
connectors at the output.

The fire risk is way too high.



...... Phil

 

[Tim Williams]

We used to run 60 Hz test equipment on 400 Hz power on the flight line.
Good stuff, Tektronix say , ran no problem. Aircraft selected 400 Hz to
reduce the amount of iron needed in inductors, transformers, generators,
and so forth. 60 Hz supplies (transformer, rectifier, filter caps, linear
requlators) have plenty of iron and capacitor size to work on 400 Hz.

-- "Too much", actually. The transformer in my Tek 475 has quite thin
laminations. It's rated for 50-400Hz, of course. Eddy current losses go up
with frequency.

Tim

 

[Phil Allison]

"Tim Williams"
"David Starr"

-- "Too much", actually. The transformer in my Tek 475 has quite thin
laminations. It's rated for 50-400Hz, of course. Eddy current losses go
up with frequency.

** But not for a given transformer and fixed primary voltage.

Raising the input frequency by a factor of 8 reduces the core magnetisation
by a similar ratio - so core losses remain much the same or even reduce.


....... Phil

 

[Don Klipstein]

-- "Too much", actually. The transformer in my Tek 475 has quite thin
laminations. It's rated for 50-400Hz, of course. Eddy current losses go up
with frequency.

Eddy current losses are proportional to square of frequency and square
of magnetic flux. (At frequencies low enough for this to not get
complicated much by skin effect, stray inductances, whatever)

However, volts per turn is proportionate with magnetic flux and
frequency. That means eddy current loss is proportional to square of
volts per turn, unaffected by frequency.

Hysteresis losses are actually improved by using a higher frequency.
Those are *roughly* proportionate with square of magnetic flux but only
linear with frequency - so with constant volts per turn, this loss is
roughly inverse proportional to frequency.

(Hysteresis losses have messy nonlinearities, and the above is an
oversimplification.)

The usual problem with operating 50/60 Hz iron core transformers at 400
Hz is that leakage inductance in the transformer sometimes causes output
voltage to be less (with load) at 400 Hz.

- Don Klipstein ([email protected])

 

[Tim Williams]

Hmm, I'll believe that, but then, if that's the case, why ARE the
laminations thinner in my scope's transformer? Why use thin laminations at
any frequency if it factors out?

Tim

 

[Phil Allison]

"Tim Williams"

** Fuckwit TOP POSTER !!!!!

"Don Klipstein"


Hmm, I'll believe that, but then, if that's the case, why ARE the
laminations thinner in my scope's transformer? Why use thin laminations
at any frequency if it factors out?

** Wanna think that last bit of drivel through again ??




..... Phil

 

[Don Klipstein]

Hmm, I'll believe that, but then, if that's the case, why ARE the
laminations thinner in my scope's transformer? Why use thin laminations at
any frequency if it factors out?

"Don Klipstein" <[email protected]>
wrote in news:[email protected]...

Making laminations thinner is done to reduce eddy current loss. With
that loss (and source of heat) being about the same at 50, 60 and 400 Hz,
making the laminations thinner makes things better at all of these
frequencies.

I suspect the laminations are thinner either as part of the transformer
being a better one, or to counter increased core loss by using lower cost
core material or skimping turns/volt. This is to make a cheap transformer
less bad, or to make a good one better. Less heat from the transformer
may be something sorely needed in something as compact as many modern
oscilloscopes.

Someone I have worked with tells me that a pretty good quality core
material with fairly reasonable (not bargain basement) cost is "29M6".
That comes in laminations thinner than those in the usual fare from low
bid contract manufacturers.

- Don Klipstein ([email protected])