I guess many of you, if not all, heard of pure sinewave inverters which use conventional two-winding iron core transformers.

Their transformer is driven by a MOSFET bridge which, in turn, is driven by a sinewave PWM (for 60/50 Hz) whose frequency could be 16 KHz, for example.

This means that the transformer, with or without load, can also act as an effective low pass filter (perhaps with a very small ripple on its output voltage).

?

My first thought is that the transformer’s leakage inductances play the main element(s) in its function as LPF. But there must be another element to complete the LPF. Is it the stray capacitances, the core losses or both?

?

In vain, I have searched a linear model which could be used in simulating a transformer in such application.

I wonder if someone recalls that such a model could exist in the group’s archive so that I will redo my search looking for it.

?

Thank you.

Kerim

开云体育

The 'other element' is mostly the load resistance reflected into the primary circuit.

The leakage inductance, in conjunction with the resistive part of the load, constitutes a 1st-order LPF. The stray capacitances are in most cases negligible in this respect.

开云体育

Sounds like a ferroresonant power conditioner. Does that ring a bell? I don't think they are very efficient, but they are quite effective at producing sine waves from all kinds of nasty inputs.

Donald.

开云体育

Sounds like a ferroresonant power conditioner. Does that ring a bell? I don't think they are very efficient, but they are quite effective at producing sine waves from all kinds of nasty inputs.

OP specified "pure sinewave inverters which use conventional two-winding iron core transformers."

On Wed, Jul 5, 2023 at 11:48 AM, John Woodgate wrote:

The 'other element' is mostly the load resistance reflected into the primary circuit.

Your answer is logical.
But I was surprised when I tested a ready-made inverter, its HV output (of a conventional iron core transformer) was sinewave even without a capacitor as a load!

开云体育

That's because the leakage inductance is rather high.

On Wed, Jul 5, 2023 at 01:04 PM, Jerry Lee Marcel wrote:

That's because the leakage inductance is rather high.

So, in your opinion, what could be the other element in the transformer to complete its LPF response since there is no load at its HV output?

On Wed, Jul 5, 2023 at 01:45 PM, Alan Pearce wrote:

Don't forget that the iron core will have losses at the switching frequency, along with the inductance of the windings which will have a significant impedance to the high frequency switching. These will all have an affect on the output waveform, forcing it to be more sinusoidal.

Now the question is how we can translate this to an equivalent circuit.
None of the actual various equivalent circuits of a transformer seems being able to simulate its function as a stand-alone LPF!

开云体育

Losses as a resistor in parallels with the winding and less significantly paraitic capacitance.
Remember that losses increase largely with frequency, so harmonics and switching residuals are attenuated more than the fundamental. The basic linear model cannot represent that accurately. You have to consider one value of Rloss for fundamental, and other values for harmonivs and residuals.

开云体育

On 05/07/2023 13:06, Kerim via groups.io wrote:

Now the question is how we can translate this to an equivalent circuit.
None of the actual various equivalent circuits of a transformer seems being able to simulate its function as a stand-alone LPF!

Are you adding a shunt resistance (Rpar) to simulate core loss?

开云体育

I understand that, but the resonant winding is not visible to users. Ferroresonant xformers appear to be two-winding devices for all practical purposes.

Donald.

开云体育

Rpar = R(af+bf^2) near enough, for suitable values or R, a and b. You need data on the core material to determine them. af is eddy-current loss and bf^2 is hysteresis loss.

It seems all here agreed that the other element in LPF is Rpar(hi) which represent the various core losses at the PWM high frequency.
Now I have to find out a rather simple practical test to measure, even approximately, this Rpar(hi) of the unknown cores that the local retailers offer.

For instance, aren't the leakage inductances independent of frequency (at least for this application), so that they can be measured simply at 50 Hz?.

You may insist, but the? OP gives contrary evidence:

"pure sinewave inverters which use conventional two-winding iron core transformers. Their transformer is driven by a MOSFET bridge"... " driven by a sinewave PWM"

Just like the nominal inductance largely varies between 50Hz and 16kHz, so does the leakage. The magnetic permeability decreases significantly at HF, so te fraction that goes through the air becomes proportionally higher.

On Wed, Jul 5, 2023 at 03:59 PM, Jerry Lee Marcel wrote:

Just like the nominal inductance largely varies between 50Hz and 16kHz, so does the leakage. The magnetic permeability decreases significantly at HF, so te fraction that goes through the air becomes proportionally higher.

So, you mean that the impedance Z of the leakage inductance, is somehow proportional to F^2 (since Z=wL is proportional to F already).
For instance, to allow some flux leakage to exist, the transformer should have 2 separate bobbins, not one.
?

On Wed, Jul 5, 2023 at 02:30 PM, Donald H Locker wrote:

I understand that, but the resonant winding is not visible to users. Ferroresonant xformers appear to be two-winding devices for all practical purposes.

Yes, this is a practical solution in many applications.
In fact, my first thought was to look if there is a hidden third resonant winding. But I found out that the inverter transformer is indeed a normal one, though it has two bobbins instead of one which is usually used to reduce the flux leakage.

开云体育

On Wed, Jul 5, 2023 at 03:59 PM, Jerry Lee Marcel wrote:

Just like the nominal inductance largely varies between 50Hz and 16kHz, so does the leakage. The magnetic permeability decreases significantly at HF, so te fraction that goes through the air becomes proportionally higher.

So, you mean that the impedance Z of the leakage inductance, is somehow proportional to F^2 (since Z=wL is proportional to F already).

Not so much. Actually leakage inductance does not vary too much, but since the effective inductance of the winding decreases, the leakage factor increases.

For instance, to allow some flux leakage to exist, the transformer should have 2 separate bobbins, not one.

Leakage exists even when there is a single bobbin. The flux lines cannot be concentrated 100% in the core.

Two separate bobbins allow reducing mutual capacitance, but increase leakage.

?