Flyback transformer power capability

[MRMILSTAR]

I know that it varies with the size of the flyback but does anyone know the "typical" amount of power that a flyback transformer can supply? I have two identical AC units which appear to be large (2.38" x 1" coil dimensions) from my limited experience with them. I am considering paralleling them just to be conservative. I can never find any specs on flybacks whether AC or DC.

This will be used to power the 14-stage CW multiplier that I am currently building. There are 28 capacitors with 1.7 nF each.

[Da_Stier]

A small warning at first, my answer might not be too helpful. 

Now that that is out of the way, I always try to calculate the output current by looking at the input power and measuring the output voltage. By assuming, I'm not saturating the transformer and an efficiency of 80%, I calculate the current.

A typical ZVS driver that I like to use can draw around 50W at 12V on the input.
By using the top assumptions, that would give me 4mA at 10kV. However the flyback gets pretty hot pretty quick under these circumstances.

I can't guarantee that any of this is accurate, however the current that is needed in a TV is pretty darn small, as far as I know. 

[davekni]

AC flyback transformers are rare, so I have no idea what yours were originally designed for.  I think 50W is about right for a typical DC flyback for a 19" TV, around 2mA at 25kV.  (Da_Stier:  4mA at 10kV may be running hot due to higher current through the windings and diodes.  Or, more likely, ferrite losses due to sine-wave excitation.  Typical flyback drive has less average voltage for a given peak output voltage.)

[Da_Stier]

(Da_Stier:  4mA at 10kV may be running hot due to higher current through the windings and diodes.  Or, more likely, ferrite losses due to sine-wave excitation.  Typical flyback drive has less average voltage for a given peak output voltage.)

Makes sense, I guess it might have been a better idea to test it at the voltage it is designed for, thanks.

I asked a friend who used to work  a lot on CRT TVs, according to him, a smaller TV (like 15" or something like that) uses around 1.3 to 1.8mA.

Another source might be something like this:
https://www.genvolt.com/application-information/5147347e0e033/CRT-Power-Supply
It's a piece of testgear / a powersupply for work on CRTs.
It provides 1.3mA at 20 to 28kV.

So I guess the 50W still seems pretty realistic.

[SteveN87]

This thread from 4HV might be useful:

https://4hv.org/e107_plugins/forum/forum_viewtopic.php?165666

With DC flybacks, I've found that 150W seems to be about the limit for ones without internal capacitors or screen/focus outputs.

[hammertone]

Steve,

The only limit to the power that you can pass with a transformer is set by the copper losses, the core doesn't care, with a few limitations: (Ferrite core)
The core must be kept from saturation, this means that deltaB (fluxvariation) should be below ~0.4T, but if it is driven at a frequency higher than 20kHz,  the limit to the flux variation has to be lowered because the core will heat up. At 100kHz you may well have to lower the fluxvariation to 0.1T. So below 20kHz the core is *saturation limited*, above 20kHz it is *core loss limited*.

Look at it this way: the transformer has an ideal transformer, with the primary parallelled by the magnetizing inductance, like this:



It is the magnetizing inductance that saturates the core if too much current is passed through it, or it heats the core up, if the frequency is high enough, so as long as there are enough turns on the primary coil to produce enough inductance, so that the flux stays below a safe limit, then the ideal transformer will happily transfer all the power you want.
But at some point, the coils will get hot, start to smoke, burst in flames, ultimately, fuses will blow.

I highly recommend you read this fine piece by Dr. Ray Ridley, he is an expert on magnetics :

https://ridleyengineering.com/design-center-ridley-engineering/39-magnetics/281-106-custom-transformers%E2%80%93one-design-equation.html
 
Hope this helps,
Cheers, Finn Hammer

[SteveN87]

OK. Got it.

[nix85]

This thread from 4HV might be useful:

https://4hv.org/e107_plugins/forum/forum_viewtopic.php?165666

With DC flybacks, I've found that 150W seems to be about the limit for ones without internal capacitors or screen/focus outputs.

Do you know some model without internal capacitors?

I got two different models both with caps and the output waveform is nothing like it should be (pulsed sawtooth with negative ramp).

One of these "AC flybacks" would also be perfectly fine since i got external diodes, as long as it has no internal caps.

BTW i just broke my flyback with hope of extracting the ferrite and secondary winding.

Like everyone says it is EXTREMELY hard to break it, i literally broke a concrete block into pieces on which i was hitting it before the plastic/ceramic whatever it is even began to crack. When i finally got to the secondary and peeled off some of the transparent plastic covering i realized these superthin wires are glued to it, so i have screwed it, and yet it is impossible to find the terminals without peeling it off, only way would be to keep the original external terminals but you can forget about that cause you will surely destroy the connections if you ever want to get to the secondary.

If you ever do this and i do not recommend it by any means, wear protective glasses.

[nix85]

This guy says...

If you compare a television and a monitor flyback transformer, you will observe that the monitor flyback have an internal high voltage film capacitor built into it. Generally, television flyback do not have this internal capacitor.

http://www.noahtec.com/discharging-flyback-transformer.htm

I had two TV flybacks BSC24 and FSV20A001 and they definitely had internal caps,  had to be discharged (i got few small shocks) and look at their waveforms.



I wish he was right, i want no caps in my flyback for my needs but as far as i see and as others say most if not all have 'em.

Again, does anyone know of ANY available flyback model that definitely does not have internal caps?

[SteveN87]

I have a Sony flyback (from a Trinitron portable TV - sorry, don't know the part number) which has no capacitor - it produces a hissing arc rather than sparks.

Have you connected the bottom end of the cap to ground? How are you driving the flyback (e.g. ZVS, bridge, true flyback)?

[nix85]

ZVS. I know the rise time is bit slow due to sinewave in the primary, but other than that, it should produce intermittent halfsine, yet it produces these weird large duty cycle waveforms which is purely due to internal cap.

Speaking of waveforms, it is interesting that if you feed pulsed DC to a normal transformer you will get a full wave AC meaning as current in the primary gets to the peak of the half-sinewave, voltage on the secondary has already reached the peak and fell back to zero. And as current in the primary falls from peak to the zero voltage on the secondary produces another opposite half-sine resulting in full sinewave output. Yet, when we feed full sine to the primary, of course, secondary follows it, no alterations. Just an interesting tip. There is something else even more interesting about transformers but i'll leave it for a future post.



So yea, ZVS is technically not driving it in flyback mode

Not exactly sure what you mean by "hissing arc rather than sparks" but sounds like low current output. I also get hissing thin arcs when i increase the voltage and lower the current, and fat, electric lighter type flames when i lower the voltage and increase the current.

Also not clear what you mean "bottom end of the cap to ground", "bottom" of the cap already is connected to negative side of the inductor. Here is a schematic below. All was connected properly. I been playing with 2 flybacks in series, it was clear to me that internal caps don't see the full voltage since from their perspective secondaries are in parallel. I also saw that adding a third flyback in series would change that and now they would see double the normal voltage. But morning haze, i forgot and connected 3....it worked for a short time, i got monster arcs (i don't care about monster arcs i see them as noisy waste of energy), more than 10cm, maybe 12-13cm (~5 inches), but then the internal caps dielectrics just broke and they just shorted in all 3 flybacks resulting in no output voltage. I had 10 external diodes each rated for 30kV as additional protection so i know it was not diodes that burned but caps.



In any case, i did not like those waveforms and so i am looking for capless FB.

[nix85]

As for another interesting thing about transformers i hinted at... Yea, turns ratio, eddy currents, hysteresis, magnetostriction, voltage drop across primary proportional to flux, that is, inductance, as load increases the counter flux primary flux drops and so does the voltage across the primary which makes more current flow as if a resistor appears in parallel with the primary inductance, additional current trying to bring the primary flux value to the original (no load) value which it never fully manages....all sweet and cool, basics. BUT have you every wondered how does the secondary "know" the number of turns, voltage and current in the primary? ALL that secondary sees is change of flux in the core. Now, this SAME flux in the core can be produced by many (theoretically infinite) combinations of turns, voltage and current in the primary.

[SteveN87]

Also not clear what you mean "bottom end of the cap to ground",

Like this one:

[nix85]

Also not clear what you mean "bottom end of the cap to ground",

Like this one:

Still not clear what you mean. Cap, at least in my fbs, is obviously connected to "ground" (meaningless term usually denoting 0/minus, reference point) internally cause, i only used two terminals, hv + and -, and still internal cap got charged, so it's clearly not dependent on me connecting it to "ground". If it was not internally connected to "ground", then it would obviously not get charged and would not affect the output waveform.

BTW i thank you for the reference, i searched for that tv flyback and found HR 6050 on Ebay, probably different model than you have but according to diagram it seems there are no caps in the HV winding, between pins 7 and H.V. I sent hrdiemen.com mail to confirm this, no response so far.

https://www.hrdiemen.com/reparation/flyback/scheme/6050

TO FORUM STAFF, it would be nice if you increase the editing time without the ugly label to say 10min (altho many forums have it much longer), as it is, i go back to correct some typo like 10sec after posting and then it says edited, ugly, unnecessary.

[SteveN87]

...Cap, at least in my fbs, is obviously connected to "ground"

So the answer to my question "Have you connected the bottom end of the cap to ground?" is "no, the manufacturer has already done it internally."

[nix85]

So the answer to my question is "no, the manufacturer has already done it internally."

Yes, obviously.

[nix85]

I been playing with 2 flybacks in series, it was clear to me that internal caps don't see the full voltage since from their perspective secondaries are in parallel. I also saw that adding a third flyback in series would change that and now they would see double the normal voltage.

I looked at this again and i was wrong above. Yea, when two flybacks are in series from internal cap's persepctive two secondaries are in parallel but their polarity is opposite.

Let's take one of the caps in the flyback on the left.

One secondary is trying to break through the cap downward (relatively speaking looking at the diagram) and the other one upward, so cap should see 0V.

When 3 flybacks are in series, it's own secondary is trying to break through it downward and other two in series upward, so overall it should see 1 normal voltage.

So that should not be the issue.

Diodes also should not be the issue, since each sees only voltage of it's own secondary, voltages being equally spread out. Well, one of my flybacks was smaller and burned first, i could attribute this to it's diodes but why did other two flybacks burn too, then.

It is true i used only 2 turns on the primary, voltage per secondary might've been as high as 70kV, which is on the border of what secondary insulation can take.

But fact is they burned when i put 3 in series. It's hard to be certain.

[klugesmith]

... BUT have you every wondered how does the secondary "know" the number of turns, voltage and current in the primary? ALL that secondary sees is change of flux in the core. Now, this SAME flux in the core can be produced by many (theoretically infinite) combinations of turns, voltage and current in the primary.

What is this voltage of which you speak?   The flux in transformer core is proportional to (primary turns * primary current) - (secondary turns * secondary current).   Same formula for flux in a flyback converter core.
We could extend your question to ask, how does primary circuit know about the secondary winding?  Start with powered transformer in no-load condition, with some current in magnetizing inductance.   By definition, the primary ampere-turns generate enough core flux F, and dF/dt, to induce voltage in primary turns enough to match the applied voltage.  (Impossible if core can't carry that much flux at saturation.)

Now connect a load that draws current from secondary.   Secondary ampere-turns tends to reduce core flux; primary ampere-turns needs to increase by the same amount.  In order to keep core flux dF/dt the same, so primary volts/turn continues to match the applied volts/turn.  Magnetizing current is a small fraction of rated primary current in normal transformers, where we get large magnitude primary and secondary ampere-turns mostly cancelling each other.   That's not true in flyback converters.

BTW, I don't mind the automatic "Last Edit" timestamp if I go back to fix or polish a posting, even if nobody saw the post yet.   It is not ugly, not a badge of shame.  It marks the writer as a perfectionist who chooses not to bother composing the post offline.

[MRMILSTAR]

Just about every post that I make is edited, usually multiple times. When did ensuring that something is correct or properly worded become ugly?

[nix85]

To start from the end, for me "edited" is ugly and unnecessary. Ok if it appeared after an hour or two from posting, but mere seconds later, when you just went back to fix some typo. Anyway, irrelevant.

All you did is basically repeat what i said and missed the point.

"What is this voltage of which you speak?

What voltage would it be, as i wrote, voltage drop across inductance of the primary, across it's inductive reactance determined by the formula I = (V-E)/Z in which I is current through the primary, V is voltage of the source, E voltage drop across the primary and Z it's inductive reactance.

The flux in transformer core is proportional to (primary turns * primary current) - (secondary turns * secondary current).   Same formula for flux in a flyback converter core."

We all know amper turns determines the flux. As for the deduction of flux i wrote that already...

"as load increases the counter flux primary flux drops"

This is lenz' law, flux from the secondary is reducing overall flux in the core which in turn makes the voltage across the primary inductance drop which in turn makes more current flow through it....

"as if a resistor appears in parallel with the primary inductance, additional current trying to bring the primary flux value to the original (no load) value which it never fully manages"

We could extend your question to ask, how does primary circuit know about the secondary winding?  Start with powered transformer in no-load condition, with some current in magnetizing inductance.   By definition, the primary ampere-turns generate enough core flux F, and dF/dt, to induce voltage in primary turns enough to match the applied voltage.  (Impossible if core can't carry that much flux at saturation.)

Now connect a load that draws current from secondary.   Secondary ampere-turns tends to reduce core flux; primary ampere-turns needs to increase by the same amount.  In order to keep core flux dF/dt the same, so primary volts/turn continues to match the applied volts/turn.  Magnetizing current is a small fraction of rated primary current in normal transformers, where we get large magnitude primary and secondary ampere-turns mostly cancelling each other.   That's not true in flyback converters.

All you say has been addressed already and all the key points of transformer operation clearly noted. All the stuff relating to flux reduction in the core due to lenz has already been addressed. Of course core must not saturate, of course magnetizing current is tiny compared to load current, of course flyback converter is not really a transformer but a coupled inductor...why even say such basic assumed things.

You totally missed the main point

Secondary has no way of knowing which combination of amper turns is producing the flux it sees. It can be literally any, say 16 amps 16 turns or 32 amps 8 turns or 64 amps 4 turns etc...all these and many many others will produce exactly the same flux. Obviously changing the number of turns changes inductive reactance and thus different voltages will be needed to "push" current through primary at particular frequency...

Secondary has no way, well, it clearly HAS a way, but no way we can determine, of knowing which among theoretically infinite combinations it is. We may get into the A-field and Magnetic vector potential but this does not explain the phenomena either. It implies some kind of INTELLIGENCE underlying this usually taken-for-granted phenomena.

[Uspring]

Secondary has no way of knowing which combination of amper turns is producing the flux it sees. It can be literally any, say 16 amps 16 turns or 32 amps 8 turns or 64 amps 4 turns etc...all these and many many others will produce exactly the same flux. Obviously changing the number of turns changes inductive reactance and thus different voltages will be needed to "push" current through primary at particular frequency...

Yes.

Secondary has no way, well, it clearly HAS a way, but no way we can determine, of knowing which among theoretically infinite combinations it is. We may get into the A-field and Magnetic vector potential but this does not explain the phenomena either. It implies some kind of INTELLIGENCE underlying this usually taken-for-granted phenomena.

You seem to imply, that the secondary knows, which of the many possible primary configurations produces the flux. How, e.g., does a secondary voltage measurement tell you details about the primary circuit? Except, of course, the total flux time derivative it generates in the secondary.

[nix85]

Yes.

You seem to imply, that the secondary knows, which of the many possible primary configurations produces the flux. How, e.g., does a secondary voltage measurement tell you details about the primary circuit? Except, of course, the total flux time derivative it generates in the secondary.

I don't only imply, obviously turns ratio "law" is respected and output voltage is predictable.

Secondary voltage does not really tell us anything about the primary circuit.

ALL it sees is certain flux change, what is generating it, as far as we can tell, it has absolutely no way of "knowing".

[Uspring]

I don't only imply, obviously turns ratio "law" is respected and output voltage is predictable.

That is, because

a) secondary voltage is proportional to the flux rate of change and proportional to the secondary turn count
and
b) the flux rate of change is proportional to the primary voltage and inversely proportional to the primary turn count.

Together this leads to the turns ratio law. The only thing, that the secondary "knows" about, is from a) . But we have the turns ratio law anyway.

[nix85]

That is, because

No it's not.

a) secondary voltage is proportional to the flux rate of change

I been saying from first post on this subject that voltage across the primary is proportional to rate of change of flux, this is Faraday's law we all know

V=--N*dΦ/d

Question is, however, is it violated in the secondary. I'll address this below.

and proportional to the secondary turn count

Obviously, turns ratio is respected, as said before.

and
b) the flux rate of change is proportional to the primary voltage and inversely proportional to the primary turn count.

Rate of change of flux is proportional to frequency of primary voltage and consequently current and of course number of turns. God knows why you wrote inversely proportional to primary turns, as if more turns means less flux. More turns means more inductive reactance and thus more voltage is needed to push the same current at the same frequency, but that is another thing.

Together this leads to the turns ratio law.

No it does not.

The only thing, that the secondary "knows" about, is from a) . But we have the turns ratio law anyway.

Secondary (as far as we can tell) knows nothing but flux in the core. It has no idea if there is a primary coil there at all, not to mention how many turns it has, what is the voltage and current across/through it, the flux it sees may as well be produced by spinning a permanent magnet inside a gap in the core. And if there is a primary there, again, many different combinations of turns, voltage and current can produce the same flux, secondary has no idea which combination it is, and yet it knows.

Back to the Faraday's "law".

V=--N*dΦ/d

According to the "law" particular rate of change of flux will always produce the same voltage across N number of turns.

But we know that secondary is respecting the turns ratio, even if change of flux is the same, it will generate different voltage and current clearly violating the "law".

If Faraday's "law" was valid for the secondary, it would mean there is only one combination of voltage, current and number of turns that can produce particular rate of change of flux. But that is not the case.

[Uspring]

Both of my statements above are versions of Faradays law. Explicitly:

a) secondary voltage is proportional to the flux rate of change and proportional to the secondary turn count
i.e.:
Vsec ~ Nsec*dΦ/dt

b) the flux rate of change is proportional to the primary voltage and inversely proportional to the primary turn count
i.e.:
dΦ/dt ~ Vpri/Npri
Note that multiplying this equation by Npri on both sides will lead to the familiar version of Faradays law. In this case for the primary coil.

Rate of change of flux is proportional to frequency of primary voltage and consequently current and of course number of turns. God knows why you wrote inversely proportional to primary turns, as if more turns means less flux.

What I meant to say is, that for a given primary voltage, an increase in turn count will lead to less flux. That is because the current will drop with rising turn count, due to the larger primary inductance. In equations:

Ipri ~ Vpri/Lpri ~ Vpri/Npri^2   due to the inductance being proportional to the square of turn number
Φ ~ Ipri*Npri ~ Vpri/Npri      using Ipri from above

Again, Npri appears in the denominator here as it must in order not to violate Faradays law.

More turns means more inductive reactance and thus more voltage is needed to push the same current at the same frequency, but that is another thing.

The turns ratio law relates primary voltage, secondary voltage and turns ratio. So the effect of more voltage needed to push the same current cannot be disregarded.

[klugesmith]

b) the flux rate of change is proportional to the primary voltage and inversely proportional to the primary turn count.

Rate of change of flux is proportional to frequency of primary voltage and consequently current and of course number of turns. God knows why you wrote inversely proportional to primary turns, as if more turns means less flux. More turns means more inductive reactance and thus more voltage is needed to push the same current at the same frequency, but that is another thing.

Should we stop feeding the troll?    Uspring is right: for given primary voltage, more turns means less volts per turn, so less flux, and less voltage in secondary.  Also, rate of change of flux is proportional to primary voltage and _not_ proportional to frequency.
We could forget about the "inductance"view and say: Core flux will be whatever it needs to be, so the induced voltage in primary winding matches the applied voltage.

We could further avoid cause-and-effect arguments by using the coupled inductor model.
L_pri = 1 henry.  L_sec = 100 henries.  K (L1,L2) = 0.99.  When an AC voltage is applied to L_pri, L_sec (even if unintelligent) has a voltage greater by a factor of about 10.  It's just the solution of circuit equations, not much fancier than a resistive voltage divider.

[nix85]

Both of my statements above are versions of Faradays law...

Ok, the way you formulated b) was not obvious that you were just stating the Faraday's law rearranged.

What I meant to say is, that for a given primary voltage, an increase in turn count will lead to less flux. That is because the current will drop with rising turn count, due to the larger primary inductance.....

Well, i been saying exactly the same thing. Higher inductance > higher inductive reactance > smaller current/flux, that is, unless voltage is increased to compensate for additional impedance. To remind of few basic formulas

Inductive reactance

XL= 2πfL

Alternating current flowing through inductor is applied voltage / inductive reactance

I= V/XL

​In equations:

Ipri ~ Vpri/Lpri ~ Vpri/Npri^2   due to the inductance being proportional to the square of turn number
Φ ~ Ipri*Npri ~ Vpri/Npri      using Ipri from above

Again, Npri appears in the denominator here as it must in order not to violate Faradays law.

I know very well that inductance is proportional to turns squared just like it is inversely proportional to coil length, as well as Faraday's law and turns ratio law, these are basics.

The turns ratio law relates primary voltage, secondary voltage and turns ratio. So the effect of more voltage needed to push the same current cannot be disregarded.

We all know what law of turns ratio relates, again, turns ratio is respected. As for more voltage to push.. i brought that up first, i'm sure not disregarding it but that is not the main point.

Again main point is many different volt amp turn combinations can produce the same flux change and secondary has no way of differentiating.

Unless you maybe want to claim there is only one volt amp turn combination that can produce any particular change of flux.

Should we stop feeding the troll?

I may call you a fool, but i'm gonna remain polite. I am not a troll.

Uspring is right: for given primary voltage, more turns means less volts per turn, so less flux, and less voltage in secondary.

No one is denying that more turns on the primary means smaller voltage in the secondary, again, turns ratio is respected, obviously.

Also, rate of change of flux is proportional to primary voltage and _not_ proportional to frequency.

In fact, i brought that up first when i mentioned need for higher voltage to "push" same amount of current through an inductor at higher frequency, basically the same thing in other words...so, i'll call semantics on this one.

We could forget about the "inductance"view and say: Core flux will be whatever it needs to be, so the induced voltage in primary winding matches the applied voltage.

We could further avoid cause-and-effect arguments by using the coupled inductor model.
L_pri = 1 henry.  L_sec = 100 henries.  K (L1,L2) = 0.99.  When an AC voltage is applied to L_pri, L_sec (even if unintelligent) has a voltage greater by a factor of about 10.  It's just the solution of circuit equations, not much fancier than a resistive voltage divider.

You are also missing the main point....many different volt amp turn combinations can produce the same flux change and secondary has no way of differentiating which one it is.

[nix85]

I may also remind of Farraday's law for moving conductor

where V is voltage B magnetic field L length of a conductor/coil and v relative velocity of a magnet and a coil

V = BLv

If we were to draw a parallel, we may say that there are many combinations of these 3 values that will generate the same voltage. It is not a perfect analogy cause situation is very different, but somewhat portrays the point.

[klugesmith]

OK, I apologize for saying the T word.  And we don't want readers to start fetching popcorn.

>>You are also missing the main point....many different volt amp turn combinations can produce the same flux change and secondary has no way of differentiating which one it is.

Agreed. So what? Who says a secondary coil does differentiate, or needs to differentiate, between primary cases that have same outcome in secondary circuit?   I am not seeing an unexplained mystery.

Secondary "sends" signal to primary, just like primary "sends" signal to secondary, by having ampere turns that magnetize the shared core.  Under normal loading conditions the core flux from secondary amp-turns is about the same magnitude as that from primary amp-turns, with opposite sign.

[nix85]

What do you mean so what. If there are multiple variables on the primary side that can result in same flux change and flux change is all that secondary sees....do i need to draw it that secondary has no way of telling which combination is generating the input, yet, it behaves as if it does.

You must either disagree with me that multiple combinations of volt amp turns can produce the same change of flux in the primary or if you agree, you must deny the validity of Faraday's law, you can NOT have it both ways.

Faraday's law demands that for certain rate of change of flux same number of turns will always result in same voltage across those turns.

Also, i know well "Under normal loading conditions the core flux from secondary amp-turns is about the same magnitude as that from primary amp-turns, with opposite sign." and that max-load flux is slightly lesser than no-load flux etc.

[nix85]

Here are few more basic formulas

flux density = amper x turns x core permeability x core area / m² (T)

F = ILxB force on a conductor in a magnetic field - laplace
as load increases, current in the conductor must increase to balance the forces: I = F/BL

v = L(di/dt) BACKEMF from an inductor

t = L/R inductor time constant, after ~5t (transient time) current reaches 99.5%

etc

[klugesmith]

>>You must either disagree with me that multiple combinations of volt amp turns can produce the same change of flux in the primary or if you agree, you must deny the validity of Faraday's law, you can NOT have it both ways.

I agree with the first part, but don't see Faraday's law breaking down.   Would you mind reviewing the following example, designed with "flyback-like" size and frequency, and point out what I'm missing?

The core has effective length 10 cm, area 1 cm^2, permeability 796.   To make coil inductances 1 microhenry per turn squared.
Let's operate at 15.9 kHz, for radian frequency of 10^5.

Our first primary coil has 10 turns (L = 100 uH), intended to operate at 10 V (peak) with magnetizing current of 1 A (peak).  MMF is 10 amp-turns, H is 100 A/m, Bmax is 0.1 T, peak flux is 10 uWb, dF/dt (peak) = 1 Wb/s, inducing 1 volt/turn.  The unit "webers per second" is dimensionally identical to the volt unit.

Our alternate primary coil has 5 turns (L=25 uH), to operate at 5 V (peak) with magnetizing current of 2 A (peak).   MMF is 10 amp-turns, ..., dF/dt is 1 Wb/s as before.

Our secondary has 40 turns (L = 1600 uH), and will have peak voltage of 40 V no matter which primary coil is used.   Each turn of the secondary goes around 1 Wb/s of flux change rate, so Faraday says 1 volt per turn.

If a secondary load draws 1A, that would increase the primary current by 4 A (original coil) or 8 A (alternate coil).   Net MMF (ampere turns) and net dF/dt are about the same as the unloaded case, no matter which primary coil is used.

What suggests to you that Faraday's law isn't valid here?  If the example is missing a pathological case, what can we vary to expose the failure?

Note 1: normally the secondary load current and corresponding change in primary current are in phase with the voltage, while the magnetizing currents have 90 degree phase lag wrt the voltage.

[nix85]

Our secondary has 40 turns (L = 1600 uH), and will have peak voltage of 40 V no matter which primary coil is used....

Now that you gave a practical example, i gotta say my bad. I knew this, i even said it from post one that voltage across the primary is proportional to it's inductance, that is, flux, so turns ratio is automatically respected.

Note 1: normally the secondary load current and corresponding change in primary current are in phase with the voltage, while the magnetizing currents have 90 degree phase lag wrt the voltage.

We all know when secondary is open IV through the primary are 90° outta phase, in reality little less due to various magnetizing losses. Loading the secondary with resistive load brings primary IV more into phase, like i in the first post, "as if resistor appears in parallel with the primary inductance". Ofc magnetizing IVs are always 90° outta phase and load currents are always in phase, well, at least if load is purely resistive.

If load is inductive then picture is not so clear, but extending what happens in the core to the load, we can assume this inductive load will also appear as an inductor and resistor in parallel and larger the work done larger the virtual resistor in parallel will appear again bringing IV in phase.

I assume we can extend the last paragraph to capacitive loads altho i never read about this anywhere, of course, electric field just like magnetic field can also be used to do work and as you all probably know there are various electrostatic motors, some newer ones of significant power (some even speculate about replacing magnetic ones). Anyway..

[Mads Barnkob]

Insults and massive double posting got this thread locked. Some posts will be deleted to keep the initial thread within site rules.

Read the rules before posting on HVF: https://highvoltageforum.net/index.php?topic=31.0



EDIT: Thread has been unlocked, for a second chance to stay on topic.

[nix85]

To recap on the subject, for "normal" (non flyback) transformer, i'd summarize it like this to cover all the angles.

When transformer is unloaded (secondary open) only magnetizing current flows through the primary which is tiny compared to load current and is always (almost) 90° outta phase with the driving voltage, almost but not 90° due to various losses, namely, eddy currents, hysteresis, magnetostriction and copper losses (P = I²R). Average flux in the core is maximum in no-load state and slightly smaller in the full-load state.

When secondary is loaded, counterflux developed by the secondary demagnetizes the core and this makes the voltage across the primary to drop since this voltage is directly proportional to rate of change of flux, we all know Faraday's (or should i say Henry's) law V=--N*dΦ/dt.

Voltage across an inductor can also be expressed as V = L(di/dt)

And current through an inductor I = (V-E)/Z where V is voltage of the source driving the primary and E is voltage drop across primary's inductance. Clearly, when secondary demagnetizes the core and flux through the primary drops, so does it's inductive reactance Z and voltage across it E, V remaining the same means current must rise and so it does trying to bring the flux back to the original value but it never fully manages to do so, so, as said before, max load flux is slightly less than no load flux.

To the circuit driving the primary, it appears as if a resistor appears in parallel with the inductance of the primary, bigger the load smaller the resistor appears, obviously.

As said above magnetizing current is always almost 90° out of phase with the driving voltage while the load current is always in-phase. At least when the load is purely resistive.

If load is inductive then picture is not so clear, but extending what happens with the resistive load, we can assume this inductive load will also appear as an inductor and resistor in parallel and larger the work done larger the virtual resistor in parallel will appear again bringing IV in phase.

I guess we can extend the last paragraph to capacitive loads too, of course, electric field just like magnetic field can also be used to do work and as you all probably know there are various electrostatic motors, some newer ones of significant power (some even speculate about replacing magnetic ones).

As for flyback, the only difference is, as said before in the thread, induction in the secondary happens with a delay (due to internal diodes blocking the current in one direction) when the primary flux collapses. Ignition coil uses the same principle, store, collapse, get 10x (or more) voltage in the primary and x turns ratio in the secondary.

Talking of ignition coil, the reason for cap across the switch (be it mechanical relay or a MOSFET) is to limit the peak voltage, cap appearing as a short for an instant and lower the resistance in the circuit of the collapsing flux, lower will be the peak, obviously, higher the resistance higher the peak and faster will it burn out.