ASK: Determining transformer's windings phase

[BravoV]

Scope of my question is on common mains step-down transformer which has 2 windings on the primary side.

The case here is, at the primary mains side it has 2 x 110 V windings, and to use it at 220 V mains, both windings must be connected in series with the right phase connection (the left illustration below, instead of the right one), cmiiw.




Questions :

1. How to determine the right phase connections on both windings "easily/practically", if the transformer doesn't have the proper labeling, say like the label is torn off or lost ? (only for the primary side, assuming the secondary side winding already properly identified).

2. What will happen if the primary side is connected like the one at the right at above picture ? Magic smoke ?

TIA

[Messtechniker]

Simply apply some low level 50 or 60 Hz AC from a signal generator and measure or scope the output.

The connection with the higher output on the secondary side is the proper series connection.

To double check:
When reversing, the output on the secondary will be lower (half). i.e. no scope needed.

[Ian.M]

The incorrectly phased primaries on the right will act pretty much like a dead short (limited only by the lowish winding resistance), as their magnetic fields cancel out.   Result: one dead smoking transformer.

For first powerup of a mains transformer, if you aren't 100% certain of *all* winding phasing,  its a good idea to have a mains voltage incandescent^* bulb (ideally of a wattage between 5% and 50% of the transformer's VA rating) in series with the supply, and to *not* connect the secondaries to any load till you've checked voltages and phasing.  If there is a  fault, the bulb will limit the current so its low enough not to burn out the primary.  If the bulb lights, you've got a problem . . .

* It must be a real tungsten filament bulb.  CCFL and LED bulbs are totally unsuitable, because you need the non-linear resistance characteristic of a heated metal filament.

[mvs]

1. How to determine the right phase connections on both windings "easily/practically", if the transformer doesn't have the proper labeling, say like the label is torn off or lost ? (only for the primary side, assuming the secondary side winding already properly identified).

Connect both primary windings in series, apply some small AC voltage (you may use signal generator or another transformer) to one primary winding only and then measure voltage over both.
If the connection was correct, you will get voltage doubled.

[sorin]

You can use a LCR meter (use 100 Hz test frequency).
The correct winding's should have higher Inductance.

[ArthurDent]

Use the method Ian.B describes. Connecting the two primary windings out of phase and plugging the transformer directly into the wall will be very bad because you will have almost a dead short.

[BravoV]

Thanks for all replies, really appreciate it. 


Especially to Ian.M, thanks alot, this is what I'm expecting, to test it using simple stuffs. 

The reason is I'm helping a friend that lives far away, that is helping me to procure used transformers, and he has only DMM in hands, so your method is really handy as he need to test it on the sale site 1st.

The incorrectly phased primaries on the right will act pretty much like a dead short (limited only by the lowish winding resistance), as their magnetic fields cancel out.   Result: one dead smoking transformer.

For first powerup of a mains transformer, if you aren't 100% certain of *all* winding phasing,  its a good idea to have a mains voltage incandescent^* bulb (ideally of a wattage between 5% and 50% of the transformer's VA rating) in series with the supply, and to *not* connect the secondaries to any load till you've checked voltages and phasing.  If there is a  fault, the bulb will limit the current so its low enough not to burn out the primary.  If the bulb lights, you've got a problem . . .

* It must be a real tungsten filament bulb.  CCFL and LED bulbs are totally unsuitable, because you need the non-linear resistance characteristic of a heated metal filament.

Noted, on real incandescent lamp, and we still have these plenty those here, even though they're starting to fade out.

So if the connection is correct, the bulb supposed to be dimly lighted ? (low current as the secondary is not loaded at all), cmiiw.

[Ian.M]

You probably wont see much of a glow, if any, under normal room lighting, unless its right at the bottom of the wattage range.  An excessive glow indicates high magnetizing current, which may be due to over-voltage causing saturation or excessive eddy currents in the core.   It will light for a bad windings with a shorted turn.

[BravoV]

You probably wont see much of a glow, if any, under normal room lighting, unless its right at the bottom of the wattage range.  An excessive glow indicates high magnetizing current, which may be due to over-voltage causing saturation or excessive eddy currents in the core.   It will light for a bad windings with a shorted turn.

Noted, thanks.

Assuming there is no bad shorted at the internal primary windings, and also the connection is correct, and also with secondary is not loaded at all, what level of current I should be expecting ? Say I put the connection in series with a DMM to measure the current.

[ArthurDent]

Check minutes 9 through 13 of this video to see what happens when Dave plugs a 120VAC transformer into 240VAC. This was a commercially made product that had a serious design flaw of no line fuse. It shows why you should be careful and why proper fusing is important. This is why posters say to put an incandescent bulb in series when testing an unknown transformer.

[BravoV]

ArthurDent, thanks, I'm aware of that video. 

Just want to know if there is a simple method without involving advanced tools like scope + sig-gen, even though I own it, its just I want to help friend that is at remote area, that has only DMM to test the transfomer on site, and the incandescent lamp method pointed by Ian.M is perfect for my case.

[Gyro]

It's going to be hard to give a close figure, but with correctly phased primaries and open secondaries, all you're left with is magnetizing current. That will depend on the frame size and quality of the transformer - the steady state current ought to be in the 10's of mA range or less. A shorted turn would certainly make its presence known.

A big NOTE though... At switch-on the current will peak very high - worst case, the transformer will get switched on at a voltage zero crossing and the core may well saturate in the first half cycle, peak current then will be AMPS. With separate A and mA jacks on the meter, that's going to be impossible to handle.

Difficult to arrange, but if you really want to measure quiescent current, you need to bypass the meter with a switch or something during power-up. Difficult to arrange safely - especially when trying to tutor a friend at a distance! You may well be advised to stick with the dim bulb tester - you can switch a low wattage mains bulb.

[Jwillis]

The general practice when transformers are wound with a Live ,Common and Tap on the primary side is a bifilar winding.Meaning 2 parallel wires are wound together at the same time.This method give a more uniform balanced coil with matching number turns. These turns start at the center and tap out on either side.
Even if the coils wound monofilar from  one side to center to other side, your center  taps  are always tied to together (series coils)for 240V.For 120V the center tap of coil 1 is tied to outside of coil 2 and center tap of coil 2 is tied to outside of coil 1 (parallel coils).
Regardless whether its a bifilar or monofilar windings, the coils are always wound the same direction. And the Live and Neutral lines are on the outside in both cases (120V or 240V). This is the proper procedure used when winding E core or Toroidal transformers.

[basinstreetdesign]

The incorrectly phased primaries on the right will act pretty much like a dead short (limited only by the lowish winding resistance), as their magnetic fields cancel out.   Result: one dead smoking transformer.

For first powerup of a mains transformer, if you aren't 100% certain of *all* winding phasing,  its a good idea to have a mains voltage incandescent^* bulb (ideally of a wattage between 5% and 50% of the transformer's VA rating) in series with the supply, and to *not* connect the secondaries to any load till you've checked voltages and phasing.  If there is a  fault, the bulb will limit the current so its low enough not to burn out the primary.  If the bulb lights, you've got a problem . . .

* It must be a real tungsten filament bulb.  CCFL and LED bulbs are totally unsuitable, because you need the non-linear resistance characteristic of a heated metal filament.

Has anybody ever had this situation with the dual primaries wired in series but out of phase?  I have and the result is non-climactic.  With all secondaries open, if the primaries are wired out of phase their magnetic fields are in opposition and cancel within the core.  All secondaries then show next to zero voltage.  Except for magnetizing current, the turn-on transient is minimal.  Swapping the leads on one primary brings it right.  Testing this requires nothing more than a meter.

Carrying this further, I have used some of the low voltage secondaries wired in series with, and, in or out of phase with the primary(s) to adjust the voltage on other secondaries as needed.
The only time there may be smoke is if there is a shorted turn or three.

[Jwillis]

I did make a mistake re winding primaries once and all it did was trip the breaker.On secondaries all you get is the difference .For example with to coils one at 12.5 volts and one at 12 volts all you get is around 0.5 volts.Your right its not very dramatic.Theirs a few rewind businesses locally so I just went down to a few to get some practical information for my experiments.But they showed me how to keep things in phase. It was really interesting and the winding machines are really cool.Especially the toroid winder.

[Ian.M]

Miswired smaller transformers may not
draw enough primary current to trip the breaker, as the fine wire of the primary can fuse at the leadout or an internal thermal/overcurrent fuse can blow before the breaker trips.   

I don't have any decent sized mains transformers hanging around at the moment so I grabbed my 250VA portable isolating transformer (toroidial 240V, 1A), took the plug off it and put a batten lamp holder in series.  A 100W bulb didn't glow at all.  A 40W bulb flashed dimly with the initial surge current then didn't glow.  Unfortunately I don't have any lower wattage mains incandescent bulbs.

On a cheap DMM's 10A range the surge flicked up to several hundred mA then the magnetizing current very rapidly dropped away below 10mA, the limit of its resolution.  If I'd tried a lower current range the surge would have almost certainly blown the multimeter fuse, so I *DO* *NOT* recommend attempting to measure the magnetizing current in the field with a DMM. 

If you want to do it on the bench, either use a 5 digit or better meter on its 10A range and don't use a transformer over 250VA, 240V or 120VA, 120V (allowing for an x10 surge at switchon, unless you have an essentially free source of meter fuses!  ) or rig a shorting switch across the ammeter that you can close before power on, and when changing meter ranges, only opening it to take a reading.   

Scoping the magnetising current is a whole different ballgame.  There's little point of doing so unless you are investigating transformer design and core saturation.  You'll need an isolated current probe and an isolated HV differential voltage probe to see the phasing with respect to the supply voltage.  Trying to cheap-skate it without the proper probes is likely to result in the death of your scope, and yourself if careless.  An expert engineer can build suitable probes but their design is not really a fit topic for this forum as due to the Dunning–Kruger effect, presenting such a design is an attractive risk to all the idiots who shouldn't be trusted to even wire a plug.

[Ian.M]

@basinstreetdesign,
I question your results with antiphase series primaries - IMHO without current limiting the result will at best be a blown fuse.

Consider a pair of wires side by side, joined at one end, then wound together onto a core.  As their magnetic fields cancel out at any significant distance from the wires, the presence of the core has no significant effect on the impedance seen at the free ends so if you apply voltage across the two wires at their free ends the current will be no less than it was before they were wound on the core.  This is the principle of the bifilar choke which has negligible effect on the load current, as it cancels out, but presents a high impedance to common mode conducted EMI.

[ledtester]

Here's a youtube video which covers the topic along with the dangers of connecting the primary side windings in series with opposite phasing.

Discussion of the test procedure starts around 5:50.

[BravoV]

A big NOTE though... At switch-on the current will peak very high - worst case, the transformer will get switched on at a voltage zero crossing and the core may well saturate in the first half cycle, peak current then will be AMPS. With separate A and mA jacks on the meter, that's going to be impossible to handle.

Difficult to arrange, but if you really want to measure quiescent current, you need to bypass the meter with a switch or something during power-up. Difficult to arrange safely - especially when trying to tutor a friend at a distance! You may well be advised to stick with the dim bulb tester - you can switch a low wattage mains bulb.

Well, my friend works as technician, developed a technique at the field, that he used for years on skipping the initial big inrush using his meter, he works mainly on AC motors.

What he told and taught me with this technique, when he forgot to bring his AC clamp meter, is to use an ordinary DMM, and connect it like this illustration below, using a sacrificial wire and a pair of safe isolated wire cutter.



His DMM has those expensive Hirschmann cable piercing clips and bite into the sacrificial wire touching the copper, and on the initial power on at AC motor, especially big one will hum and slowly starting to rotate, he will wait until the motor's spin stabilized, and then cut the wire in the middle to measure the current.

Of course this is not totally fool proof, but he claimed he never blown any fuse, as he knows & aware of the load. He used this method only occasionally and with great care as he aware of the dangerous.

He will be using this technique. 

[IanMacdonald]

Might be worth clarifying that transformers (at least, mains transformers) do not cause any phase change. It is simply a question of the winding polarity (as in, having righthand or lefthand turns as seen from the same end) being the same or reversed.

The word phase tends to get bandied-about, and it confuses beginners into thinking there is some kind of resonance or time delay going on.  No, it's just the same as batteries in series. Except with AC.

[Gyro]

A big NOTE though... At switch-on the current will peak very high - worst case, the transformer will get switched on at a voltage zero crossing and the core may well saturate in the first half cycle, peak current then will be AMPS. With separate A and mA jacks on the meter, that's going to be impossible to handle.

Difficult to arrange, but if you really want to measure quiescent current, you need to bypass the meter with a switch or something during power-up. Difficult to arrange safely - especially when trying to tutor a friend at a distance! You may well be advised to stick with the dim bulb tester - you can switch a low wattage mains bulb.

Well, my friend works as technician, developed a technique at the field, that he used for years on skipping the initial big inrush using his meter, he works mainly on AC motors.

What he told and taught me with this technique, when he forgot to bring his AC clamp meter, is to use an ordinary DMM, and connect it like this illustration below, using a sacrificial wire and a pair of safe isolated wire cutter.



His DMM has those expensive Hirschmann cable piercing clips and bite into the sacrificial wire touching the copper, and on the initial power on at AC motor, especially big one will hum and slowly starting to rotate, he will wait until the motor's spin stabilized, and then cut the wire in the middle to measure the current.

Of course this is not totally fool proof, but he claimed he never blown any fuse, as he knows & aware of the load. He used this method only occasionally and with great care as he aware of the dangerous.

He will be using this technique. 

Haha, I was holding back on suggesting something like that!  He'd better make sure those clips are really high integrity, otherwise he's going to be searching his remote location for a mA fuse or Aluminium foil / copper nail 

P.S. A really clean turn-off would be advisable under those circumstances too.

[spec]

Hi BravoV

[1] Connect the primary windings any way that you like.

[2] Connect an incandescent bulb (40W to 100W) between the pair of primary windings and connect the mains across the bulb and the primary windings in series.

[3] Note the brightness of the bulb

[4] Reverse the connection of the primary windings and repeat [2] above.

[5] Note the brightness of the bulb.

The primary connections that gives the dimmest bulb are the correct connections for 230V.

[ArthurDent]

Even if you think you're sure your connections to a transformer are correct you should always use an incandescent lamp in series for safety while testing.

An easy way to check to see if you have the two equal 110 VAC primaries connected in parallel correctly that only requires a DMM is to try the wiring I show in the modified drawing below. Wire the first of the two primaries through the bulb to 110 VAC with one wire of the second identical primary connected to one of the wires that is powering the first primary. Measure the AC voltage from the unconnected wire on the second primary to the single wire on the first primary. If the voltage is zero, the primaries are wired in phase which is correct  but if the voltage is 220 VAC then the primaries are out of phase and you need to revere the second primary wires and try again and you will now get zero volts. 

[T3sl4co1l]

Has anybody ever had this situation with the dual primaries wired in series but out of phase?  I have and the result is non-climactic.  With all secondaries open, if the primaries are wired out of phase their magnetic fields are in opposition and cancel within the core.  All secondaries then show next to zero voltage.  Except for magnetizing current, the turn-on transient is minimal.  Swapping the leads on one primary brings it right.  Testing this requires nothing more than a meter.

You've contradicted yourself there -- you're applying flux (voltage at frequency) to the sum primary windings; if the field in the core is forced to zero, it can only be that all windings will appear short circuited under that condition!

Since zero minus zero equals zero, the anti-series primaries will see no back EMF, and full mains current will flow (likely limited by the resistance and leakage of that winding, more than the mains supply itself).

(As it happens, this condition does not force the field to zero; secondary voltage is independent of primary connection when the primary is anti-series, and vice versa.  Simply, the amp-turns cancel out, making that winding be a lump of wire wrapped around the core, but otherwise an open circuit to the field.)

If you did in fact have this and nothing happened, then something else was wrong, like the primary was wired correctly but the secondary was opposite (+Vsec - Vsec = 0), or the fuse opened, or the circuit wasn't closed after all.  FYI.

Carrying this further, I have used some of the low voltage secondaries wired in series with, and, in or out of phase with the primary(s) to adjust the voltage on other secondaries as needed.
The only time there may be smoke is if there is a shorted turn or three.

Also known as a buck or boost transformer.  It's a normal application, and has the advantage of the autoformer: you only need as many VA as you're adding, not the whole thing.

Tim

[ArthurDent]

T3sl4co1l - "Also known as a buck or boost transformer.  It's a normal application, and has the advantage of the autoformer: you only need as many VA as you're adding, not the whole thing."

That is the way the TrippLite automatic  power line conditioners work. They have a 20A transformer with a tapped secondary that is switched by relays in the older units to add or subtract two 12 volt steps to give 2400 watts corrected line voltage in a small package (LC2400 is 7"x6"x9.5"). If the 123 VAC input drops to 101 then the sensing circuit adds 12 volts to change the output back to 123. If the input drops one more step to 89, then it adds 24 to bring it back to 123. It just reverses the secondary to subtract 12 or 24 to correct for overvoltage of 135 and 147.