120/220 Transformer Wiring Question

[bostonman]

I have a linear power supply I've been repairing (the one I needed a knob for those following another thread) and it has a 120/240 switch on the rear.

Unfortunately the switch (much like most of this power supply) is cheap and was barely making contact due to it falling apart. Upon unscrewing the switch to hopefully reassemble, it fell apart. It has two tiny metal pieces that appear to connect various metal tabs  in order to switch between the two input voltages.

The issue: the switch fell apart before I could trace the connections, so I don't exactly know which position is 120/240 or where the tiny metal pieces go. The switch is on a tiny PCB with the line voltage on one side (a red and black wire), and a four pin connector on the other side with all four wires going to the transformer.

The four wires are two sets of red/black. From what I am measuring, the red and black is one coil and the other red and black is the other coil. i.e. I don't get continuity on anything other than red/black of one set and the other red/black set.

Using a milli-ohm meter, one red/black pair measures 0.997ohms and the other red/black pair measures 1.051.

If I had to guess, I'd say the 1.051ohm pair is the 240V coil, but I'm looking for confirmation and/or a method of trying to confirm the right pair to hard wire this at 120V.

The transformer doesn't have any writing or part numbers on it. So I'm basically stuck with guessing which input pair is the 120, and, since I'm uncertain of which voltages should be on the secondary, I can't power and measure.

Update: the two black wires on the right go to a tiny Molex (?) connector that can be seen on the left side that I believe wasn't connected upon disassembling. Not sure what it's for, but it's not part of the 120/240 PCB.

[ArdWar]

Most transformers usually come with two "half" primary windings where you connect them in series for 220/240 V input or in parallel for 110/120 V input.



If I'm to guess, it would be R-[B+R]-B for 240V and [R+R]-[B+B] for 120 V.
You can check with low voltage AC at the input first if you're not exactly sure, especially if you aren't sure which winding direction (i.e. polarity) each half of the windings are. They are usually in the same direction however for the vast majority.

[bostonman]

Are you saying the two red wires would be tied together to the hot of the input and the two black tied together to the neutral of the input?

I tried to find transformer wiring diagrams online earlier before posting without luck, but this makes sense. I'll try digging a bit deeper by measuring the PCB now that I have a visual on how it should be connected.

Thanks, this is a help.

[bostonman]

The picture helps with a visual of your description.

What happens if I (accidentally) connect to S2/F2 instead of S1/F1?

[CaptDon]

There is no way in hell you should attempt to apply 240vac into the 120vac winding of the transformer. Not only is the working impedance to low but the core will more than likely go into hard saturation! It will probably smoke within seconds if it doesn't immediately trip your branch circuit breaker. Also, the way the capacitor run winding is supposed to be connected across the main winding there is no way that capacitor should hold a charge after power is disconnected!!! Something is still seriously wrong!!! It just can't get any simpler....For Christ's sake there are two windings, one is a main winding and one is a winding connected in series with a capacitor to get phase shift to make the motor start and run in a given direction. In a way it makes a single phase motor into a split phase motor. So many posts, so many different wiring attempts and it still hasn't even turned 1/4 of a rotation?? It should run although slow and feeble on 240vac!!!! Don't even think about trying 400+ volts until you see some kind of sensable activity at 240vac. As for 360vac, you simply stack the 240vac winding on top of the 120vac winding (assuming they are separate isolated windings in the first place) and you get 120 + 240 = 360 if you have the 240 phased correctly. If you connect the 240 winding with the wrong polarity I think you end up with 120 since 120 - 240 = -120 so the polarity would be flipped from the main. in that case main to neutral = 120 vac, output to neutral I believe is also 120vac, output to main = 240? Any way, you have a four wire motor and a capacitor, this isn't rocket science. More than likely if the motor is working correctly you can flip the rotation by reversing the connection of either winding relative to the other.
 

[bostonman]

Did you read the thread or did you accidentally respond to the wrong thread?

I'm not applying 240V AC to a 120V AC coil nor running a motor. Your concern for safety is appreciated, but nothing you typed applies to my question.

[CaptDon]

ooppps!!! Sorry, fat fingered the mouse pad!! Response was supposed to be to the guy playing with a motor. Wow, don't even know how that happened, the posts aren't even near each other in the list??? Sorry again!! Strange things have been happening here lately. I have a wireless router dongle plugged into a hardwired router (wife needs the wireless option for her work computer) and this Chromebook has been doing weird stuff ever since! Enjoy, cheers mate!!

[CaptDon]

Ard gave the best solution given what can be derived from the available information. For 120 vac the windings would be in parallel red to red and black to black and it isn't important if mains hot goes to the reds or to the blacks, you still get 120vac on the paralleled windings. For 240vac the F1 gets connected to S2 and input is applied to S1 and F2. Assume for 240vac operation S1 is black, that can be connection one input connection, F1 being red will then be tied to S2 which is assumed to be black and F2 being red would be the other input connection so you have a pair of 120vac windings in series and properly phased. You can determine which red and black pair are the same winding with an ohm meter so you will determine 'winding 1' and 'winding 2' to make sure you get the connections right. Hope that helps!!

[tooki]

Update: the two black wires on the right go to a tiny Molex (?) connector that can be seen on the left side that I believe wasn't connected upon disassembling. Not sure what it's for, but it's not part of the 120/240 PCB.

Just from its silhouette, it looks like a JST XH connector, FYI.

As for how to use the primary windings, there’s little I can add that hasn’t already been said expertly by others. But attached is a schematic for how to wire a DPDT switch as a voltage selector, nabbed from https://uk.rs-online.com/web/p/slide-switches/0319900

[Ian.M]

If in *any* doubt  about mains transformer primary connections/phasing, first powerup should be with an incandescent^# bulb (mains voltage) in series with the supply to limit the current.

• The bulb may flash when you apply power but should rapidly go very dim or off if all is OK.
• If it lights up bright, either you've got the phasing wrong, or there's a heavy load on one or more secondaries or the transformer is bad.  Disconnect secondaries if necessary and try again.

Keywords: dim bulb tester

# The bulb must have a tungsten filament.  Halogen bulbs are acceptable.  LED and CFL bulbs are not.

[bostonman]

CaptDon, absolutely hilarious about responding to the wrong thread. Taking time to type a response about safety (even though it was the wrong thread) is well appreciated and hopefully the OP in the other thread appreciates it just as much.

I will try the light bulb test.

Earlier I traced that tiny PCB. It has two wires on the input (hot and neutral - one is tied directly to neutral - the other I assume is hot - it’s open but I’m guessing due to going through a relay on another PCB). The other side is a four pin connector (as I mentioned going to the transformer).

I’m only guessing about the slider for 120/240 making the connections in the attached hand drawn pictures , but, if I had to bet my paycheck, would believe those are where the tiny metal tabs touch when the slider is slide to either position.

[bostonman]

If in *any* doubt  about mains transformer primary connections/phasing, first powerup should be with an incandescent# bulb (mains voltage) in series with the supply to limit the current.

I re-read your suggestion because most likely I'll have time today and wanted to confirm a few things. The transformer has several secondaries (four or five). In all cases, I only care about 120V mode and don't have intention of using 240V.

From my previous two drawings, my guess on the connections appear correct; it's wired in series for 240 and parallel for 120.

A few scenarios come to mind that make me wonder if the bulb test proves something is wired wrong, whether the power supply circuits can get damaged.

1: I'm wrong and the parallel is really series, so I power it with 120, but the transformer is in 240 mode (most likely nothing will happen and the power supply will fail to work or work correctly)

2: I get the phase wrong on both primary coils

3: I get the phase wrong on one primary coil and correct on the other

Due to wire colors, my connection guessing, measurements, and my observation on how the switch should be assembled, most likely I'm making more out of this than needed and I should just hard wire it and be done.

My other plan (besides the light bulb test) is to use a diff probe and scope the primary and secondary simultaneously to check phase and amplitude (maybe use a variac and disconnect all secondaries so it's only the transformer).

[ejeffrey]

If you get the phase wrong on both primaries it doesn't matter.  Only the relative phase matters.  If you connect the two in parallel with one phase flipped it will basically look like a short circuit and the bulb will light up normally, indicating a fault.

Another way you can do this is to disconnect the secondary windings and use a function generator to apply a low voltage (like 1 V) to one of the secondary windings.  That will induce medium voltage on the primary.  This will allow you to measure the voltages and make sure you are connecting them in parallel with the correct phase.

[bostonman]

If you get the phase wrong on both primaries it doesn't matter.  Only the relative phase matters.

Not sure I understand what you meant by the relative phase.

[bostonman]

This is weird. I am getting out of phase with  hot and neutral in one position and flipped.

My scope has channel 1 on the primary, and channel 2 on the secondary.

Regardless how the input leads are connected to the transformer, the primary and secondary remain out of phase.

I’m trying to rig up a set up for the light bulb test, but expected to see an in phase waveforms.

[bostonman]

The light bulb lights and stays lit the same in both polarities. It’s not very bright, somewhat orange, but the same brightness.

[ejeffrey]

If you get the phase wrong on both primaries it doesn't matter.  Only the relative phase matters.

Not sure I understand what you meant by the relative phase.

You are connecting the two primaries in parallel.  There are two ways to do that, one of them is wrong.  Once you do that you have two wires but it doesn't matter which one you use for live and which for neutral.

[Ian.M]

Wattage of the bulb vs VA of the transformer?

If a linear PSU you can assume the transformer VA is a bit over 1.5x max DC output power (i.e. max. Vout x Imax at that output voltage), adding all channels.

[bostonman]

I'm confused about this transformer.

The tiny PCB the 120/240V switch was on has four wires out into the transformer (as previously mentioned). I assume and assumed these are the only primary windings.

What I recently realized is on the same side on the bottom is four wires (red, black, white, and yellow) that go into a connector and two yellow wires that go into another connector. These connectors go into a board with relays and regulators. On the other side of the transformer on the top are a few more wires but smaller gauge. Two connectors are red, red, and black. Another is yellow and yellow. On the bottom are more wires that go to two more four-pin connectors also red, black, white, and yellow thicker gauge.

Having wires on four points (top, bottom, left, and right) is making this confusing.

Ideally I'm quite confident this can be wired based on my assumption with the switch positions, but hoping I can make more sense out of this to be sure.

From what I mentally concluded, the worst case is: Since this will never be powered with 240, I wire the transformer in 240 mode in which case the power supply won't work. I've already concluded that the transformer needs to be connected in parallel,and traced the two windings, so this leads me to whether I can wire the phase wrong.

I expected the phase to be 180 degrees off when the primary is wired wrong and in phase when correct. Seeing the waveforms 180 off in both cases confused me.

[Ian.M]

Err no.  If the primaries are wired antiphase, in series or parallel, the transformer will draw excessive current, close to being a short circuit.  If phased correctly, the transformer, if unloaded, will only draw magnetizing current.

Secondary phase relative to the primary feed is a different matter and is generally not relevant except in devices with isolated control circuits doing phase angle control of mains SCR controlled bridges or similar, using the secondary as a timing reference and steering signal for the SCRs..

[bostonman]

Is it safe to just connect the secondary to the circuits and power it?

This entire time I had the secondary open. If I can connect the secondary and test the connections in the primary for both in phase and out of phase, then maybe that will make this easier.

I can measure the current in both cases and see which one is higher telling me that's the wrong phase.

[Ian.M]

You haven't answered the bulb wattage and transformer VA question, so we have no idea if the glow is bad phasing or maybe a faulty transformer, or is just the magnetizing current.  However if the bulb is bright then goes a lot dimmer if you swap one primary's connections or visa versa, the correct connection is when its dim, i.e. lower total primary current.   Adding a secondary load before you determine correct primary relative phasing will just confuse things by increasing the current, making the bulb brighter and harder to distinguish from its brightness with incorrect phasing.

[ejeffrey]

Is it safe to just connect the secondary to the circuits and power it?

No.

There are exactly two ways to connect the primary windings in parallel.  That is what everyone means by "phase", the relative phase between the two primary windings.  The phase of the secondary doesn't matter most of the time.

If you have connected the two primary windings the wrong way and you apply 120 V, you will destroy the transformer unless you are lucky enough that your circuit  breaker saves it first.

You need to find the correct phase relationship between the two primary windings and connect them in parallel with the same phase orientation.  Using the filament light bulb in series with the primary is a safety factor in case you got that wrong.

All of that should be done with the secondaries unloaded.

[ArdWar]

Don't think too much about it lol. The primary leads are already color coded red and black. There's no reason to assume the phase/polarity not to be also connected as indicated. It would require special kind of evil company to manufacture otherwise.

[bostonman]

The power supply itself (Korad Power Supply - Model KA3305P) has two 30V outputs rated at 5A each and a dedicated 5V rated at 3A.

Later I'll measure the current with a 60W bulb in series.

Unless I'm misunderstanding the wording/terminology, in all cases the secondaries are disconnected (although I'm uncertain whether the wires on the bottom on one of either side of the transformer could be another primary). The only connections are the red to red, and black to black. In one case I connect HOT to red and red, and NEUTRAL to black and black.

I crank up the variac and the bulb has an orange glow. After I place the HOT on black and black, and NEUTRAL on red and red, crank the variac, and the same orange glow.

As mentioned, most likely nobody did anything evil to mess up the color codes at the factory. Ideally I can probably just wire it hot to red and red and neutral to black and black, and be done with this, but I was hoping to get the visual on seeing precisely what happens when it's wired wrong versus correct.

[Ian.M]

Its transformer is certainly over 500VA  so use a 100W bulb, or 2x 60W in parallel.   

Reversing the Line and Neutral connections has *NOTHING* to do with phasing.

Incorrectly phase the parallel primaries by reversing one primary winding (i.e. red to black x2) , or placing them in 'series' joining two wires of the same color, with Line (via the bulb) and Neutral  to wires of the other color, or short one of the secondaries, and you will see the bulb light near full brightness, limiting the current and protecting the transformer from the miswiring.

[bostonman]

Reversing the Line and Neutral connections has *NOTHING* to do with phasing.

Hahahhah oh my. This entire time I was reversing hot/neutral thinking that changed the phase in the transformer. As you know, sometimes I over think, sometimes I'm just unfamiliar with such in depth portions of electronics, or a combination.

If that's the case, then I don't need to worry about anything. This entire time I questioned whether when the switch fell apart did the two tiny metal tabs touch more than the points I assumed they touch and/or did another metal tab fall on the floor without me knowing.

From what I can tell, the connections would have been HOT to red and red, and NEUTRAL to black and black, so I may be able to finally just assemble it based off that.

I can still try a light bulb test. There is only one other way to connect this to get them out of phase (now that I believe I'm understanding this better), and that would be HOT to red, and HOT to black on the second coil. Then NEUTRAL to black, and red on the second coil. That would make it one phase in coil one, and reverse phase on coil two.

[Ian.M]

Exactly.  That's why you do the dim bulb test if the primary wire colors or terminal numbers are ambiguous, or if you are replacing the voltage selector switch and need to confirm you wired the new one correctly - test with 115V supply and check neither selector switch position makes the bulb go bright.  It will also show most internal transformer faults like windings with shorted turns, though if the faulty secondary is low current and a small fraction of the total VA, the bulb may not go full brightness so the results may be hard to interpret.

[bostonman]

I got the expected results using a 100W light bulb.

With the assumed correctly connected phase (using an inline meter), the light bulb was off: 120V AC using a VARIAC, 203mA, 17.75W, 0.741 power factor.

With the assumed incorrectly connected phase (I was too afraid to go higher in voltage) the light bulb was on: 30V AC using a VARIAC, 413mA, 12.5W, 1.0 power factor.

Afterwards, knowing red to red and black to black was correct, I hardwired all the wires, reconnected the connectors, etc... and tested the power supply. As expected it turned on, but I didn't do a load test yet (unfortunately the only adjustable load I have is a single channel 180W variable load).

A few things baffle me about this experiment. Initially I used a 60W light bulb and it was on when the wires were wired correctly. Was this due to the resistance being higher in the filament?

Also, I don't comprehend how wiring the two primary coils out of phase causes a short whereas in phase they keep the bulb from turning on.

In phase and the two coils are in parallel. Initially I measured (rounding off) 1 ohm on both, so it would be 0.5ohms and would think the bulb would be on due to the low resistance in series with the bulb.

[wasedadoc]

In phase and the two coils are in parallel. Initially I measured (rounding off) 1 ohm on both, so it would be 0.5ohms and would think the bulb would be on due to the low resistance in series with the bulb.

Surely someone with your number of posts knows that a multimeter applies DC and measures resistance whereas the bulb test has an AC voltage source and the large inductance of the transformer primaries (when correctly phased) is dominant in determining what AC current flows.

[ejeffrey]

That's totally uncalled for, clearly he doesn't know about this in particular and thats why people ask questions. Nothing wrong with that.

When the two coils are wired in parallel, they generate magetic fields in the same direction.  This generates flux in the core and magnetizes it, which is why it has a large inductance that limits the current.

When the coils are anti parallel, the magnetic fields cancel.  There is no magnetic flux in the core and the inductance is very low.  Then the current is only limited by the DC resistance which is very low.

[Ian.M]

Thanks for the clear and concise explanation.  I'm sure it will help Bostonman grock transformers better. 

Incandescent bulbs typically have a cold filament resistance an order of magnitude lower than  their hot (at nominal power) filament resistance.   
A 100W 120V bulb will have a hot resistance of  144 ohms.   Its cold resistance will typically be between 13 and 21 ohms depending on how hot its filament normally runs (color temperature).  The hotter/brighter it normally runs the lower the cold resistance, so a halogen spotlight bulb will give you a significantly lower cold resistance than a rough service inspection lamp bulb of equal voltage and power rating.

To a first approximation the resistance increases linearly with absolute filament  temperature.  However (by Stefan–Boltzmann law) the emitted power (radiated heat and light) of the filament in vacuo^*, is proportional to the fourth power of the filament temperature.  In steady state, the filament must be in thermal equilibrium, which means that at low powers (relative to rated power), a small increase in current results in an increase in temperature and thus a large increase in resistance and power so the bulb acts as a current limiter. 

The trip current is hard to determine without a lot of math. Its easiest to get a feel for it with a small low voltage bulb and a bench CC/CV PSU (as most of us don't have good continuously variable mains AC loads of near unity power factor).  Set the voltage to the nom. bulb voltage and turn the current *way* down, then slowly bring it up plotting resistance (V/I) vs I on graph paper as you do so.   As low voltage bulbs typically run at lower color temperatures than most mains bulbs, unless you use something like a krypton torch bulb, the result wont be directly applicable to your 'dim bulb' but should give you a decent idea.

Once 'tripped' the bulb limit the current to a bit less than the bulb's normal operating current, so  around 0.8A for that 100W 120V bulb.   The estimated  500VA transformer must draw at least 4.17A total primary current in normal operation, so unless it relies on fan cooling, the 'tripped' current from the dim bulb tester is safe indefinitely.

* In practice, nearly all incandescent bulbs have a low pressure inert gas fill, so one also has to consider heat losses due to convection, but the fourth power term typically dominates.

[bostonman]

I appreciate you sticking up for me. I'm also sure sometimes the very active users who respond to questions can sometimes get frustrated with posts that are clearly from others who don't take time to conduct basic research before posting.

Sometimes I've been guilty of jumping the gun and posting about an anomaly in a measurement/test/etc only to realize I didn't look at the problem with both eyes open after getting a reply.

To his point, and your technical reply, I was thinking in terms of resistance and not magnetic fields. So basically when the magnetic fields cancel, I'm left with two 1ohm coils in series, and a light bulb that lights as a result due to higher current.

[bostonman]

ncandescent bulbs typically have a cold filament resistance an order of magnitude lower than  their hot (at nominal power) filament resistance. A 100W 120V bulb will have a hot resistance of  144 ohms.

On a side note, months ago I did a cold start 60W light bulb test with a current probe connected to my scope. I was actually shocked to not see a higher initial current draw.

Also, I didn't realize that a cold filament was so high in resistance until I measured one of the bulbs before applying power. Until then, and your explanation now, I thought they were well under 10ohms at initial turn on.