UPS transformer with 3 wires altogether

[max.wwwang]

Hi,

I torn town another Eaton UPS today. I'm surprised it uses a transformer with only 3 wires altogether. Refer to the photos, they are Red, Brown & Blue. I measured the resistance between all of them. It's about 12 Ohm Brown-Red and Brown-Blue, about 3 Ohm Red-Blue. I guess it's an ”autotransformer" with only one winding. I hooked Brown/Red with mains 240V AC and tested voltage between Red/Blue I got about 32V AC. The voltage between Brown/Blue is about 270V.

From this, I think Brown/Blue is the full winding, and Red is in the middle. If I hook Brown/Blue to 24V AC [this should be 240V AC], I would be able to get 28V AC between Reb/Blue and 212V between Brown/Red. Is my understanding correct? I can easily have another test.

My main question, though, is - what's the point of not using a transformer with two separate windings which provides electrical isolation between the primary and secondary? Also, is this transformer good to be used in a power supply? (Its high capacity looks useful)

Thanks.

[max.wwwang]

More photos.

[andy3055]

Your reasoning seems correct on the first part. As for feeding 24 volts between Brn/Blu, you will not get anything more than that between any of the taps because those are the 2 ends of the complete winding. If you are feeding 240 volts between Brn/Blu, you will get 211.76 volts from Brn/Blu. I believe you missed a '0' there.

As for why they used an auto-transformer in this case, it is because it is cheap to produce and it is not supposed to provide mains isolation in this case anyway. And so, for that very reason, this will not be a good candidate for a bench power supply.

[max.wwwang]

Your reasoning seems correct on the first part. As for feeding 24 volts between Brn/Blu, you will not get anything more than that between any of the taps because those are the 2 ends of the complete winding. If you are feeding 240 volts between Brn/Blu, you will get 211.76 volts from Brn/Blu. I believe you missed a '0' there.

As for why they used an auto-transformer in this case, it is because it is cheap to produce and it is not supposed to provide mains isolation in this case anyway. And so, for that very reason, this will not be a good candidate for a bench power supply.

Yes, I missed a 0; it should be 240V AC. And another test confirmed this. Pity this good beast is not suitable for a DIY bench top power supply!

[andy3055]

You can always work out the turns ratio and rewind a secondary with a smaller wire to suit your need discarding the smaller winding from the existing to make room. It is not that hard. If you go that route, make sure to measure/note the voltage from the smaller section you are discarding and count the number of turns so that you can work out the turns ratio.

[Gregg]

It is possible that the brown is the 240V line; the blue is neutral which is common.  The red / blue pair is for charging the battery and powering the circuitry; again the blue is common for reference.
Looking at the picture of the transformer, you may be able to peel off the outer layer of yellow tape and find a splice between the two windings that can be separated.

[ArthurDent]

If you look at the photo of the transformer where you can see the colored leads coming out toward the top of the photo (shown below) you can figure out how the transformer is wired, which is close to what others have said. The blue wire is connected to 2 darker colored magnet wires on the right which is one end of the low voltage winding and the end of the entire 240 volt winding.

Near the middle you can see where the red lead is connected to the other end of the 2 darker colored magnet wires which is the other end of the low voltage winding and these 2 wires are in parallel to carry the high current when the unit is running on the 24 volt battery. Also connected to the red lead is the start of the lower current, higher voltage remainder of the 240 volt winding. This magnet wire has a lighter colored enamel insulation.

To the left of the red wire that is connected to the 3 mentioned magnet wires is a black sleeved wire that is connected to the brown wire. To the left of this is the high voltage end of the lighter colored magnet wire which is connected to the lead inside the adjacent piece of black sleeving. What is between the 2 pieces of black sleeving is a thermal fuse lying against the core so if the transformer overheats, the fuse opens and the transformer is dead but you haven’t started a fire.

While you could physically separate the start of the high voltage winding at the red lead and bring that out as one end of the primary, there probably isn’t adequate insulation to have 240 volts between those 2 separated windings and it would represent a safety hazard. Also putting 240 volts on a winding designed for 212 volts would probably cause the transformer to overheat and the thermal fuse would blow.

So this was a rather long explanation of saying the transformer should be used as intended or scrapped and don’t try to modify it.

https://www.eevblog.com/forum/beginners/ups-transformer-with-3-wires-altogether/?action=dlattach;attach=928620;image

[max.wwwang]

You can always work out the turns ratio and rewind a secondary with a smaller wire to suit your need discarding the smaller winding from the existing to make room. It is not that hard. If you go that route, make sure to measure/note the voltage from the smaller section you are discarding and count the number of turns so that you can work out the turns ratio.

Indeed I wanted to rewind one or two transformers to suit my needs (I'm sort of obsessed with the 'powerfulness' of these beasts). I have already collected some and may one day choose from them the right candidates for the core and wires. Thanks.

[max.wwwang]

It is possible that the brown is the 240V line; the blue is neutral which is common.  The red / blue pair is for charging the battery and powering the circuitry; again the blue is common for reference.
Looking at the picture of the transformer, you may be able to peel off the outer layer of yellow tape and find a splice between the two windings that can be separated.

According to this video --

(10:38),

it seems this is a "line interactive" UPS and the autotransformer is not used for charging the battery but for stepping up or down the mains line voltage when needed (Screenshot below). The difference is the transformer in the video has 4 leads while mine with 3. But with more relays it's still possible to step up and down voltage with only 3 wires.

If it's really of no sue in its current form, I'll do a bit more tear down of the transformer itself. Thanks.

[max.wwwang]

If you look at the photo of the transformer where you can see the colored leads coming out toward the top of the photo (shown below) you can figure out how the transformer is wired, which is close to what others have said. The blue wire is connected to 2 darker colored magnet wires on the right which is one end of the low voltage winding and the end of the entire 240 volt winding.

Near the middle you can see where the red lead is connected to the other end of the 2 darker colored magnet wires which is the other end of the low voltage winding and these 2 wires are in parallel to carry the high current when the unit is running on the 24 volt battery. Also connected to the red lead is the start of the lower current, higher voltage remainder of the 240 volt winding. This magnet wire has a lighter colored enamel insulation.

To the left of the red wire that is connected to the 3 mentioned magnet wires is a black sleeved wire that is connected to the brown wire. To the left of this is the high voltage end of the lighter colored magnet wire which is connected to the lead inside the adjacent piece of black sleeving. What is between the 2 pieces of black sleeving is a thermal fuse lying against the core so if the transformer overheats, the fuse opens and the transformer is dead but you haven’t started a fire.

While you could physically separate the start of the high voltage winding at the red lead and bring that out as one end of the primary, there probably isn’t adequate insulation to have 240 volts between those 2 separated windings and it would represent a safety hazard. Also putting 240 volts on a winding designed for 212 volts would probably cause the transformer to overheat and the thermal fuse would blow.

So this was a rather long explanation of saying the transformer should be used as intended or scrapped and don’t try to modify it.

https://www.eevblog.com/forum/beginners/ups-transformer-with-3-wires-altogether/?action=dlattach;attach=928620;image

Thanks for your good observation and thorough explanation. I didn't know there is a thermal fuse embedded in this thing - certainly it's a good idea (I've seen similar thermal protection in various motors including universal motor, which simply works as a switch passively and similar thing on sinks of power transistors which works with a sensing/reacting circuitry).

Will have a closer look at the wires with different colours and with or without sleeves!

[max.wwwang]

Another beautiful thing is it seems like the UPS I have here has the very similar topology and uses the similar technology as described in this patent document --
https://patentimages.storage.googleapis.com/fb/75/34/13fab3664adb39/WO1993012570A1.pdf
which is a highlight of Dave's video in my opinion!

One thing that I don't understand is that when I was very excited and tried to hook the two wires of the second biggest transformer in this unit (the two wires the opposite side of the battery wires), the main switch of my house immediately tripped. It didn't work as I thought.

I can see clearly a H bridge with 12 MOSFET in 4 groups (3 in each group wired in parallel), right next to the two fat battery wires, and I don't see an apparent diode rectifier bridge or a charging transformer as Dave said in this video.

Any idea?

[ArthurDent]

The "second biggest transformer in this unit" isn't a transformer at all but probably a dual inductor in series with the A.C. line and has almost zero voltage drop so you put a dead short across your A.C. line, tripping your breaker.

Here's two drawings showing how I think your backup unit works. I didn't show the switching/relays involved but this is a crude diagram of how the charge and carryover functions work. Your backup unit may have separate diodes but the patent shows using the internal diodes in the MOSFETs so I just copied that schematic to illustrate the point. There would also be some sensing circuits and current limiting but this is basically how I think it works. By combining the two windings as an autotransformer they saved some money and because everything is totally enclosed you don't need any line isolation.

In the top drawing the 240V comes and is stepped down to about 24V and the diodes rectify this voltage to charge the batteries. In the bottom drawing things are switched around when the loss of line voltage is sensed so the battery is powering the inverter circuit and you get 240V out of the transformer.   

[SeanB]

The transformer also looks like the low current section of the windings is done in copper wire, but those parallel pairs used for the 24VAC section look suspiciously like CCA wire, the curse of modern electronics and motors, as it fails quite quickly from insulation failure, breaking from fatigue at the ends, or simply corroding away from the terminals.

Otherwise the transformer at least has a thermal fuse, and also has double insulation between the winding sections, and probably will actually work with the thinner winding disconnected from the thick, though it will definitely run hot.

[max.wwwang]

Obviously the circuitry should not be that simple. Fortunately my house wiring is idiot-proof to some degree!

Had a bit more reading of the patent document and more tracing of the PCB, and its topology becomes clearer to me. Overall the unit is faulty but it may still be perfect to be salvaged as a high power inverter with my own Arduino PWM input to the H-bridge. (It seems a bit too complicated to me to re-engineer it with external sensing and signalling circuits based on Arduino, etc.)

It's amazing that the patent document is so well-written and so instructive, so good that it is a perfect reference for anyone who wants to understand how a UPS works in general ("prior art", without this invention) and the new idea of this invention as well!

I hope Dave will make another video dedicated to this patent idea.

[max.wwwang]

The "second biggest transformer in this unit" isn't a transformer at all but probably a dual inductor in series with the A.C. line and has almost zero voltage drop so you put a dead short across your A.C. line, tripping your breaker.

Here's two drawings showing how I think your backup unit works. I didn't show the switching/relays involved but this is a crude diagram of how the charge and carryover functions work. Your backup unit may have separate diodes but the patent shows using the internal diodes in the MOSFETs so I just copied that schematic to illustrate the point. There would also be some sensing circuits and current limiting but this is basically how I think it works. By combining the two windings as an autotransformer they saved some money and because everything is totally enclosed you don't need any line isolation.

In the top drawing the 240V comes and is stepped down to about 24V and the diodes rectify this voltage to charge the batteries. In the bottom drawing things are switched around when the loss of line voltage is sensed so the battery is powering the inverter circuit and you get 240V out of the transformer.

Thanks very much for the effort you have taken in drawing this up and explaining. I see what you mean. By saying the "second largest transformer" I mean the lower left one in the photo of the PCB. Not the common mode (or differential mode?) noise filter at the right hand side of the photo.

[Based on what I understand from here https://techweb.rohm.com/knowledge/emc/s-emc/01-s-emc/6899 and the wiring, it's common mode.]

Indeed the Brown/Blue wires are connected to the mains input via a relay and it seems like the 24V AC is used somewhere. There are 8 IRF740 MOSFETs close to the traces of this voltage but it doesn't seem like they are in a formation of an H-bridge (with each pair in parallel). It's very clear, however, that another 8 IRF3205 MOSFETs (at the left hand side) are directly hooked up with the battery wires (and the second largest transformer) which I have identified are in a formation of an H-bridge (each pair in parallel).

[max.wwwang]

The transformer also looks like the low current section of the windings is done in copper wire, but those parallel pairs used for the 24VAC section look suspiciously like CCA wire, the curse of modern electronics and motors, as it fails quite quickly from insulation failure, breaking from fatigue at the ends, or simply corroding away from the terminals.

Otherwise the transformer at least has a thermal fuse, and also has double insulation between the winding sections, and probably will actually work with the thinner winding disconnected from the thick, though it will definitely run hot.

If the thinner winding is disconnected from the thick (so isolated), why will it run hot? Thanks.

[Jwillis]

I think ArthurDent has it correct. The secondary is tapped off the primary common. Its a 240 to 24 Volt auto transformer.
  Sorry I had to correct a mistake

[max.wwwang]

I've taken two more photos of the board (both sides) for easy analysis. For easy reference, I've flipped the top side (up side down) so the traces and components are in the same places in the photos.
I've also added red lines showing connected traces (at the other side) and big components with simple symbols using Windows Paint.

[max.wwwang]

The other side.