Transformer Woes

[Andrew Leigh]

Hi,

I salvaged a 24V to 220V transformer from a failed UPS (650W) as I wish to use it on an inverter I plan on building. On the 220V side there are two seperate windings, each of the same resistance, and each half of the total resistance, from this I assume that the 220V side is OK.

On the LV side my meter, when set to "diode mode" does not scream an audible warning for a dead short. Unfortunately it is an el cheapo and does not have the resolution to measure very low resistances so I thought of applying voltage to the LV side and to measure the output.

I applied 12V d.c. to the 24V side of the transformer and the leads sure got smokin' real quick so I hastily disconnected before the stuff hit the fan. Surely the magnatising current cant be that high? Would I therefore be correct in assuming that I have a shorted turn somewhere on the LV winding as neither side of the LV winding is down to earth?

How can I varifiy the integrity of the LV winding without an accurate meter / bridge?

The truth is, I will be rewinding the 24V winding with a 12V so if the winding has failed it is not material but my understanding of what is happening is.

Regards
Andrew

 

[ericgibbs]

Hi,

I salvaged a 24V to 220V transformer from a failed UPS (650W) as I wish to use it on an inverter I plan on building. On the 220V side there are two seperate windings, each of the same resistance, and each half of the total resistance, from this I assume that the 220V side is OK.

On the LV side my meter, when set to "diode mode" does not scream an audible warning for a dead short. Unfortunately it is an el cheapo and does not have the resolution to measure very low resistances so I thought of applying voltage to the LV side and to measure the output.

I applied 12V d.c. to the 24V side of the transformer and the leads sure got smokin' real quick so I hastily disconnected before the stuff hit the fan. Surely the magnatising current cant be that high? Would I therefore be correct in assuming that I have a shorted turn somewhere on the LV winding as neither side of the LV winding is down to earth?

How can I varifiy the integrity of the LV winding without an accurate meter / bridge?

The truth is, I will be rewinding the 24V winding with a 12V so if the winding has failed it is not material but my understanding of what is happening is.

Regards
Andrew

hi Andrew,
The transformer is designed for 'ac' not 'dc' input.

With a dc input all you have is the winding resistance,,,,, and smoke.!

 

[Andrew Leigh]

Hi Eric,

thanks for giving me that "slap palm of forehead moment".

I KNOW you can't put d.c on a transformer winding ............. have no idea what I was thinking.

Thanks
Andrew

Could this be the result of excessive alcohol abuse now shrinking my brain?

 

[Andrew Leigh]

Hi Eric

Not to hijack my own thread but ........

Surely a UPS is merely a sophisticated sine wave inverter with additional controls etc.

All the inverter circuits I have downloaded show the primary LV winding as having a centre tap. My transformer from the UPS had no centre tap, any thoughts would be appreciated.

Thanks
Andrew

 

[Nigel Goodwin]

Hi Eric

Not to hijack my own thread but ........

Surely a UPS is merely a sophisticated sine wave inverter with additional controls etc.

Unlikely to be sinewave, far too inefficient, and not required for a UPS.

All the inverter circuits I have downloaded show the primary LV winding as having a centre tap. My transformer from the UPS had no centre tap, any thoughts would be appreciated.

It's most probably a high frequency switch-mode transformer.

 

[MrAl]

Hi there,


Without a center tap you need to drive the transformer with an "H" bridge.
If you dont know what that is, look it up for much more info then i can
give here, but basically it's a four transistor circuit that drives the transformer
winding both plus and minus, which as you found out, is required.


The older (and cheap) UPS supplies used (use) a pulse instead of a sine
wave. The pulse goes plus and minus, and equals the average current
that would be delivered with a real sine.

The better UPS supplies use a modified sine wave, which is really a bunch
of pulses whos average equals the average of a sine wave at the point
in time where the pulse occurs (usually the center). With modern MOSFETs,
there can be anywhere from 3 pulses per half cycle to 3000 (roughly) per
cycle, and the more pulses the better as the total harmonic distortion is
lower with the more pulses...some are as low as 1 percent.

If you want to do a pulse UPS you might be able to do it with a simple
circuit, but if you want to do a modified sine wave UPS then you probably
need a microcontroller with a sine lookup table programmed in.
The program would look up the time sine value in the table and provide a pulse
width output according to that times sin(wt) value.

The modified sine converter is not for someone who never designed an
inverter before however, so perhaps you should start with a pulse
inverter, or even a square wave inverter. Once you get that going
you can move on to the more sophisticated circuit.

 

[Andrew Leigh]

Thanks guys,

The load is a heating element for my caravan fridge, from what I have read it will only need to be a square wave design?

Will check the H Bridge configuration for interest, but as the primary is 24V I need to rewind it to 12V, at this point I can introduce a centre tap. But then again it may be more convenient to run it with an H bridge off two 12's.

I do not have the capacity to design will use an existing circuit. I was planning on using one of the two as attached. I favoured the second one as it uses two more 2N3055 per side which would mean a less aggressive loading. Having said which, the duty cycle will be extremely low as I only intend using the inverter whilst travelling. Generally 6 hours trips, perhaps 10 per annum.

Would appreciate comments as to which would be more suitable for my purpose.

Thanks again
Andrew

 

[ericgibbs]

hi,
Whats the fridge heating element watts/power rating.?

My caravan fridge is gas or electric, I need 24Vac at about 3-4 amps for the heater,[ there is no compressor motor]
Be aware IF the 'old' transformer is a hf, SMP design it will not work the same at 50Hz

EDIT: this a £20GBP 150W continous rated invertor.

**broken link removed**

 

[Andrew Leigh]

Hi Eric

Whats the fridge heating element watts/power rating.?
180W

My caravan fridge is gas or electric, I need 24Vac at about 3-4 amps for the heater,[ there is no compressor motor]
Mine is the same type, an absorbtion fridge but at 12V.

Be aware IF the 'old' transformer is a hf, SMP design it will not work the same at 50Hz.
Excellent point.

This a £20GBP 150W continous rated invertor.
I know they are cheap but there is often more satisfaction in making the thing.

Or circumvent the whole issue, get someone to make a dual 12 /220V replacement element. Make a thermostat control (complete with low battery cutout) and run it directly from the battery without any charging from the tow vehicle. Before departure one would need to get the fridge loaded and to temperature on the normal 220V supply. I estimate that the duty cycle will probably be about 30%. A 50Ah battery should last the trip, it can then get the charge replenished on arrival.

Watcha think.


Andrew

 

[chemelec]

Is your transformer a Ferrite Core, Or an Iron Core?

I Wouldn't recommendEither of those 2N3055 Designs.

Mosfets are Much More Efficient for this application.

 

[Andrew Leigh]

Hi,

Iron core (E & I)

Thanks for that advice.

Andrew

 

[audioguru]

The "500W" inverter circuit from Electronics-Lab has many problems and does not work.
This one works with a load of up to 720W.
If it used Mosfets then it would be much more efficient.

 

[chemelec]

Hi,

Iron core (E & I)

Thanks for that advice.

Andrew

Andrew, This C-Mos circuit of mine works well.
Actual Power output Depends on Which mosfets are used and How Many of them.
Also the Power rating of your transformer.

**broken link removed**

Gary

 

[Andrew Leigh]

Hi,

thats awesome. Perfect for a "build by numbers" guy like me. Even has the PCB layout.

Thanks
Andrew

PS: What are the losses?

 

[Andrew Leigh]

Hi,

getting back to the HF issue on the transformer, how could I tell if it was? The link below is the unit from which the transformer was removed.

**broken link removed**

Thanks
Andrew

 

[ericgibbs]

Hi,

getting back to the HF issue on the transformer, how could I tell if it was? The link below is the unit from which the transformer was removed.

**broken link removed**

Thanks
Andrew

hi,
Looked at the link,,1000VA.

If its a 'big' transformer it could 50HZ or if small its a HF type.

What are the dimensions of the transformer.? and a rough weight.?

 

[grim]

if the windings 'smoked' then it could have kn@ckered it, as we in the trade say

sometimes you get away with it, sometimes you don't. but if it has shorted the winding, then it would be a shame to stick it onto your inverter circuit and blow that up as well.

 

[chemelec]

Hi,

thats awesome. Perfect for a "build by numbers" guy like me. Even has the PCB layout.

Thanks
Andrew

PS: What are the losses?

Losses: Depends on how good you wind the transformer.
With a Good transformer, Efficiency should be better than 90%.

 

[audioguru]

The inverter in your UPS has a sine-wave output. Then its transformer will produce the wrong voltage if it is in a simple square-wave circuit. The sine-wave transformer might be a small high frequency type instead of the huge and heavy one for a simple square-wave circuit.

Many electronic products do not work properly when fed a square-wave.

 

[MrAl]

Hi,

As i have said before, when an inverter like this is made and it is to be
made with only a square wave, the best choice is really a pulse who's
width is shorter than the full half cycle. The idea is usually to get
either the same average voltage or the same rms voltage as you
would get with a real sine wave.
A square wave of the same frequency contains more energy than a
sine wave, and it has harmonics that are quite large. Using the
pulse, the lower order harmonics can be reduced somewhat to
help the situation out a little too.
This means an oscillator that can put out two pulses out of phase
with each other could be used to drive the output transistors.
Each pulse could start at the normal start of the half cycle, but
end before the half cycle really ends. This would mean a shorter
pulse and reduced energy getting to the transformer as well as
the output.

As a side note, sometimes the pulses reaching the transformer
primary are integrated and this is used as a feedback signal
to reduce the net dc in the primary. The effect is to
increase efficiency and reduce the audible noise coming
out of the transformer itself.