## Back-to-back two transformers

Expand Messages
• To get a floating 220V AC supply (for very low power), I connected two small transformers (220/12V, 3W) back to back; the two terminals of the low voltage
Message 1 of 9 , Jun 22, 2013
To get a floating 220V AC supply (for very low power), I connected two small transformers (220/12V, 3W) back to back; the two terminals of the low voltage side of T1 (12V) to their counterparts of T2 (12V).

I supplied the high voltage side of T1 with 220V and I expected getting 220V (rms) on the high side terminals. I got 146V (rms) only.

I decreased the input voltage to 25V, the output became 10V!!! (more tests below).

I wonder which characteristic of the transformer affects the most on this voltage drop.

On the other hand, I assume the higher the voltage we get from this simple configuration, the better the type of the transformer under test is (if the comparison is made among transformers that are rated for the same power).

Kerim

Voltages are RMS and measured with a digital scope.

Vin , Vout, Vout/Vin
025 , 010 , 0.400
050 , 025 , 0.500
075 , 042 , 0.560
100 , 060 , 0.600
150 , 097 , 0.647
200 , 133 , 0.665
220 , 146 , 0.664
250 , 165 , 0.660
• Do you have a load on the output of the second transformer besides the volt meter? When not tied together, what is the measured output voltage of your first
Message 2 of 9 , Jun 22, 2013
Do you have a load on the output of the second transformer besides the volt
meter? When not tied together, what is the measured output voltage of
your first transformer at 220V in? Same question for your second transformer.
When tied together, what do you measure at the tie points? What do you
measure at the tie points and the output if you cross the tie points ( 220v = !!
X !! = ?)

Wharpt

In a message dated 6/22/2013 5:05:03 P.M. Central Daylight Time,
ahumanbeing2000@... writes:

To get a floating 220V AC supply (for very low power), I connected two
small transformers (220/12V, 3W) back to back; the two terminals of the low
voltage side of T1 (12V) to their counterparts of T2 (12V).

I supplied the high voltage side of T1 with 220V and I expected getting
220V (rms) on the high side terminals. I got 146V (rms) only.

I decreased the input voltage to 25V, the output became 10V!!! (more tests
below).

I wonder which characteristic of the transformer affects the most on this
voltage drop.

On the other hand, I assume the higher the voltage we get from this simple
configuration, the better the type of the transformer under test is (if
the comparison is made among transformers that are rated for the same power).

Kerim

[Non-text portions of this message have been removed]
• ... There is a property (nonlinear magnetizing inductance??) of iron core power transformers that causes their efficiency to drop when supplied with lower
Message 3 of 9 , Jun 22, 2013
Kerim F <ahumanbeing2000@...> wrote:

> I decreased the input voltage to 25V, the output became 10V!!! (more tests below).

There is a property (nonlinear magnetizing inductance??) of iron core
power transformers that causes their efficiency to drop when supplied
with lower voltages. At least that explains the drop in Vout/Vin at
lower voltages.

Now, just guessing about the lower voltage in general. You'd think
the turns ratio should be 4.28:1 to get the 12 V. But let's say it is
less, maybe 4.0:1 ... so one might expect 13.75 V unloaded. But you
get only 12 V because of leakage inductance (not all the flux is
coupled between windings). In effect there is a built-in voltage loss
due to less than perfect mutual inductance. That loss gets you both
ways, stepping down and again stepping up.

The same explanation works for I*R voltage drops when the transformers

Regards,
Andy
• ... Thank you, Andy, for your detailed answer. I measured the high voltage of T2 without any load (besides the 10M Ohm of the scope probe). I have no idea
Message 4 of 9 , Jun 22, 2013
--- In Electronics_101@yahoogroups.com, Andy <ai.egrps@...> wrote:
>
> Kerim F <ahumanbeing2000@...> wrote:
>
> > I decreased the input voltage to 25V, the output became 10V!!! (more tests below).
>
> There is a property (nonlinear magnetizing inductance??) of iron core
> power transformers that causes their efficiency to drop when supplied
> with lower voltages. At least that explains the drop in Vout/Vin at
> lower voltages.
>
> Now, just guessing about the lower voltage in general. You'd think
> the turns ratio should be 4.28:1 to get the 12 V. But let's say it is
> less, maybe 4.0:1 ... so one might expect 13.75 V unloaded. But you
> get only 12 V because of leakage inductance (not all the flux is
> coupled between windings). In effect there is a built-in voltage loss
> due to less than perfect mutual inductance. That loss gets you both
> ways, stepping down and again stepping up.
>
> The same explanation works for I*R voltage drops when the transformers
>
> Regards,
> Andy
>

I measured the high voltage of T2 without any load (besides the 10M Ohm of the scope probe).

I have no idea about the turn ratio because the two transformers are ready-made. As you said, if this ratios is known, one can deduce the mutual inductance by comparing it to the voltage ratio.

I wished to know if it is just my transformers that are of a rather bad quality. Or perhaps being far from perfect mutual inductance is a general case for almost all `small' transformers in the market.

Perhaps someone will test this drop too if he has two small similar transformers. The experiment is simple ;)

Kerim
• I used my old digital scope (FLUKE 199C) and its 10M probes. Also I set its readings for Vac (RMS). I forget measuring the low voltage, as you suggested,
Message 5 of 9 , Jun 22, 2013
I used my old digital scope (FLUKE 199C) and its 10M probes. Also I set its 'readings' for Vac (RMS).

I forget measuring the low voltage, as you suggested, before and after connecting the two transformers. Doing this is not easy now since I already enclosed the two transformers after covering all internal connections :(.

When I noticed this relatively large drop of voltage between input and output (with no load) I thought that, unlike I, many of you may expect it from experience hence already explained it. But there is always the possibility that my two ready-made small transformers are rather a special case, I mean they could be of a bad quality. If this is the case I think this simple experiment can quickly compare the quality of similar transformers made by different manufacturers.

Kerim

--- In Electronics_101@yahoogroups.com, wharpt@... wrote:
>
> Do you have a load on the output of the second transformer besides the volt
> meter? When not tied together, what is the measured output voltage of
> your first transformer at 220V in? Same question for your second transformer.
> When tied together, what do you measure at the tie points? What do you
> measure at the tie points and the output if you cross the tie points ( 220v = !!
> X !! = ?)
>
> Wharpt
>
>
> In a message dated 6/22/2013 5:05:03 P.M. Central Daylight Time,
> ahumanbeing2000@... writes:
>
>
>
> To get a floating 220V AC supply (for very low power), I connected two
> small transformers (220/12V, 3W) back to back; the two terminals of the low
> voltage side of T1 (12V) to their counterparts of T2 (12V).
>
> I supplied the high voltage side of T1 with 220V and I expected getting
> 220V (rms) on the high side terminals. I got 146V (rms) only.
>
> I decreased the input voltage to 25V, the output became 10V!!! (more tests
> below).
>
> I wonder which characteristic of the transformer affects the most on this
> voltage drop.
>
> On the other hand, I assume the higher the voltage we get from this simple
> configuration, the better the type of the transformer under test is (if
> the comparison is made among transformers that are rated for the same power).
>
> Kerim
>
• I think Andy is correct (although he got the turns ratio wrong - should be about 18:1 not 4:1). Playing around with circuit simulators can be enlightening. The
Message 6 of 9 , Jun 22, 2013
I think Andy is correct (although he got the turns ratio wrong - should be

Playing around with circuit simulators can be enlightening. The following
can be saved as a .asc file and loaded into LTSpice.

This simulation shows two transformers connected as described, with
coupling factors of .9 which gives approximately the results that Kerim
reported.

Version 4
SHEET 1 880 680
WIRE 240 64 192 64
WIRE 336 64 240 64
WIRE 32 80 -32 80
WIRE 80 80 32 80
WIRE 496 80 448 80
WIRE 512 80 496 80
WIRE -32 112 -32 80
WIRE 80 112 80 80
WIRE 192 112 192 64
WIRE 336 112 336 64
WIRE 448 112 448 80
WIRE -32 208 -32 192
WIRE 80 208 80 192
WIRE 80 208 -32 208
WIRE 192 208 192 192
WIRE 192 208 80 208
WIRE 336 208 336 192
WIRE 336 208 192 208
WIRE 448 208 448 192
WIRE 448 208 336 208
FLAG -32 208 0
FLAG 32 80 VP
FLAG 240 64 VS
FLAG 496 80 VT
SYMBOL ind2 64 96 R0
SYMATTR InstName L1
SYMATTR Value 600
SYMATTR Type ind
SYMATTR SpiceLine Rser=100
SYMBOL ind2 176 96 R0
SYMATTR InstName L2
SYMATTR Value 1.5
SYMATTR Type ind
SYMATTR SpiceLine Rser=1 Rpar=1meg
SYMBOL voltage -32 96 R0
WINDOW 3 -85 145 Left 2
WINDOW 123 0 0 Left 2
SYMATTR InstName V1
SYMATTR Value SINE(0 240 50 0 0 0)
SYMATTR SpiceLine Rser=1 Cpar=0
SYMBOL ind2 320 96 R0
SYMATTR InstName L3
SYMATTR Value 1.5
SYMATTR Type ind
SYMBOL ind2 432 96 R0
SYMATTR InstName L4
SYMATTR Value 600
SYMATTR Type ind
TEXT -118 262 Left 2 !.tran 5
TEXT 136 -8 Left 2 !K1 L1 L2 .9
TEXT 352 0 Left 2 !K2 L3 L4 .9

On Sat, Jun 22, 2013 at 6:45 PM, Andy <ai.egrps@...> wrote:

> Kerim F <ahumanbeing2000@...> wrote:
>
> > I decreased the input voltage to 25V, the output became 10V!!! (more
> tests below).
>
> There is a property (nonlinear magnetizing inductance??) of iron core
> power transformers that causes their efficiency to drop when supplied
> with lower voltages. At least that explains the drop in Vout/Vin at
> lower voltages.
>
> Now, just guessing about the lower voltage in general. You'd think
> the turns ratio should be 4.28:1 to get the 12 V. But let's say it is
> less, maybe 4.0:1 ... so one might expect 13.75 V unloaded. But you
> get only 12 V because of leakage inductance (not all the flux is
> coupled between windings). In effect there is a built-in voltage loss
> due to less than perfect mutual inductance. That loss gets you both
> ways, stepping down and again stepping up.
>
> The same explanation works for I*R voltage drops when the transformers
>
> Regards,
> Andy
>

[Non-text portions of this message have been removed]
• ... Oops! Yes indeed. I took the square root of the voltage ratio. My mistake. I was in the habit of doing that when going from impedance ratio to turns
Message 7 of 9 , Jun 22, 2013
> I think Andy is correct (although he got the turns ratio wrong - should be

Oops! Yes indeed. I took the square root of the voltage ratio. My
mistake. I was in the habit of doing that when going from impedance
ratio to turns ratio.

Andy
• I think you are seeing the effect of magnetizing current for the second core, passed through the resistance of three windings, instead of just one (in addition
Message 8 of 9 , Jun 22, 2013
I think you are seeing the effect of magnetizing current for the second
core, passed through the resistance of three windings, instead of just one
(in addition to the resistive losses, supplying the magnetization current
for the first core).

You might try connecting a 400 volt mylar capacitor across the 220 volt
output to supply some of that current, via resonance, to see if it pumps
the voltage up a little. There is some optimum value that puts the system
into resonance, so, if you have a few sizes, you might get some idea what
the optimum value is.

It should work best if half that resonance current was supplied across the
12 volt windings, also, but that would require a lot higher value of
capacitance (I think ((220/12)^2)/2 times as much to reduce the 220 volt
resonating capacitor by half.

--
Regards,

John Popelish

On Sat, Jun 22, 2013 at 6:04 PM, Kerim F <ahumanbeing2000@...> wrote:

>
> To get a floating 220V AC supply (for very low power), I connected two
> small transformers (220/12V, 3W) back to back; the two terminals of the low
> voltage side of T1 (12V) to their counterparts of T2 (12V).
>
> I supplied the high voltage side of T1 with 220V and I expected getting
> 220V (rms) on the high side terminals. I got 146V (rms) only.
>
> I decreased the input voltage to 25V, the output became 10V!!! (more tests
> below).
>
> I wonder which characteristic of the transformer affects the most on this
> voltage drop.
>
> On the other hand, I assume the higher the voltage we get from this simple
> configuration, the better the type of the transformer under test is (if the
> comparison is made among transformers that are rated for the same power).
>
(snip)

[Non-text portions of this message have been removed]
• We were not given a lot of information on the transformers. Some transformers that are intended for installation under building codes, have current limiting
Message 9 of 9 , Jun 24, 2013
We were not given a lot of information on the transformers. Some transformers that are intended for installation under building codes, have current limiting resistors installed in series with the windings. Low Voltage wiring is often installed in buildings without conduit or other protection. These resistors are to limit the current and therefore the danger of creating a fire in the case of a short circuit. These resistors are generally not visible or accessible from the outside, probably to prevent the users from removing them. If these transformers are of that, current limiting, type then there would be two such resistors in series with the "back-to-back" secondary windings and that would drop a significant fraction of the output Voltage generated in the first transformer.

You will always have some losses, but with the proper, identical transformers (not current limited ones), I would expect that under no load conditions the output Voltage of such a connection would be within 5 or 10 percent of the input Voltage. Adding your load will change that number.

Paul A.

--- In Electronics_101@yahoogroups.com, John Popelish <jpopelish@...> wrote:
>
> I think you are seeing the effect of magnetizing current for the second
> core, passed through the resistance of three windings, instead of just one
> (in addition to the resistive losses, supplying the magnetization current
> for the first core).
>
> You might try connecting a 400 volt mylar capacitor across the 220 volt
> output to supply some of that current, via resonance, to see if it pumps
> the voltage up a little. There is some optimum value that puts the system
> into resonance, so, if you have a few sizes, you might get some idea what
> the optimum value is.
>
> It should work best if half that resonance current was supplied across the
> 12 volt windings, also, but that would require a lot higher value of
> capacitance (I think ((220/12)^2)/2 times as much to reduce the 220 volt
> resonating capacitor by half.
>
> --
> Regards,
>
> John Popelish
>
> On Sat, Jun 22, 2013 at 6:04 PM, Kerim F <ahumanbeing2000@...> wrote:
>
> >
> > To get a floating 220V AC supply (for very low power), I connected two
> > small transformers (220/12V, 3W) back to back; the two terminals of the low
> > voltage side of T1 (12V) to their counterparts of T2 (12V).
> >
> > I supplied the high voltage side of T1 with 220V and I expected getting
> > 220V (rms) on the high side terminals. I got 146V (rms) only.
> >
> > I decreased the input voltage to 25V, the output became 10V!!! (more tests
> > below).
> >
> > I wonder which characteristic of the transformer affects the most on this
> > voltage drop.
> >
> > On the other hand, I assume the higher the voltage we get from this simple
> > configuration, the better the type of the transformer under test is (if the
> > comparison is made among transformers that are rated for the same power).
> >
> (snip)
>
>
> [Non-text portions of this message have been removed]
>
Your message has been successfully submitted and would be delivered to recipients shortly.