Transformers in parallel: What are the pitfalls?

[JonTalbain]

Greetings,
I need a transformer which steps down the 230VAC mains to 18VDC @ 10A. I am unable to locate a transformer of that current rating in my locale, and I am considering connecting multiple 18V/3A transformers (4 of them) in parallel so as to hit the current required.

I have never attempted such a connection before, and I would appreciate any tips on how to go about this sucessfully. BTW, I did come across a 18V/13A transformer, but it was huge and for the cost, I could have purchased an off the shelf power supply. I am building a 12V 10A regulated power supply, regulated by a LM7812.

Thanks for any help!

Regards,
Alvin

 

[electricpete]

As you may know parallel transformers will share real load in inverse proportion to their impedance and there will be a circulating reactive current if the turns ratio's do not match. With identical transformers that shouldn't matter to you (particularly since it sounds like you've got 20% overload capacity which should handle any slight circulating current). Of course you'll have to pay attention to make sure you don't wire together leads of opposite polarity which would create a short circuit. I can't think of any other problems.

 

[JonTalbain]

Hi eletricpete!
Thanks for the response. No, I am not aware of such ill effects of transformers; this will be the first time I am attempting to build a multi transformer PS, previous circuits were pretty small capacity and used a single transformer.

The 18V3A transformers will be obtained off the shelf, I am not certain of their turn ratios will be matched...

If so, would it be safe to use even more transformers e.g. 5 of them so I will get 5 x 3 = 15A theoretical max? Or should i just leave it at 4 transformers? The output of the power supply I am building will be used to power a battery charger which will operate at 12VDC, with 9A constant current draw.

Thank you!

Regards,
Alvin

 

[maximka]

From my own experience: a little voltage imbalance will create large currents. You will have to match your voltage outputs. It is hard, but I managed to do that with two DC supplies using their own regulators.

 

[electricpete]

The circulating current in 2 paralle transformers can be calculated as the difference in voltage on low-voltage side (when energized from hi-side), divided by the sum of the impedances of the two transformers (references to low voltage side). I'm not sure how you would calculate for more than two. But you could hook it up with 4 and measure currents in the individual transformers to make sure you have enough margin (I suspect 4 would be enough).

Have you given any thought to overcurrent protection of the transformer. Perhaps it would be conservative to include individual fuse for each transformer. A single protective device common to all four would not be enough to protect a single transformer from overload in the event of turn-to-turn short or other unexpected failure... possible fire hazard? I'm not too sure what is conventional approach.

 

[jmarko]

Hi, JonTalbain!

If you reaaly can not find appropriate transformer, you can consider next tip:
If you connect parralel multiple secondary windings on one transformer or even multiple transformers, as in your case, I suggest that you apply separate four-diode bridge (Graetz) to each winding, and then connect together positive otputs and together negative outputs. In that way, if first winding has slightly lower voltage output, no current will flow, until second winding (with higher voltage) will be so loaded, that a voltage drop will appear; with even higher load current a lower voltage winding (first one) will start to conduct and thus help maintaining output voltage at increased current.
In that way you can connect as many windings as you like; transformers can be even with different power.
I do reccomend winding voltage differences are max. cca 5% , as the winding with higher voltage will be maximum loaded all the time load is applied and winding with minimum voltage only when full load is applied.


Best regards, JMarko

 

[jbartos]

Suggestions:
1. Locate a similar transformer with a higher secondary voltage, e.g. 24 Volts and remove right number of secondary turns so that you will have 18VAC
2. Order one custom made. A manufacture of the custom transformer is a very simple task.
3. Try

http://www.thomasregister.com

and type Transformer: Custom under Product or Services which will return 237 companies to choose from
4. Any connections of more transformers in series, in parallel or a combination will never be as good as the single transformer. There will always be some interactions, e.g. harmonics, transients, balance currents, etc. as hinted in the previous postings.

 

[koolkid]

jmarko was the closest to a decent solution. In that situation, a difference in output voltage of the transformers will cause one to work very hard while the other only steps in when the output impedance of the first transformer causes sufficient voltage drop as load current increases, to the point where its output voltage, under heavy load, is the same as the second output voltage at light load.

To share the load, put a filter choke after the bridge rectifiers, before the common filter cap.

For best result, use a switch-mode power supply designed for charging a battery.
For low-cost design see

http://www.1stquadrant.com/services.htm

For 1z 2z, it's cheaper to buy a charger or to buy at HSC electronics.

 

[salec]

I agree with JMarko, and you can even get by with just one Graetz and two diodes:

The Graetz is applied on one of the secondaries.

Two remaining secondaries are to be tied in series to form a-sort-of one secondary with middle out (observe the winding directions).

The middle gets tied directly to Graetz negative out.

Remaining ends get a diode anode each, with cathodes tied together with Gratz positive out.

Regards, Salec

 

The best way is to rectify the indivdual transformer outputs and then parallel to avoid circulating currents in the transformers

 

[walker1]

From my own experience I tend to agree with Power Guru.
Just recently I tried to connect 2 very different transformer secodaries in series.

1 Toroide and 1 E-I.
One way I got 2.5 volts x-tra, and reversed I got nothing x-tra at all.
OK I did it without any load, but ...

If you have a nearby electronics recycler, or what it may be called in your area,
(230 volts AC points towards Europe, where we have such enterprises. At least in the EU)
you might be lucky to find an old transformer, that goes beyond the standard 3 A maximum.
In the days before the switchmodes some quite heavy transformers vere in use in computers.

Another possibillity would be a switchmode supply. I saw one the other day giving off
2x15V / 15A + 5V / 150A.
It came from some equipment unknown, but the price was just right, if you had the need.
Equ. $40 or so.

 

[visitor]

How about wiring two step up transformers parallel? Specifically, two 120VAC to 240VAC transformers. We need to run a motor that came from Australia and is a 240 volt motor. We have several step up transformers (same type), but I'm concerned about overloading one transformer. If the current ends up pushing the limit, do you think we can wire two of these parallel?