Transformers in parralel

[Freakazo]

Hi all! I'm new here so forgive me if I'm doing something wrong.

I'm just wondering what the effect will be if I have two different power supplies connected in parallel (specifically the transformers I'm worrying about)

the circuit will be something like this:

rectifiers ->4 230V caps in parallel (about 10Farad total) -> switching transistor -> 2 transformers in parallel -> 2 rectifiers in parallel -> 2 filtering circuits in parralel.

It is basically two switch mode transformers in parallel (with a few alterations such as only one switching transistor.


I don't know the exact specifications of the transformers, will that be a big problem?
The 2 rectifiers in parallel might be a problem, but will it be big enough to worry about? they are rated for 38A and the total load will not exceed 50A, I'll check the difference in forward voltage drop tomorrow when I get my multimeter back...

Anything else I should worry about?



Oh yeah, I'm a bit new to electronics

 

[rjvh]

38 amp ,50 amp, 10 Farad, new to electronics

that's a good recipy for trouble and get zaped

what is the voltage and what are you trying to achieve

Robert-Jan

 

[Freakazo]

If it is a good recipy I'm up for it

I'm trying to mod a PC PSU, I need more juice and don't have money to buy those 800W PSUs. I have 5 PSU laying around which I can use, so to mod it will be the cheapest route.

 

[rjvh]

I would not try to conect the outputs of several PSU together

the only mods I can recomend to make is to increase the battery capacity of the PSU for longer longetivity when it have to work

the PSU work with generated sinuses on variable frequentys depending on power usage of the equipment you might try to pararalel the Mosfets in the unit and than you need to increase the transformers in the unit induvidual also but conecting the units together will probably disturb all units conected resulting in failur of the units or conected equipment

I would save for the bigger PSU

Robert-Jan

 

[Freakazo]

That's why I'm only combining them after the switching stage (right before the transformer) so that they will use a single frequency.

 

[rjvh]

You might try to put 2 transformers the input windings in serial conection and the output windings parralel

I would think that the windings must have identical spects on resistance

increase the input voltage by 2 of course, just as an experiment but i can not guarantee you anyting in this set up

you don't have any info on the driver part but normaly they work on 12V and the components used do withstand 24V

Again See it as an experiment the PSU are not designed to do this kind of things

and personaly I would not put my computer on it

but for the sake of I want to know if you do this it would be nice to waste a few not needed PSUs

and than still save for the bigger one

Robert-Jan

 

[Hero999]

4 230V caps in parallel (about 10Farad total)

Sorry, I don't believe that your capacitors are 230V 10F, they'd need to be 2.5F each and would be huge - big enough to fill a room.

 

[arhi]

I'd say it is not easy doable, if doable at all. The much easier way to add extra juice to your PC is to put two PSU's in the box (if you can find room in there) and power up different devices with different PSU's...

The easiest way is to
- connect grounds from both PSU's together
- mod the power switch to power on the second psu at the same time it powers the first one
- do not connect + or - leads from both psu's together
- power the MB/GPU/SoundCard from one PSU
- power the HDD's, DVD's and fan's from second PSU

That will do the trick (do not forget to connect ground from one PSU to another, black wire) in most cases. You will get extra juice as you will split the load and you will not have to meddle with the insides of the PSU. I have used this for over a year (one box with 2 atx psu's and one box with 1 atx and 1 at psu) without a problem, but, I do not guarantee that this solution is problem free, maybe someone more experienced can comment on this proposal

As for the "meddling" with the PSU insides, note that most PSU's if not all have few built in protections (short circuit, overload...) so you would have to modify those circuits too in order to allow PSU to put out more juice... not for noob's and easy way to release the magic smoke from your PC

 

[Freakazo]

2 of them are 4700uF and the other 2 are ~1200uF capacitors, 1000uf = 1F right?
I'm going to test two PSU on a p3 PC and see if there is any smoke cause it seems there isn't really a direct answer, or so far there isn't...

 

[JimB]

1000uf = 1F right?

Wrong!

1000uF = 1mF (milli-Farad)

1000000uF = 1F

JimB

 

[Freakazo]

doh, my bad!

so it will be 10mF not 10 farad

Thanks!

 

[Hero999]

Yes but we don't normally use mF because it can be confusing, some older capacitors and textbooks it means micro-Farad.

Generally for values greater than =>100,000:mu:F then I use Farads, for values <100,000:mu:F I use :mu:F.

For example instead of saying 220,000:mu:F I say 0.22F.

 

[Freakazo]

We derailed a bit in the thread, let's start over:

What will the effects be if two different transformers are placed in parallel with two different ratings (one has a bit more or a bit less winding ratio than the other)?

 

[Willbe]

In any case, run the DC output of each supply to the anode of an isolation diode, and connect the cathodes together. JC Whitney has these.
If the supplies are equal voltage you might get reasonable sharing of current. To get better current sharing add some series resistance to the diode anodes.
Your output impedance will be higher than for a regulated supply but that may not matter for this task.

 

[MrAl]

Hi there,


When you connect two transformers in parallel and intend to drive
them with a voltage source it is basically the same as
connecting two voltage sources in parallel. If the two sources are the
same it works just fine, but if the two have a different voltage then
the current even without any load is equal to the difference in voltage
divided by the two source resistances added together, or
I=(V1-V2)/(R1+R2)
This means that if the winding ratio is much different for one it could
mean a big difference in output voltage for that transformer, which
would mean a huge current could flow in both transformers even without
any load.
For transformers that normally put out a lot of current the series resistance
is going to be small, so the current could be very high when they are
connected in parallel and fed with some voltage.
For example,
transformer A has turns ratio 10:1, and equivalent series Rs of 0.066 ohms,
transformer B has turns ratio of 12:1 and equivalent series Rs of 0.1 ohms,
the two are connected in parallel and connected to a source of 100 volts,
the output current is 10.0 amps even with no load.

 

[Hero999]

the circuit will be something like this:

rectifiers ->4 230V caps in parallel (about 10Farad total) -> switching transistor -> 2 transformers in parallel -> 2 rectifiers in parallel -> 2 filtering circuits in parralel.

There's nothing wrong with that.

It is basically two switch mode transformers in parallel (with a few alterations such as only one switching transistor.

So they're switched mode power supplies now. In that case it proably won't work, the two supplies will fight each other can become unstable.

Connecting two transformers with rectifiers and capacitors on the output is fine but doing the same with switched mode power supplies isn't because they're active circuits and are a lot more complicated than simple transformers with rectifiers and capacitors.

 

[MrAl]

......
Connecting two transformers with rectifiers and capacitors on the output is fine ......

Hello Hero,

I dont think i can agree with that, because if the two transformers are
too different then one transformer has to deliver most of the current
and gets hot while the other sits there not having to do anything
not delivering much current.
Perhaps some sort of load balancing would help, if the two transformers
were not too much different.

 

[Hero999]

It doesn't matter because in this instance each transformer has a rectifier on the output which will take care of that.

 

[MrAl]

It doesn't matter because in this instance each transformer has a rectifier on the output which will take care of that.

Hi again,


Well here is the way i see it...

If we had a 20v output transformer and a 10v output transformer and
each had rectifiers to convert the ac to dc and a filter cap to help
smooth out the dc, and we wired the two caps in parallel and connect
a load, the 20v output transformer would keep the output caps charged
up and supply most of the load while the 10v output sat there idle not conducting any (or very little current) because the higher voltage
output keeps the rectifier diodes of the 10v output reverse biased.

Thus, although it doesnt seem to hurt anything it also doesnt
help anything.

 

[Freakazo]

I=(V1-V2)/(R1+R2)

For example,
transformer A has turns ratio 10:1, and equivalent series Rs of 0.066 ohms,
transformer B has turns ratio of 12:1 and equivalent series Rs of 0.1 ohms,
the two are connected in parallel and connected to a source of 100 volts,
the output current is 10.0 amps even with no load.

Ah thanks! That's exactly what I wanted to know!

How would one measure series resistance, or exactly what is the series resistance? (google isn't working, all I get is transformers the series ) Is it the resistance of just a single coil or the total resistance across both coils?