Will 0.05 volt difference be a issue

[Wynand]

Hi. I've just rewound a transformer with 4 x 25v secondaries.

The difference between them is about 0.05 vac.

Would I be able to run them two of these in parralel.

i.e 2 x 2 to give 25-0-25v secondaries.

I wound them like this because I'd like to add two more channels to my amp later on and didn't want to rewind the tranny again.

Thanks
Wynand.

 

[Holger Barske]

There's an easy way to get rid of problems with currents between paralleled secondaries: put a rectifier on each secondary and parallel afterwards.

 

[eds65gto]

Hi:
It seems that you want to have a pair of secondaries in series to get the 25 - 0 - 25 volts you want. However I do not think .05 volt difference will make any difference.
Hope this helps
Ed

 

[Christer]

It probably depends on the VA rating of the transformer too. For small transformers it is usually OK and even sanctioned by the manufacturers to parallell the secondaries, since their resistance is high enough. For a large transformer, even a small difference in voltage might cause quite a big current to flow just between the windings.

 

[AndrewT]

Hi,
is that the difference from a one turn error?

I wonder if you can get no closer than that since you can only work with whole (integer) turn windings.

I would aim for a one half turn match between paralleled windings.
Adding a small resistor to each winding will try to force current sharing between windings. RC also offers a small attenuation of HF residuals (that's an acceptable word in the Forum dictionary) from the mains.

If you parallel unmatched windings the problem in unequal sharing comes about due the difference in voltage between the partially charged smoothing capacitor voltage less the diode voltage drop and the instantaneous voltage on the upper part of the sinusoidal waveform.
The winding with the higher voltage will push a current based on this voltage difference divided by the charging circuit impedance. This higher voltage difference could be just a few hundred millivolts. If the lower voltage winding were just half of this few hundred millivolts less, then the current would also be halved, since the circuit impedance will be almost identical (same length of copper).
Since heat is I^2R then the heating in the higher voltage winding will be four times higher. If the voltage match were worse than half the effective charging voltage then the heating difference becomes much worse. Take care to match parallel secondaries accurately.

 

[Wynand]

All four secondaries were turned at the same time, so it can't be a 1 turn error. The difference is also too small for that. Each turn is +-0.50 volt and the difference between these secondaries is less way les than that.

 

[quasi]

Fair Dinkum !!

Tolerances of components down the line (rectifier, filter caps etc) will likely have higher error. Even a very small load will drop the voltage more than 0.05 volt. Remember 0.05 / 25 is an error of 0.2% and this figure is better than any tolerance you are likely to achieve with the rest of your build.

So I would convert your re-winding concern to re-winding pride in achieving an error of 0.2%.


Cheers

 

[Eva]

First, in case all the secondaries have the exact same amount of turns, any apparent voltage mismatch can be considered a measurament error rather than a voltage mismatch.

Second, in case of a true voltage mismatch, the current flowing from one winding to another will be limited by the sum of the ohmic resistances of both windings. For example, if each winding has 0.5 ohms and we have a voltage mismatch of 50mV, then only 50mA are going to flow, thus not producing any harm.

Third: There is not such a thing as a half turn producing half the voltage of an entire turn in conventional toroid and E-I transformers. Half a turn produces nothing there.

 

[AndrewT]

Hi,

There is not such a thing as a half turn producing half the voltage of an entire turn in conventional toroid and E-I transformers. Half a turn produces nothing there

has my statement

aim for a one half turn match between paralleled windings

caused confusion?

By that I mean that the measured voltage difference between windings can be reduced to less than half the voltage generated by one turn. If the error exceeds the voltage of one turn then a further turn more or less can be wound on/off to reduce the difference. i.e. aim for a one half turn match between paralleled windings.

 

[AndrewT]

If you parallel unmatched windings the problem in unequal sharing comes about due the difference in voltage between the partially charged smoothing capacitor voltage less the diode voltage drop and the instantaneous voltage on the upper part of the sinusoidal waveform.
The winding with the higher voltage will push a current based on this voltage difference divided by the charging circuit impedance. This higher voltage difference could be just a few hundred millivolts. If the lower voltage winding were just half of this few hundred millivolts less, then the current would also be halved, since the circuit impedance will be almost identical (same length of copper)

and

Second, in case of a true voltage mismatch, the current flowing from one winding to another will be limited by the sum of the ohmic resistances of both windings. For example, if each winding has 0.5 ohms and we have a voltage mismatch of 50mV, then only 50mA are going to flow,

I think these two variants produce different results.
Which is correct?

 

[jan.didden]

and
I think these two variants produce different results.
Which is correct?

0.5 ohms per sec with 50mV delta VAC gives indeed 50mA 'blind' current.

The other issue is a non-issue: BECAUSE of the 50mA delta current, the secondaries now have the same sec VAC so they both charge the caps equally.

Jan Didden