Transformer Rewinding.. Help Needed

[Ayne]

I have a transformer. It is 40-40 Volts(RMS) center tape transformer. I check the voltages with Voltmeter with no load.

I want to get 30 Volts DC at 2A currnet from it after full wave rectification.
What should be the transformer output volts(RMS)???

Is there any need for rewinding the transformer for reducing the it's output voltages???

 

[tunedwolf]

You will need a 20-0-20 transformer to produce ~ +/- 30v full wave dc with a capacitive filter.

Assuming your transformer is rated for 3A, you will need to remove half of the turns on each leg of the secondary. Therefore if it is a single, or split bobbin type core, unless they have devised some sneaky way of putting the windings on that has escaped me the last 30+ years, this means you will have to remove completely one winding and half of the first one, make a new centre tap, and wind on the correct number of turns to complete the second winding again. If it doesn't buzz under load, I'll be surprised, because I doubt you will be able to hand wind it with even tension from start to finish. Although a good dip in transformer varnish works wonders.

It would be far easier, and probably much safer for you, to just buy a 20-0-20 3A transformer. Remember one nick in the enamel between turns and the transformer goes up in smoke...

If you have to re-wind a transformer, never be tempted to use second hand enamelled wire. Always replace the windings completely with new wire.
Never ever join wire inside the transformer winding, always bring the tails out neatly and make any joint outside of the winding window.
Always keep correct and even tension on the wire as you wind, any slack or sloppy winding that creates an anomaly will only be amplified in later layers, possibly resulting in lack of window space for the windings.
Always properly insulate your windings when complete, including a final tape layer, this will help with mechanical strength too.

 

[Hero999]

You will need a 20-0-20 transformer to produce ~ +/- 30v full wave dc with a capacitive filter.

Do the calculations.

[latex]V_{RMS} = \frac{V{DC}+V_{DROP}}{sqrt 2}[/latex]

Assuming your bridge rectifier has a volt drop of 1.2V:

[latex]V_{RMS} = \frac{30+1.2}{1.414}=22.06V[/latex]

But this is only the case under no load conditions. Under full load conditions the bridge rectifer can have a volt drop as high as 2.2V, the transformer voltage will drop a bit and there will be significant voltage ripple on the capacitor so if your regulator requires 30V to regulate properly then it won't regulate properly. All of these factors need to be taken into account. Luckilly the transformer will already designed to provide a slightly higher voltage off load so if you unwind it completely then rewind it for 20V counting the number of turns then it will provid 20V under full load.

Before you decide on the AC inout voltage you need to think about the circuit connected the DC side.

What's the minimum voltage required?

What's the maximum tollerable voltage?

Is a regulator?

What's the maximum current required?

Please be aware that assuming an infinately large smoothing capacitor the RMS input current is always greater than the output current by a factor of root 2.

 

[tunedwolf]

uhuh...but where is he going to get a 22-0-22 3A transformer off the shelf ?
The choice he will have off the shelf, will be 20-0-20 or 24-0-24, unless he finds a surplus place with old amplifier transformers that have just the right spec.

I suspect this is related to his other posts about the amplifier...so I figured a smidge less than 30-0-30 won't hurt...

As for counting the turns, he needs to count the turns coming off the secondary and calculate the turns per volt ratio, based on his input voltage, then calculate how many are required for 20 volts. The output current available after rectification will be ~ 1.24A with a 2A transformer, or 2A with a transformer capable of producing ~ 3.25A.
I didn't want to get into his % of turns per volt per 10VA of loading etc. I figure that will confuse the issue for him, and he hasn't supplied enough detail about the stack used to give him an accurate figure anyway. There's no guarantee that the transformer he is planning on rewinding is even capable of the current, because he hasn't supplied us with enough details...

I would say if he needs 2A, after rectification, get a 3A 20-0-20 off the shelf, cheaper, faster and definately safer for him.

 

[Ayne]

I have opened my transformer.

It's Turns/Volt = 3.82 Turns

Now i am rewinding again for

Two windings of 26 Volts(RMS).
Total Turns = Turns/Volt * Volts
Total Turns = 3.82 * 26
Total Turns = 100 Turns approx.

 

[tunedwolf]

did you not bother reading the previous posts ?

you need 22 V per winding, if you want 30 V per winding after the rectifier, therefore 3.82* 22 = 84 turns

 

[Hero999]

uhuh...but where is he going to get a 22-0-22 3A transformer off the shelf ?
The choice he will have off the shelf, will be 20-0-20 or 24-0-24, unless he finds a surplus place with old amplifier transformers that have just the right spec.

I suspect this is related to his other posts about the amplifier...

Well I though it was for a small linear power supply but we'll never now unless he tells us.

so I figured a smidge less than 30-0-30 won't hurt...

Either way given the choice he's better off with that 2V extra.

As for counting the turns, he needs to count the turns coming off the secondary and calculate the turns per volt ratio, based on his input voltage, then calculate how many are required for 20 volts.

I agree.

The output current available after rectification will be ~ 1.24A with a 2A transformer, or 2A with a transformer capable of producing ~ 3.25A.

Where did you get your figures from?

I would say if he needs 2A, after rectification, get a 3A 20-0-20 off the shelf, cheaper, faster and definately safer for him.

That depends on whether all the componants are rated to 35V and if they are he should go for 24V, otherwise he should use 20V to be on the safe side.

 

[Ayne]

Infact i want to make Variable power supply from +,-30 to 0 Volts.
Using LM317.
Now my plan is Rewind a transformer with four windings
two winding of 26-0-26 Volts RMS for DC 30 Volts Approx
two winding of 11-0-11 Volts RMS for DC 12 Volts Approx

My idea is when LM317 gives it's out put lower than 10 volts it will switch to second winding(12 Volts winding).

 

[audioguru]

The peak of 26VAC is 36.8V. The rectifier bridge will drop about 2V at full load leaving 34.8VDC, not 30VDC. The LM317 needs an input of at least 32.5V for a 30V output so it will be fine.

 

[Hero999]

You'd need pretty big capacitors though.

A better solution would be to have one 26-0-26V winding with a 11V tap at either side of 0V.

Now let's see what size smoothing capacitors you need, what frequency is the supply in your area? I'll assume 50Hz because the values I calculate will work for 60Hz as well.

The LM317 always requires the input voltage to be 2.5V greater than the output when given a 1.5A load. I'm assuming you want 1.5A given the other thread?

Your regulator behaves like a constant current source to the capacitors will discharge in a linear fashion so the size of capacitor can be given by the following formula:
[latex]C = \frac{I}{2FV_{RIPPLE}}[/latex]

F being the mains frequency and Vripple being the maximum allowable ripple below the unloaded DC value.

Unloaded your 26V transformer will give:
[latex]V_{DC} =(26-2.2) \times sqrt 2=33.9V[/latex]

Your regulator needs 32.5V to regulate properly so the maximum allowable ripple below 33.9V is 33.9-32.5 = 1.44V.

Now let's claculate the minumum capacitor value:
[latex]C = \frac{1.5}{2 \times 50 \times 1.44}=10417 \mu f[/latex]

Note this is the minimum value, you can't use 10,000:mu:f because the the ripple would appear on the output, anyway even if it was you'd need to add on 20%, giving 12,5004:mu:f (to account for the +-20% tolerance margin). You'd need to go add another 3300:mu:f capacitor in parallel both being quite large capacitors when rated at 40V.

Increasing the AC voltage by 1V to 27V would give you a 36V DC output so the ripple can be as high as 3.48V.

Let's calculate the capacitor for that:
[latex]C = \frac{1.5}{2 \times 50 \times 3.48} = 4306 \mu f[/latex]

So go for 4,700:mu:f with 1000:mu:f in paralell to account for the 20% tolerance.

 

[tunedwolf]

my mistake...I dunno where I got it in my head that he was planning on using this transformer for an amplifier...I think I must have my threads tangled...

in any case...he asked for 30V after rectification, without reference to what kind of filter etc he was using, or supplying any proper information about the transformer characteristics etc...

I see now however, that he actually wants around 33 volts...

 

[Ayne]

Thanks Hero999
and
I have rewind my transformer successfuly.
Now it is,
25-0-25 volts
10-0-10 volts

without reference to what kind of filter

Only capacitors for filtering prupose

I am going to add
4 capacitor of 3300uF 35 volts for 25-0-25 volts windings
4 capacitor of 3300uF 15 volts for 10-0-20 volts windings

Infact i am going to make power supply for my projects
You can say "Work Bench power Supply"
Thanks

 

[Hero999]

You need either need bigger filter capacitors of more turns on that secondary for a higher voltage.

Try perfroming the calculations I went through in my last post to you figures and you'll see what I mean.

I don't think you understood my 26V-11V-0V-11V-26V tapped transformer idea. You use the same rectifier and filter capacitor and just switch between the 11V and 26V taps to change the voltage range. Note that these voltages should be the voltages when fully loaded so aim for 27V to 28V off load.

 

[toddbailey]

why not just use a voltage regulator?