Transformer selection for dual output power supply

[dcbrown73]

Hello,

I would like to design and build a dual output power supply that can produce 30v at up to 5amps per output.  (though 3 amp is easily probably enough)

I was looking at the Hammond MFG toroidal transformers, but on their datasheet it says the following:

The transformers have dual primary and secondary windings.

Note: Units are designed to have all windings engaged (either series or parallel
connected) or connected as an autotransformer. Connection sheet included with
transformer.

Does this mean I cannot use the two secondary windings independently as two voltage sources?  Or is it just talking about the primary windings should be connected in either parallel or in series and not only one winding connected?

If I can, then I have the following question.  If I want up to 30v on the output.  This would mean I would have to select a transformer that provided a bit more than that, but I see them show 30v output then the next step up is 60v output.    A LM338 takes a maximum of input of 40v.    How exactly could I go about lowering it to an acceptable voltage in the most efficient way?   

Or a better question would be, what is the best course of action here?

Thanks!

[Bassman59]

Dual primaries are provided so you can use the transformer with 120 V or 240 V mains. For the former, you parallel the primary windings, and for the latter you put them in series. But yes, please follow the connection sheet, as Hammond suggests.

As for the dual secondary. Do you want a single +30 V supply or a +/- 15 V supply? A transformer with two secondary windings (or a single secondary winding with a center tap) will do both, you just have to wire it correct -- as per that connection sheet. But make sure you get a transformer with adequate output voltage.

[Kleinstein]

The nominal power is for using both primaries and both secondaries. One can still use the secondaries independent. This only means that one can not get the full power from just one secondary. If the transformer is something like 2x30 V and 120 VA, one can get about 2 A (AC) from each primary. If one uses only one primary one can not get the full power (e.g. 2x2 A) form that one winding, but there is no problem drawing 2A (or even 2.5 A) from one and 0 from the other.


To get 30 V out after the regulator is needs some 33 V DC at the filter cap. To get this is takes about a 28-30 V transformer. The filter cap will increase the voltage by up to a factor of 1.4 as the peak to RMS ratio. However there are possible losses like the rectifier, main voltage lower than nominal, ripple. With no load the voltage would rise to around 40- 45 V. So this may be to much for the LM338 if there is a short.

While filter caps makes the voltage to go up, the RMS current is higher than the DC current by about a factor of about 1.6 ( it can be a little more for a rung core transformer with a large cap). So for  3 A DC out it takes around 5 A AC current.

The LM338 like many other integrated regulator has limited power handling capability. For the LM338 the practical limit is somewhere around 70 W, more is difficult and would need a really large heat sink / fan.

For a variable supply one should consider ways to reduce the heat loss, e.g. by using 2 or more transformer windings depending on the needed voltage.

[floobydust]

It's complicated. Consider a full-wave bridge rectifier as the most efficient for transformer utilization.
Working backwards, for a 30V output the LM338 would like 3-5V headroom so 33-35Vin.
Example 1182M12 24VCT 3.33A 80VA transformer gives about 32VDC under full load and 35VDC under no-load, and good for 2A DC.

But note the IC's available output current drops Fig. 2 with high >30V input-output differentials, due to SOA protection, so at a 5V output you it would get only around 1.2A with >30V input. The IC has to drop over 25V and it can't do that at full current.

This is why (tap) switching to lower voltage transformer winding is done.
If you DPDT switched the transformer secondaries from 24V to 12V, you could get 14.6VDC under full load and 16.2VDC under no-load, so I would expect around 12VDC max. output voltage and good for 4.1A DC.

My numbers are just rough calculations, AC mains varies and can be 115V or 120V and I used 9.7% regulation spec for the 80VA transformer.

[LaserTazerPhaser]

Instead of transformers a fused capacitor dropper with mov across the film cap can be fullwave rectified then filtered and fed into a buck converter. The output power is much greater and this arrangement will output much less heat and be more compact than a transformer.

[Jwillis]

It's complicated. Consider a full-wave bridge rectifier as the most efficient for transformer utilization.
Working backwards, for a 30V output the LM338 would like 3-5V headroom so 33-35Vin.
Example 1182M12 24VCT 3.33A 80VA transformer gives about 32VDC under full load and 35VDC under no-load, and good for 2A DC.

But note the IC's available output current drops Fig. 2 with high >30V input-output differentials, due to SOA protection, so at a 5V output you it would get only around 1.2A with >30V input. The IC has to drop over 25V and it can't do that at full current.

This is why (tap) switching to lower voltage transformer winding is done.
If you DPDT switched the transformer secondaries from 24V to 12V, you could get 14.6VDC under full load and 16.2VDC under no-load, so I would expect around 12VDC max. output voltage and good for 4.1A DC.

My numbers are just rough calculations, AC mains varies and can be 115V or 120V and I used 9.7% regulation spec for the 80VA transformer.

I just happen to have a 30V toroidal on my current project and its putting out after rectification 44V.I ball park the out put with the 70% rule for E cores which would work out as
AC output of transformer multiplied by 1.4 equals DC voltage after full bridge rectification . Toroids are more efficient so the out put is going to be higher than a E core  so a ball park figure would be more like 1.46 to 1.5 times more voltage after rectification. Consider the peak to peak voltage minus losses after rectification.Just to be clear this is an estimation but it does come close to actual measurements.

[Damianos]

Hello,

I would like to design and build a dual output power supply that can produce 30v at up to 5amps per output.  (though 3 amp is easily probably enough)

I was looking at the Hammond MFG toroidal transformers, but on their datasheet it says the following:

The transformers have dual primary and secondary windings.

Note: Units are designed to have all windings engaged (either series or parallel
connected) or connected as an autotransformer. Connection sheet included with
transformer.

Does this mean I cannot use the two secondary windings independently as two voltage sources?  Or is it just talking about the primary windings should be connected in either parallel or in series and not only one winding connected?
...

In the datasheet they specify only this:

Hi-Pot test of 4,000VAC RMS between primary & secondary

So there is not defined the insulation between primaries or secondaries (more important for independent power supplies).
If they are connected in series or parallel, as the manufacturer suggests, there is not any voltage difference between them.
If you make a dual supply, be sure that is not applied a high voltage difference between them, to be safe.
For example, don't consider them as fully floated and connect them in different parts of a circuit, that the voltage difference is more than 60V (I'm "guessing" here and keep the number low to be safe!).

[dcbrown73]

It's complicated. Consider a full-wave bridge rectifier as the most efficient for transformer utilization.
Working backwards, for a 30V output the LM338 would like 3-5V headroom so 33-35Vin.
Example 1182M12 24VCT 3.33A 80VA transformer gives about 32VDC under full load and 35VDC under no-load, and good for 2A DC.

But note the IC's available output current drops Fig. 2 with high >30V input-output differentials, due to SOA protection, so at a 5V output you it would get only around 1.2A with >30V input. The IC has to drop over 25V and it can't do that at full current.

This is why (tap) switching to lower voltage transformer winding is done.
If you DPDT switched the transformer secondaries from 24V to 12V, you could get 14.6VDC under full load and 16.2VDC under no-load, so I would expect around 12VDC max. output voltage and good for 4.1A DC.

My numbers are just rough calculations, AC mains varies and can be 115V or 120V and I used 9.7% regulation spec for the 80VA transformer.

I just happen to have a 30V toroidal on my current project and its putting out after rectification 44V.I ball park the out put with the 70% rule for E cores which would work out as
AC output of transformer multiplied by 1.4 equals DC voltage after full bridge rectification . Toroids are more efficient so the out put is going to be higher than a E core  so a ball park figure would be more like 1.46 to 1.5 times more voltage after rectification. Consider the peak to peak voltage minus losses after rectification.Just to be clear this is an estimation but it does come close to actual measurements.

So, based on this.  I should probably be looking at a 24v version of the toroid transformer?    A 30v version would push me way over the 40v max of a LM338.

[Kleinstein]

...
I just happen to have a 30V toroidal on my current project and its putting out after rectification 44V.I ball park the out put with the 70% rule for E cores which would work out as
AC output of transformer multiplied by 1.4 equals DC voltage after full bridge rectification . Toroids are more efficient so the out put is going to be higher than a E core  so a ball park figure would be more like 1.46 to 1.5 times more voltage after rectification. Consider the peak to peak voltage minus losses after rectification.Just to be clear this is an estimation but it does come close to actual measurements.

Toroid transformers are not that different. So about the same factor of 1.4.  The lower efficiency of EI core or small transformers lead to the no load voltage to be higher, especially with small transformers. So with low load the factor max reach 1.5. With heavy load the factor will drop to maybe 1.3 and due to ripple the useful voltage will be even lower.
The higher efficiency of ring core transformers leads to more peaked current, so power factor is a little smaller, so a slightly higher current rating is needed.

To stay below 40 V the transformer should be no higher than some 26 V (if > 50 VA).  So a 24 V transformer would be the next common value.

[schmitt trigger]

Instead of transformers a fused capacitor dropper with mov across the film cap can be fullwave rectified then filtered and fed into a buck converter. The output power is much greater and this arrangement will output much less heat and be more compact than a transformer.

Are you talking about a transformer-less capacitive supply?
If so, these devices don't provide powerline isolation.

[soldar]

I was looking at the Hammond MFG toroidal transformers, but on their datasheet it says the following:

The transformers have dual primary and secondary windings.

Note: Units are designed to have all windings engaged (either series or parallel
connected) or connected as an autotransformer. Connection sheet included with
transformer.

Does this mean I cannot use the two secondary windings independently as two voltage sources?  Or is it just talking about the primary windings should be connected in either parallel or in series and not only one winding connected?

I am quite sure the exact same question was asked not too long ago and it started a long thread on the topic. I probably posted in it but now I cannot find it for the life of me.  Maybe someone with better Google fu?

[rstofer]

One of the more credible attempts at building a PS.  It is worth reading about the prior incantations for which the author provides links.

https://www.eevblog.com/forum/projects/eez-h25005-a-possible-successor-of-eez-h24005-programmable-power-supply/

Dave did a video or three on power modules with digital displays  Ignore the smoke leak, that has been resolved


Search all of Dave's videos for 'Power Supply'.  I got 28 hits...

[soldar]

Have a look at this thread:
 https://www.eevblog.com/forum/beginners/transformer-question-167192/