+24V, +48V PSU from 48V Trafo

[Dualflip]

Normally when i do a PSU, i have a low voltage trafo, say 18V 0 18V  and with a doubler or tripler i get up to 50 something volts for my +48V psu. Now i have it the other way around, i have a big trafo and i need to lower for my 24V psu.

So rather than just injecting all of that voltage to the 24V psu and "dump" it into heat by the Voltage regulator dropping it down to 24V, and by that also risking that it excedes the max 40V input to output differential of the LM317.  The solution im thinking is using a standard bridge rectifier on the 48V psu, and only 2 diodes in series with the transformers legs that feed the 24V PSU, thus having a full wave rectifier but with half the voltage or so....

Is a very simple circuit, however  eventhough it makes sense to me,  and ive done it "on paper" several times, for some reason im unsure on how "audio worthy" or clean will that PSU be, as opposed to using a bridge rectifier.

Thoughts?

[Dualflip]

Bueller? Bueller? Bueller?

[PRR]

> i have a big trafo

OK, _how_ "big"? As big as a bus? 477 Volts? 57 pounds over ideal weight for its height?

> on: August 08, 2012, 01:53:28 pm
> on: August 08, 2012, 08:55:41 pm

And how much of a rush? This week? 7 hours? Yesterday?

[Dualflip]

The size of the trafo is in the title

no rush

Thanks ?

[jdbakker]

Is a very simple circuit, however  eventhough it makes sense to me,  and ive done it "on paper" several times, for some reason im unsure on how "audio worthy" or clean will that PSU be, as opposed to using a bridge rectifier.

So show us.

It may make sense to you, but from your (sparse) description I cannot picture any scheme that might come close to doing what you want. Post a schematic (hand-drawn is better than nothing), so at least we know what you're talking about.

JD 'a schematic says more than 1000 hand-wavings' B.

[joaquins]

With 48V (rms) you will have a peak of about 68V, so for a low current 48V (phantom) will work fine, but you don't need such high voltage, with a 36/40V trafo will be enogh, but beacouse it's low current you wont have troubles.

In the other hand, if you use a bridge rectifier or 2 diodes halfe wave rectifier you'll have de same DC voltage, but using half wave or the full wave, so you still have the 68V less 2V from the diodes so 66V, the LM317 wont be happy. if you have no other choice than use that trafo you can put a lot of diodes (about 40 instead of only 2) so each diode will disipate low power or a reversed 40V zenner in series with a 1n4007 but it will work for low currents... IMO it doesnt seem a good idea.

[Harpo]

The size of the trafo is in the title

Seems only the transformers secondary voltage rating is in the title. A transformer (EI-core, toroidal, autotransformer, ..) has some more data.
Is 'the trafo' singular? A little insane, but if you have two, maybe feed the primary windings of both transformers -so these exist, whatever VA these might be rated and whatever the load current might be- in series to get your secondary windings voltage down.

[ruffrecords]

If this PSU is meant to provide dc for the active electronics and 48V phantom power from the same secondary then there is another potential problem. For best noise performance,the 0V of the phantom power should be connected to the pins 1 at the mic input sockets and nowhere else then the signal 0V should be connected to the chassis where the mains enters. If you have a common transformer then the phantom 0V is already connected to the signal 0V within the PSU itself so you cannot do this.

Cheers

Ian

[ricardo]

I'm with JD and PRR.  Haven't a clue what you are talking about.  How about a pic?  Maybe we're stupid   

[Dualflip]

Im terribly sorry for not posting a schematic, here it is, what do you think?

Filter caps and voltage regulators are not shown, the caps before the 48V diode bridge are 220uf. Trafo is somewhere betwen 48-52V with center tap (26 0 26)

Definitely not the most efficient PSU design, Im aware that there will be a lot of wasted energy, since the voltage feeding both psus will still be high, but not high enough to melt the LM317. I could always take the DC voltage output (daisy chaing) of the 48V psu and just add an extra LM317 lowering it down to 24V that way i can have my 2 voltages, but that adds even more to the "brute force", since i dont have a high current demand for this particular psu, the 48V psu voltage regulator could be able to handle it with no big problems.

[ricardo]

Dual, have you built this?  The 24V supply is a conventional full wave supply.  But the '48V supply' will have less voltage as it receives the same rectified AC but via an extra diode & capacitor. ie the 48V supply dun wuk.  [EDIT] Rubbish

You may be better off arranging the normal voltage doubler chain to get your 48V  [EDIT] It IS a voltage doubler

Try taking the other end of the bridge rectifier, not to earth, but to the 24V supply capacitor.

[Harpo]

Looks just like Keith's 9K psu with fullwave C_ockcroft-Walton / Villard multiplier for the +48V part and two diodes left out for the left out negative rail.

[Dualflip]

Ricardo,yes the 24V has s standard full wave rectifier but it was just an idea or gimick to try and lower the rectified voltage as opposed to using a bridge.The 48V psu gets around 63 volts, why do you say it doesnt work? remember that the trannie is in fact 52V not 48V

I havent built this, just simulated it and it works. Exactly how do you suggest taking the other end of the bridge rectifier to which 24V supply cap? im not following...

Harpo, no idea what you are talking about....

[joaquins]

Sorry, my post was wrong, i didn't think you have a CT point, this trafo is called 24+24, 24-0-24 or 48CT, that's why my confussion. It should work as shown, but still have the problem of grounding, you should split grounds of 48V at the output of the bridge maybe.

Cheers

JS

[ricardo]

[EDIT]My apologies.  Your circuit is indeed a voltage doubler, giving over 70Vp on your '48V supply'.  I'll remove my doubting posts.

As with all voltage doubler/multipliers, regulation is poor so you'll do well to simulate your max. load current too.  Don't forget transformer resistance and supply capacitors.

taking the other end of the bridge rectifier, not to earth, but to the 24V supply capacitor

is taking the other end of the bridge rectifier to the '24V supply' output.  There will be a capacitor there that you haven't shown.  This will give even higher output voltage but you don't need this.

And YES.  This is a good way to get 24V & 48V from a single transformer.

[Harpo]

Harpo, no idea what you are talking about....

pg.5 from http://www.studio21.ch/diy/neeno/ssl9k/DIY_9k_Preamp.pdf
Some more on the Co.kcroft-Walton multipliers in http://users.tm.net/lapointe/coXkcroft_Walton.htm (substitute 'X' with 'c')

[Dualflip]

Harpo, no idea what you are talking about....

pg.5 from http://www.studio21.ch/diy/neeno/ssl9k/DIY_9k_Preamp.pdf
Some more on the Co.kcroft-Walton multipliers in http://users.tm.net/lapointe/coXkcroft_Walton.htm (substitute 'X' with 'c')

Yes, it looks very similar, i guess it must be a common design. That Co.kcroft-Walton link is very cool!

ricardo, i get what you mean by the cap on the 24V, this would raise the voltage considerably.

[PRR]

>>> i have a big trafo and i need to lower for my 24V psu.
>> OK, _how_ "big"?
> The size of the trafo is in the title

Um.

If you want to _solve_ a problem, you list the needs and resources. If you work alone, maybe you can keep the list in your head. If soliciting help from others, you want to lay-down most key details so everybody pulls the same way.

I still don't know what "big" means. Even the "48V" in the title is insufficient.

JD and Ricardo were willing to help but, like me, didn't have enough information for a good suggestion.

> 48-52V with center tap

The CT is vital information and should have gone on-stage sooner.

Ignoring the still-unknown current available from the transformer and required on the two outlets, there is a _simple_clean_ way to get what you want.

[Dualflip]

Thanks for the diagram PRR, its a lot more simple than mine and makes more sense.... Sorry for not laying out all of the details, since most of the transformers nowadays have a CT, I somehow assumed that by saying 48V it would be understood that it has a CT. I apollogize, the current needs are very small: 48V for phantom, no more than 10-15mA, and the 24V PSU would deliver no more than 300mA, another reason why i didnt consider it as relevant, again sorry for omitting the details.

Thanks again!

[PRR]

> 48V for phantom, no more than 10-15mA, and the 24V PSU would deliver no more than 300mA

That's an issue, though minor.

"My way", the 68V is somewhat more efficient than the 34V. The 68V uses all the winding on both cycles, the 34V uses half the winding on each cycle. That's not a big difference (much less than 2:1 as you might first think). Your transformer should be about 50% bigger (in VA) than if it were designed solely for a 34V/24V load.  If the load was one Phantom and a complete BBC-style 24V broadcast console and rack, then it might be bucks cheaper to right-size the 24V (these days with single 24V AC winding plus FWB) and make Phantom with smaller iron or multiplier.

300mA is not necessarily "small", but it is down in the range where transformers don't get cheaper when they get smaller, so you are probably OK.