85-265VAC and 24VDC combined power supply

[Eeyore]

Therefore I have to supply an average power of 40W/0.9 = 44.4W at the
internal power rail. Choosing 16V as the internal voltage when the AC
supply is used gives me an estimate for the transformer with

PTrans = 1.2 * 2.8A * 17 (2 * 0,5V for schottky) = 57,12VA

You can't do that. Because the load on the transformer is non-sinusoidal, the
rms load is greater than you imagine (see crest factor of waveforms).
http://en.wikipedia.org/wiki/Crest_factor

Because the peak power will only be used for short times I though about
using a 50VA toroidal transformer. The one which came to my mind is form
talema, the 70083 (approx 20$ for a single unit). Or is this is to low I
would go with the next bigger size rated at 80VA with a diameter of 90mm.

80VA would probably be OK.

Graham

 

[Phil Allison]

[email protected]>

Eeysore



Any sane person will consider that it's about a switched psu. And that the
turns ratio for 85VAC rms vs 24VDC will be too huge to be resonably
compensated by pulse width.

** Such babble could well be the voice of one possessed by a multitude

of evil demons .....

Rather use coil part A if 24VDC, and part A+B if input is 85VAC =>
120Vpeak.

** Those demons are sure having a real good party in there !!!

Maybe you should fart some energy into your manners.

** One dumbass Swede needs to stay off the weed, piss or mushrooms that he
has had far too much of.




......... Phil

 

[Fred Bloggs]

I am currently working on a device where I am evaluating options for the
design of the power supply. On the secondary side we need two power
supply output with 5V/4A and 13,8V/2A (uncommon voltage). The input
voltages are the standard 85V-265VAC range and 24VDC.

We want every unit to be capable of being used in both environments.
Therefore the power supply should allow both inputs. Because of 24VDC
input I will need two DC/DC converters at the output side. One from
24VDC/5VDC and one from 24VDC/13,8VDC. These converters are easy to
design and can have a high efficiency (>90-95%).

What I am unsure about is how to handle the transformation from the
primary 85V-265V to the intermediate internal voltage of 24V. A
regulated switched mode power supply is certainly a overkill because I
don't need the 24V to be regulated. So I thought I could use a standard
50/60Hz input stage with a simple full bridge rectifier and a good
transformer. Size is not the problem but power efficiency and costs are.
Or are there any better options? I searched for quite some time in
goggle for possible AC input stages but did not come up with a good
solution.

Sounds like it may just require a single isolated flyback with dual
outputs. The flyback primary is powered by both the rectified line AC
and the 24VDC input. The primary coil will be tapped with separate feeds
for the two DC inputs. You can use a standard IEC AC receptacle with
integral DPST switch that opens the 24VDC feed when a plug is inserted.
It will be important to open the GND connection of the 24VDC input when
the AC is applied to prevent shorting the rectifier.

 

[Fred Bloggs]

Have you considered an off-the-shelf universal input line to 24VDC
module? You can get a cULus etc. approved Meanwell 65W open frame
supply for USD 18. 1-9, and as you note the DC-DC converters can be
cookbook designs.

That sounds like the simplest way to do it, but he still needs an IEC
receptacle with integral switch to cut out the 24VDC feed to the output
of the Meanwell when the plug is inserted, break before make style. The
24VDC then drives two little cookbook buck regulators.

 

[MooseFET]

Hello,

I am currently working on a device where I am evaluating options for the
design of the power supply. On the secondary side we need two power
supply output with 5V/4A and 13,8V/2A (uncommon voltage). The input
voltages are the standard 85V-265VAC range and 24VDC.

We want every unit to be capable of being used in both environments.
Therefore the power supply should allow both inputs. Because of 24VDC
input I will need two DC/DC converters at the output side. One from
24VDC/5VDC and one from 24VDC/13,8VDC. These converters are easy to
design and can have a high efficiency (>90-95%).

What I am unsure about is how to handle the transformation from the
primary 85V-265V to the intermediate internal voltage of 24V. A
regulated switched mode power supply is certainly a overkill because I
don't need the 24V to be regulated. So I thought I could use a standard
50/60Hz input stage with a simple full bridge rectifier and a good
transformer. Size is not the problem but power efficiency and costs are.
Or are there any better options? I searched for quite some time in
goggle for possible AC input stages but did not come up with a good
solution.

265/85 = 3.12

It is not very hard to make a DC-DC converter work with in input that
varies over that much range. If your quantities are low, you won't
want a custom transformer in the mains to 24V section but lets assume
you are free to pick the tranformer you want.

Many parts max out at 35V so lets assume that 270V gives you just
about 30V DC
This makes the secondary have

(30V+0.7) / sqrt(2) = 21.7VAC

At 85VAC you would have:

(21.7 * 85/270) * sqrt(2) - 0.7 = 9.66VDC

I have left out the losses in the resistance of the windings at this
point.

Now you need to make a first cut on the design of the DC-DC
converters. I assume they don't have to be isolated so a simple
bucker can be used for the 5V. The 13.8 needs something like a SEPIC
converter.

Once you have the first cut on the DC-DC design you need a good
estimate of its efficiency at the 9.66V end of the span. You then
will know how much current the circuit will draw at that end and can
figure out a corrected value for the 9.66V and see if you are in
trouble.

 

[Fred Bloggs]

"Christian Walter the Wanker "






** You were clear - but unbelievably STUPID !!








** What about the fucking 85V to 265 VAC requirement ???????????


IMBECILE !!!!!!!!!!!!




........ Phil

By definition anyone who would ask a question of this newsgroup is an
imbecile.

 

[Phil Allison]

"MooseFET"


( snip load of putrid smelling, verbal diarrhoea )


** More very smelly Mooooooosseee droppings !!

Splaaaaaaattttttttttt .......




....... Phil

 

[Spehro Pefhany]

That sounds like the simplest way to do it, but he still needs an IEC
receptacle with integral switch to cut out the 24VDC feed to the output
of the Meanwell when the plug is inserted, break before make style. The
24VDC then drives two little cookbook buck regulators.

Maybe combine the two inputs with a dual Schottky, which gives
polarity reversal protection on the 24VDC input for free.

http://www.vishay.com/docs/93216/9321612c.pdf


Best regards,
Spehro Pefhany

 

[Phil Allison]

"Fred Bloggs"

By definition anyone who would ask a question of this newsgroup is an
imbecile.

** By definition - a genuine question asker does not arbitrarily decide
which is the right answer.

Only fucking TROLLS do that.





......... Phil

 

[Christian Walter]

You can't do that. Because the load on the transformer is non-sinusoidal, the
rms load is greater than you imagine (see crest factor of waveforms).
http://en.wikipedia.org/wiki/Crest_factor

Yes - at least I thought I do but obviously it do not-( That is the
reason why I added the factor of 1.2 but as you pointed out this does
not look correct. I have to admit that I have taking this formula from a
book and have not though about it (german book, p923, u. tietze -
halbleiterschaltungstechnik, 3540428496). Looking at it again the say
that the value spawns between 1,2 and 2 and they have chosen to use 1.5
in our of their examples. Thanks for pointing that out. In any case I
will have to think about that again because the load to seen by the
rectifier and the buffer cap is not a constant load.

Anyway the idea from Fred Bloggs sounds great by using a single flyback
converter with the tapped primary winding. It requires the smallest
amount of components. The only drawback is that it is probably the most
complicated solution (regarding development time and testing time) from
the ones presented so far.
I will start again to think about the open frame version because I think
the mechanical problem is solvable and the supplies are cheap. This is
probably the easiest way to go.

Thanks everybody so far for your help and valuable suggestions and comments,
Christian

 

[Michael A. Terrell]

By definition anyone who would ask a question of this newsgroup is an
imbecile.

Don't be modest fred. Anything you asked on SEB would still make you
an imbecile.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Phil Allison]

"Christian Walter"

Yes - at least I thought I do but obviously it do not-( That is the reason
why I added the factor of 1.2 but as you pointed out this does not look
correct. I have to admit that I have taking this formula from a book and
have not though about it (german book, p923, u. tietze -
halbleiterschaltungstechnik, 3540428496). Looking at it again the say that
the value spawns between 1,2 and 2 and they have chosen to use 1.5 in our
of their examples. Thanks for pointing that out. In any case I will have
to think about that again because the load to seen by the rectifier and
the buffer cap is not a constant load.

** ROTFLMAO !!


Fucking Kraut software engineers ............




........ Phil

 

[Michael A. Terrell]

That sounds like the simplest way to do it, but he still needs an IEC
receptacle with integral switch to cut out the 24VDC feed to the output
of the Meanwell when the plug is inserted, break before make style. The
24VDC then drives two little cookbook buck regulators.

No, he needs a pair of high current Schottky diodes from each power
source to isolate the two possible power sources. Do it simple, and do
it right. Switches on power connectors cause LOTS of field failures.
It will also allow a battery backup system, as long as the output of the
switcher is slightly higher than the 24 VDC source. I did this for an
'All channel emergency alert system' in a CATV headend 25 years ago.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[MooseFET]

That sounds like the simplest way to do it, but he still needs an IEC
receptacle with integral switch to cut out the 24VDC feed to the output
of the Meanwell when the plug is inserted, break before make style. The
24VDC then drives two little cookbook buck regulators.

BTW:
On a lot of HP equipment, there was a switch to switch to DC
operation.


If the mains to 24VDC is a fairly simple minded one, no switch is
needed. The output side can be held positive when the input side is
unpowered. All that is really needed is a diode to keep the supply
from powering the DC input.

 

[Eeyore]

That sounds like the simplest way to do it, but he still needs an IEC
receptacle with integral switch to cut out the 24VDC feed to the output
of the Meanwell when the plug is inserted, break before make style. The
24VDC then drives two little cookbook buck regulators.

I don't see any reason/need to 'cut' either supply. If they are diode OR'd it'll
work fine under all circumstances.

Graham

 

[Eeyore]

Yes - at least I thought I do but obviously it do not-( That is the
reason why I added the factor of 1.2 but as you pointed out this does
not look correct. I have to admit that I have taking this formula from a
book and have not though about it (german book, p923, u. tietze -
halbleiterschaltungstechnik, 3540428496). Looking at it again the say
that the value spawns between 1,2 and 2 and they have chosen to use 1.5
in our of their examples. Thanks for pointing that out. In any case I
will have to think about that again because the load to seen by the
rectifier and the buffer cap is not a constant load.

Anyway the idea from Fred Bloggs sounds great by using a single flyback
converter with the tapped primary winding. It requires the smallest
amount of components. The only drawback is that it is probably the most
complicated solution (regarding development time and testing time) from
the ones presented so far.

Yes exactly. It's neat but not entirely trivial to design.

I will start again to think about the open frame version because I think
the mechanical problem is solvable and the supplies are cheap. This is
probably the easiest way to go.

I'd agree with that.

Graham

 

[MooseFET]

No, he needs a pair of high current Schottky diodes from each power
source to isolate the two possible power sources. Do it simple, and do
it right. Switches on power connectors cause LOTS of field failures.
It will also allow a battery backup system, as long as the output of the
switcher is slightly higher than the 24 VDC source. I did this for an
'All channel emergency alert system' in a CATV headend 25 years ago.

With a few more parts, you can also make it work as a "float charger"
to keep the batteries up when the mains supply is there.

 

[Eeyore]

No, he needs a pair of high current Schottky diodes from each power
source to isolate the two possible power sources. Do it simple, and do
it right. Switches on power connectors cause LOTS of field failures.
It will also allow a battery backup system, as long as the output of the
switcher is slightly higher than the 24 VDC source. I did this for an
'All channel emergency alert system' in a CATV headend 25 years ago.

Goodness. You said something sensible.

Graham

 

[Rich Grise]

I forgot too mention. There can be a switch for selecting the 110V/230V
variant and I will connect the primary windings in parallel or in
series. Customers either use 110V and 24V/DC or 230V and 24V/DC.

This simplifies things dramatically, if you're allowed to have a 120/240
switch. Then just use a transformer with two primaries, and connect them
in series for 240 and parallel for 120, and just an ordinary rectifier/
cap for your 24VDC, which you say doesn't have to be regulated.

Covering the whole range from 85-250VAC without some kind of switch
is beyond my skills. )-;

Good Luck!
Rich

 

[Jamie]

Hello,

I am currently working on a device where I am evaluating options for the
design of the power supply. On the secondary side we need two power
supply output with 5V/4A and 13,8V/2A (uncommon voltage). The input
voltages are the standard 85V-265VAC range and 24VDC.

We want every unit to be capable of being used in both environments.
Therefore the power supply should allow both inputs. Because of 24VDC
input I will need two DC/DC converters at the output side. One from
24VDC/5VDC and one from 24VDC/13,8VDC. These converters are easy to
design and can have a high efficiency (>90-95%).

What I am unsure about is how to handle the transformation from the
primary 85V-265V to the intermediate internal voltage of 24V. A
regulated switched mode power supply is certainly a overkill because I
don't need the 24V to be regulated. So I thought I could use a standard
50/60Hz input stage with a simple full bridge rectifier and a good
transformer. Size is not the problem but power efficiency and costs are.
Or are there any better options? I searched for quite some time in
goggle for possible AC input stages but did not come up with a good
solution.

Kind regards,
Christian Walter

We buy devides made in France and most likely others do it also that
accepts 24V up to 260 V AC/DC on the same input.
No switching of selections required.