Home

 
From 24 volts to 5 volts, which is best most economical method...
(<-- Back to message list)

lcdpublishing
Sat, 09 Aug 08 11:59:10 +0000
Hi guys,

I am trying to figure out the best way to deal with various power
levels for a project. At present I need two voltage levels, 24Vdc
and 5Vdc for the logic. From our past discussions, I understand it
is a bad idea to use a 7805 to take 24Vdc down to 5.

On previous projects like this, I would use various power supplies to
create 24, 12, and 5 volts etc. However, I would like to try and get
away from that sort of thinking and just end up building everything
with a 24 volt supply and from there create what I need.

I have run across DC-to-DC converters in some of the catalogs which
(without research) seem to do this. The one issue is that they seem
very expensive for what they do.

I know there has to be a "typical" and economical method to do this
as so many devices must already do this. I just am not sure which is
the way to go - any suggestions?

Thanks
Chris



16 replies folowing
Nelson Johnsrud
Sat, 09 Aug 08 13:52:31 +0000
Would an emitter-follower stage pass enough current for your 5VDC? Maybe a
3055? Seems you could pull 5V directly off the emitter, or at least cut it
down to a lower level that the 7805 could deal with. I guess it depends on
how much current you are asking that transistor to pass.

Bear in mind I have very little experience designing transistor circuits,
but I just did a similar circuit for a project here on the bench this week.
I will follow this thread to see what solutions are proposed for my own
reference. Maybe my idea is full of holes (or something).


Nels Johnsrud
Manitowoc WI
gop4evr@lsol.net


> From: lcdpublishing
> Reply-To:
> Date: Sat, 09 Aug 2008 11:59:10 -0000
> To:
> Subject: [Electronics_101] From 24 volts to 5 volts, which is best & most
> economical method...
>
> Hi guys,
>
> I am trying to figure out the best way to deal with various power
> levels for a project. At present I need two voltage levels, 24Vdc
> and 5Vdc for the logic. From our past discussions, I understand it
> is a bad idea to use a 7805 to take 24Vdc down to 5.
>
> On previous projects like this, I would use various power supplies to
> create 24, 12, and 5 volts etc. However, I would like to try and get
> away from that sort of thinking and just end up building everything
> with a 24 volt supply and from there create what I need.
>
> I have run across DC-to-DC converters in some of the catalogs which
> (without research) seem to do this. The one issue is that they seem
> very expensive for what they do.
>
> I know there has to be a "typical" and economical method to do this
> as so many devices must already do this. I just am not sure which is
> the way to go - any suggestions?
>
> Thanks
> Chris
>
>
>
> ------------------------------------
>
> Please trim excess when replyingYahoo! Groups Links
>
>
>



dicsEE
Sat, 09 Aug 08 14:08:52 +0000
 
Ron Wright
Sat, 09 Aug 08 14:08:52 +0000

Nwl,

Ron Wright, N9EE
Nel,

This is known as a pass transitor where the larger transistor is driven
at the base with whatever voltage you have, the collector is driven by
the voltage & current supply and the output is the emitter. Used in
most linear power supplies.

The problem is there is a .6 volt drop from the bast to emitter...that
is the emitter is 0.6 volts lower than the base. So if driving base
with 7805 5V source the emitter will put out 4.4 volts.

Putting a diode in the 7805 ground lead can offset this making the 7805
put out 5.6 volts. The ground lead draws little current so about any PN
diode will work. You would have to isoltate the 7805 case for it is
ground. Some versions of 7805 have insulated cases with good material
for heat conduction so isolating their grounds is already done.

73, ron, n9ee/r


727-376-6575

MICRO COMPUTER CONCEPTS

Owner 146.64 repeater Tampa Bay, FL

No tone, all are welcome.




On Sat, Aug 9, 2008 at 9:52 AM, Nelson Johnsrud wrote:

Would an emitter-follower stage pass enough current for your 5VDC?
Maybe a
3055? Seems you could pull 5V directly off the emitter, or at least cut
it
down to a lower level that the 7805 could deal with. I guess it depends
on
how much current you are asking that transistor to pass.

Bear in mind I have very little experience designing transistor
circuits,
but I just did a similar circuit for a project here on the bench this
week.
I will follow this thread to see what solutions are proposed for my own
reference. Maybe my idea is full of holes (or something).


Nels Johnsrud
Manitowoc WI
gop4evr@lsol. net

> From: lcdpublishing < lcdpublishing@ yahoo.com
> > Reply-To: < Electronics_
> 101@yahoogroups. com > Date:
> Sat, 09 Aug 2008 11:59:10 -0000 To: < Electronics_ 101@yahoogroups.
> com > Subject: [Electronics_
> 101] From 24 volts to 5 volts, which is best & most economical
> method...
> Hi guys,
> I am trying to figure out the best way to deal with various power
> levels for a project. At present I need two voltage levels, 24Vdc and
> 5Vdc for the logic. From our past discussions, I understand it is a
> bad idea to use a 7805 to take 24Vdc down to 5.
> On previous projects like this, I would use various power supplies to
> create 24, 12, and 5 volts etc. However, I would like to try and get
> away from that sort of thinking and just end up building everything
> with a 24 volt supply and from there create what I need.
> I have run across DC-to-DC converters in some of the catalogs which
> (without research) seem to do this. The one issue is that they seem
> very expensive for what they do.
> I know there has to be a "typical" and economical method to do this as
> so many devices must already do this. I just am not sure which is the
> way to go - any suggestions?
> Thanks Chris
>
>
> ------------ --------- --------- ------
> Please trim excess when replyingYahoo! Groups Links
>
>






[Non-text portions of this message have been removed]

Shawn Upton
Sat, 09 Aug 08 14:16:07 +0000
I recall the LM317 datasheet having some SMPS circuit in it; you could look at that. Alternatively, pick an SMPS buck convertor, and start using that. If you pick a wide output range part, odds are the output voltage is determined once again by a simple resistor network. Or you could just buy some fixed voltage versions (say, 5V and 12V) and keep them on hand. Lastly, I just found TI's DCR series of convertors. 1W and 2W versions, kinda pricey but at low powers they are quite small and require fewer parts.

Oh, and I recall seeing some 78xx convertors on the market. Meant to replace 78xx regulators, so three lead; but they are actually self contained SMPS's. You might want to look at those too.

As for best and most economical, I'm going to go on a limb and say "it depends". If you just need say 5mA at 3V, sometimes the best thing to do is to put a small high wattage resistor in front of a simple linear voltage regulator and dissapate the heat. Cheap, effective, and probably not that inefficent in the scheme of things, if the 24V line is supplying an amp to something else. Or if you have a metal box, with good thermal characteristics, again it might be cheapest/easiest to just dissapate the heat. The SMPS's have the dissadvantage of either being parts-heavy, or noisy, which may impact your design. For one-off projects, parts heavy is probably fine, as the cost isn't that bad; and one must consider the cost of time if one is trying to determine how to make something as cheap as possible, versus getting it done quickly or perhaps working correctly the first time.

Shawn Upton, KB1CKT



----- Original Message ----
From: lcdpublishing
To: Electronics_101@yahoogroups.com
Sent: Saturday, August 9, 2008 7:59:10 AM
Subject: [Electronics_101] From 24 volts to 5 volts, which is best & most economical method...


Hi guys,

I am trying to figure out the best way to deal with various power
levels for a project. At present I need two voltage levels, 24Vdc
and 5Vdc for the logic. From our past discussions, I understand it
is a bad idea to use a 7805 to take 24Vdc down to 5.






[Non-text portions of this message have been removed]

Ron Wright
Sat, 09 Aug 08 14:18:47 +0000

Chris,

When asking about power one needs to know what voltage AND current. A
design for 5V at 100 ma would be totally different than 5V at 5 amps.

What current are you looking for on the 5V.

7805s can be used up to 35 volt input, but since the power in the 7805
will be:

Power going into heat = (the input voltage - 5V) x current

7805 can handle 1 A, but for (24-5) x 1A = 19 watts of heat so some heat
sinking is required. Now if say 200 ma then 3.8 W, notable less.

If 5 Amp then a different design.

73, ron, n9ee/r

Ron Wright, N9EE

727-376-6575

MICRO COMPUTER CONCEPTS

Owner 146.64 repeater Tampa Bay, FL

No tone, all are welcome.




On Sat, Aug 9, 2008 at 7:59 AM, lcdpublishing wrote:

Hi guys,

I am trying to figure out the best way to deal with various power
levels for a project. At present I need two voltage levels, 24Vdc
and 5Vdc for the logic. From our past discussions, I understand it
is a bad idea to use a 7805 to take 24Vdc down to 5.

On previous projects like this, I would use various power supplies to
create 24, 12, and 5 volts etc. However, I would like to try and get
away from that sort of thinking and just end up building everything
with a 24 volt supply and from there create what I need.

I have run across DC-to-DC converters in some of the catalogs which
(without research) seem to do this. The one issue is that they seem
very expensive for what they do.

I know there has to be a "typical" and economical method to do this
as so many devices must already do this. I just am not sure which is
the way to go - any suggestions?

Thanks
Chris




[Non-text portions of this message have been removed]

Shawn Upton
Sat, 09 Aug 08 14:25:25 +0000
I'd bias the '3055 base to say 8.2V, then the 7805 would see around 7.6V, which should keep it in regulation. The problem is, the 2N3055 has a low beta; I don't have the datasheet but it's probably around 10, maybe 20 at a 1A collector current. If it's 20, then the base current needs to be around 50mA at full load. I don't know what kind of load regulation the 24V supply has, but I'll assume that the error budget would probably still have around 10mA or so into the zener at full tilt. That means around 60mA into the zener under no load; that gives around 1/2W Pd. The resistor to the 2N3055 base would be around 270 ohms (or a step down); this resistor would be dissapating 1W normally, so a 2 or 3W resistor would be required. Lastly, the 2N3055 will need a good heatsink, as it would be dissapating over 16W at a 1A load. You'd need a heatsink with a thermal impedance of under 7 degrees C per Watt, in order to keep the junction temp below 150C.

Shawn Upton, KB1CKT



----- Original Message ----
From: Nelson Johnsrud
To: Electronics 101
Sent: Saturday, August 9, 2008 9:52:31 AM
Subject: Re: [Electronics_101] From 24 volts to 5 volts, which is best & most economical method...


Would an emitter-follower stage pass enough current for your 5VDC? Maybe a
3055? Seems you could pull 5V directly off the emitter, or at least cut it
down to a lower level that the 7805 could deal with. I guess it depends on
how much current you are asking that transistor to pass.

Bear in mind I have very little experience designing transistor circuits,
but I just did a similar circuit for a project here on the bench this week.
I will follow this thread to see what solutions are proposed for my own
reference. Maybe my idea is full of holes (or something).


Nels Johnsrud
Manitowoc WI
gop4evr@lsol. net

_,___




[Non-text portions of this message have been removed]

rtstofer
Sat, 09 Aug 08 14:44:16 +0000
--- In Electronics_101@yahoogroups.com, "lcdpublishing"
wrote:

>
> Hi guys,
>
> I am trying to figure out the best way to deal with various power
> levels for a project. At present I need two voltage levels, 24Vdc
> and 5Vdc for the logic. From our past discussions, I understand it
> is a bad idea to use a 7805 to take 24Vdc down to 5.
>
> On previous projects like this, I would use various power supplies to
> create 24, 12, and 5 volts etc. However, I would like to try and get
> away from that sort of thinking and just end up building everything
> with a 24 volt supply and from there create what I need.
>
> I have run across DC-to-DC converters in some of the catalogs which
> (without research) seem to do this. The one issue is that they seem
> very expensive for what they do.
>
> I know there has to be a "typical" and economical method to do this
> as so many devices must already do this. I just am not sure which is
> the way to go - any suggestions?
>
> Thanks
> Chris
>


There are really just two basic power supply designs: a series losser
and a switching mode design.

No matter how you get to a series losser, 2N3055, LM7805, whatever,
the power dissipation is grim. Somewhere in the system you are
dropping 24V - 5V or 19V times the load current. If you need 1A, you
are losing 19W - a ridiculous design even if it would work.

A switching mode supply is just exactly what the DC-DC converters are.
As you say, they are expensive.

Your 24V supply is going to be trashed with motors and solenoids so it
is probably not a good source for a logic supply anyway.

What's wrong with a nice little 5V wall wart? I use them all the time
for systems up to about 2A (mine are all MUCH less than this). I use
the SWPS Switching Power Supply here:
http://www.digilentinc.com/Products/Catalog.cfm?Nav1=Products&Nav2=Components&Cat=Component
(near the bottom of the page).

If you really want to build a SMPS, go for it. There are plenty of
app notes around. Or, break open that wall wart and see how they did it.

Richard


John Popelish
Sat, 09 Aug 08 15:01:15 +0000
A switching buck regulator would be an efficient way to get 5 volts
from 24. There are many integrated circuits that perform this task,
with an additional inductor and a couple capacitors and sometimes an
additional diode an transistor.
http://www.narwani.net/neil/tech/elec/78SWxx/index.html

There is also a a simple circuit that works, except that it doesn't
have over current protection:
http://www.romanblack.com/smps/smps.htm

I have also seen several totally self contained buck regulators that
plug in as if they were an LM7805.
http://www.ferret.com.au/c/Soanar-Electronic-Component-Supply-Chain-Solutions/R-78xx-0-5-series-switching-regulator-available-from-Soanar-n810799

Regards,

John Popelish

Ron Wright
Sat, 09 Aug 08 17:05:24 +0000

A company I buy surplus stuff from here in Florida, MJPA.com, has
110VAC, 5V at 5 amp supplies for $12.95, new in the box.

Again we need to know what current we are working with on the 5V. Until
this is known one cannot give concrete advise for a design or way to go.

73, ron, n9ee/r


Ron Wright, N9EE

727-376-6575

MICRO COMPUTER CONCEPTS

Owner 146.64 repeater Tampa Bay, FL

No tone, all are welcome.




On Sat, Aug 9, 2008 at 10:44 AM, rtstofer wrote:

--- In Electronics_ 101@yahoogroups. com
, "lcdpublishing"
wrote:

>
> Hi guys,
> I am trying to figure out the best way to deal with various power
> levels for a project. At present I need two voltage levels, 24Vdc and
> 5Vdc for the logic. From our past discussions, I understand it is a
> bad idea to use a 7805 to take 24Vdc down to 5.
> On previous projects like this, I would use various power supplies to
> create 24, 12, and 5 volts etc. However, I would like to try and get
> away from that sort of thinking and just end up building everything
> with a 24 volt supply and from there create what I need.
> I have run across DC-to-DC converters in some of the catalogs which
> (without research) seem to do this. The one issue is that they seem
> very expensive for what they do.
> I know there has to be a "typical" and economical method to do this as
> so many devices must already do this. I just am not sure which is the
> way to go - any suggestions?
> Thanks Chris


There are really just two basic power supply designs: a series losser
and a switching mode design.

No matter how you get to a series losser, 2N3055, LM7805, whatever,
the power dissipation is grim. Somewhere in the system you are
dropping 24V - 5V or 19V times the load current. If you need 1A, you
are losing 19W - a ridiculous design even if it would work.

A switching mode supply is just exactly what the DC-DC converters are.
As you say, they are expensive.

Your 24V supply is going to be trashed with motors and solenoids so it
is probably not a good source for a logic supply anyway.

What's wrong with a nice little 5V wall wart? I use them all the time
for systems up to about 2A (mine are all MUCH less than this). I use
the SWPS Switching Power Supply here:
http://www.digilent inc.com/Products /Catalog. cfm?Nav1= Products&
Nav2=Components& Cat=Component

(near the bottom of the page).

If you really want to build a SMPS, go for it. There are plenty of
app notes around. Or, break open that wall wart and see how they did
it.

Richard






[Non-text portions of this message have been removed]

Shawn Upton
Sat, 09 Aug 08 17:25:11 +0000
I believe he is asking in general terms. He doesn't have a particular project in mind, at least not yet. But he's probably tired of having to come up with 5 different power supplies on each and every project, and wondering what kind of cost it will be if he just sticks to one supply line, converting as necessary, to cut down on wiring and/or design steps.

Shawn Upton, KB1CKT



----- Original Message ----
From: Ron Wright
To: Electronics_101@yahoogroups.com
Sent: Saturday, August 9, 2008 1:05:24 PM
Subject: RE: [Electronics_101] Re: From 24 volts to 5 volts, which is best & most economical method...



A company I buy surplus stuff from here in Florida, MJPA.com, has
110VAC, 5V at 5 amp supplies for $12.95, new in the box.

Again we need to know what current we are working with on the 5V. Until
this is known one cannot give concrete advise for a design or way to go.

73, ron, n9ee/r

Ron Wright, N9EE

___




[Non-text portions of this message have been removed]

James M.\(Jim\) Geidl
Sat, 09 Aug 08 20:38:43 +0000
Ron,

Are you sure the URL is correct, I get some travel seminar company. It
tried a Google on MJPA and got bupkis.

Jim, K6JMG
//


> [mailto:Electronics_101@yahoogroups.com] On Behalf Of Ron Wright
> Sent: Saturday, August 09, 2008 10:05 AM
> To: Electronics_101@yahoogroups.com
> Subject: RE: [Electronics_101] Re: From 24 volts to 5 volts,
> which is best & most economical method...
>
>
> A company I buy surplus stuff from here in Florida, MJPA.com,
> has 110VAC, 5V at 5 amp supplies for $12.95, new in the box.
>
> Again we need to know what current we are working with on the
> 5V. Until this is known one cannot give concrete advise for
> a design or way to go.
>
> 73, ron, n9ee/r
>
>
> Ron Wright, N9EE
>
> 727-376-6575
>
> MICRO COMPUTER CONCEPTS
>
> Owner 146.64 repeater Tampa Bay, FL
>
> No tone, all are welcome.


rtstofer
Sat, 09 Aug 08 20:45:50 +0000
--- In Electronics_101@yahoogroups.com, "James M.\(Jim\) Geidl"
wrote:

>
> Ron,
>
> Are you sure the URL is correct, I get some travel seminar company. It
> tried a Google on MJPA and got bupkis.
>
> Jim, K6JMG


http://www.mpja.com/

Richard


Roy J. Tellason
Sat, 09 Aug 08 21:52:48 +0000
On Saturday 09 August 2008 07:59, lcdpublishing wrote:
> Hi guys,
>
> I am trying to figure out the best way to deal with various power
> levels for a project. At present I need two voltage levels, 24Vdc
> and 5Vdc for the logic. From our past discussions, I understand it
> is a bad idea to use a 7805 to take 24Vdc down to 5.
>
> On previous projects like this, I would use various power supplies to
> create 24, 12, and 5 volts etc. However, I would like to try and get
> away from that sort of thinking and just end up building everything
> with a 24 volt supply and from there create what I need.
>
> I have run across DC-to-DC converters in some of the catalogs which
> (without research) seem to do this. The one issue is that they seem
> very expensive for what they do.
>
> I know there has to be a "typical" and economical method to do this
> as so many devices must already do this. I just am not sure which is
> the way to go - any suggestions?


I've seen basically two approaches taken to dealing with something like this.
One is to have a transformer that has more than one secondary winding to give
you a power supply with a couple of different output voltages. One of the
two VCRs I just tore apart had that in it. The other is to use a switcher
with multiple outputs, which is what the other one has.

Those dc-dc converters seem to me like they'd be practical where you had a
whole big system running off one voltage and wanted to get another voltage
only in one small portion of it and at not very much power. For a couple of
voltages that are used all over a system, which is what it sounds like
you're describing here, I'd go with one or the other approach in the power
supply itself.

Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
Roy J. Tellason
Sat, 09 Aug 08 21:55:00 +0000
On Saturday 09 August 2008 09:52, Nelson Johnsrud wrote:
> Would an emitter-follower stage pass enough current for your 5VDC? Maybe a
> 3055? Seems you could pull 5V directly off the emitter, or at least cut it
> down to a lower level that the 7805 could deal with. I guess it depends on
> how much current you are asking that transistor to pass.


Either way, though, you still end up dissipating the voltage difference as a
whole lot of heat...

> Bear in mind I have very little experience designing transistor circuits,
> but I just did a similar circuit for a project here on the bench this week.
> I will follow this thread to see what solutions are proposed for my own
> reference. Maybe my idea is full of holes (or something).


Tying this briefly into the zener diode discussion that just went by, that's
exactly the approach used to boost the current handling capability, rather
than going with higher-power zener diodes. It strikes me as being practical
for relatively small currents, say up to an amp or two at most, but beyond
that it gets really inefficient.

Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
lcdpublishing
Sat, 09 Aug 08 23:22:07 +0000
Hi guys,

THanks much for replying - it seems as though there are a number of
view points on this and I guess it bears deeper consideration.
First, let me give an example....

Let's say I am making a PLC type controller. It will typically be
used with 24VDC on all of it's connections (switches, relays, lamps,
etc.). The logic chips running the PLC would be using 5 or 3.3
volt. I would like to simply have one power supply required which
would provide 24 volts. Inside my project I would like to create the
5 volts for logic.

So, typically with the gizmos I build, I have logic at 5 volts
and "controlled" items at 24 volts. The current draw has typically
been in the 300 to 600 mA range of my logic boards - but this is a
rough number.

Does that make sense in what I am trying to do?

Chris


Roy J. Tellason
Sun, 10 Aug 08 00:02:42 +0000
On Saturday 09 August 2008 19:22, lcdpublishing wrote:
> So, typically with the gizmos I build, I have logic at 5 volts
> and "controlled" items at 24 volts. The current draw has typically
> been in the 300 to 600 mA range of my logic boards - but this is a
> rough number.


Sounds to me like you should be okay with a 1A supply, giving you a bit of
headroom there...

Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
John Popelish
Sun, 10 Aug 08 02:38:23 +0000
lcdpublishing wrote:
> Hi guys,
>
> THanks much for replying - it seems as though there are a number of
> view points on this and I guess it bears deeper consideration.
> First, let me give an example....
>
> Let's say I am making a PLC type controller. It will typically be
> used with 24VDC on all of it's connections (switches, relays, lamps,
> etc.). The logic chips running the PLC would be using 5 or 3.3
> volt. I would like to simply have one power supply required which
> would provide 24 volts. Inside my project I would like to create the
> 5 volts for logic.
>
> So, typically with the gizmos I build, I have logic at 5 volts
> and "controlled" items at 24 volts. The current draw has typically
> been in the 300 to 600 mA range of my logic boards - but this is a
> rough number.
>
> Does that make sense in what I am trying to do?


Any linear regulation scheme that produces 600mA at low
volts from a 24 volt source must consume a total of
600mA*24V=14.4 watts, with most of that coming out of the
regulator. If the 24 volt supply is big and you have room
for a few cubic inches of heat sink, then a linear regulator
is simple.

But a switching regulator with 80% efficiency would produce
5 volts @ 600mA while drawing only about 156mA from the 24
volt supply and would have to get rid of only about 0.6
watts of heat. You probably don't need the input to output
isolation of a DC to DC converter, but only a buck
regulator, which is a lot cheaper.

Now, if you get the logic supply down to below 100 mA then
the power wasted by the linear regulator gets pretty small
and the added complexity and expense of the switcher maybe
harder to justify. For that case, you are talking about the
linear wasting only about 1.9 watts and needing only a
couple square inches of heat sink.

Regards,

John Popelish