# stable 24vdc power supply with low heat

#### bexy

Joined Apr 2, 2020
22
i am trying to make a 24vdc power supply, sounds simple but the 240vac to 24vdc transformer i do not have any option to use has an output that can get up to 40vac also i need to make the common line common to both ac and dc so i can not use a full bridge but can only use a single diode, so i loose half the wave and use a 1000uf cap to pull it up again, on the plus side i only need 120ma. SO using a 7824 in the usual arrangement i get my voltage but even with a heat sink i am getting 25 Deg C rise on the heat sink, due i think to the 7824 havig to drop 40-24 v so 16v. basically although this works i hate heat and consider the higher the temp the shorted life span the power supply will have.
SO any one got any idea's how i can make a simple power supply that will not get hot?

#### AnalogKid

Joined Aug 1, 2013
10,172
Any linear regulator topology will have to get rid of 1.92 W somehow. You can reduce the heatsink size by putting a power resistor in series with the input of the 7824.

You say the AC source can rise to 40 V. If that is 40 Vrms, then the peak value is over 56 V, way too high for a standard 7824. Also, what is the minimum AC voltage source value? This is what is used to calculator the series resistor value.

ak

#### crutschow

Joined Mar 14, 2008
31,111
You can reduce the loss by using a switching regulator instead of a linear regulator.

#### Irving

Joined Jan 30, 2016
3,187
What's the reason for the common AC/DC line? There are potentially several ways around that if we understood your thinking...

Joined Feb 20, 2016
4,251

#### Deleted member 115935

Joined Dec 31, 1969
0
i am trying to make a 24vdc power supply, sounds simple but the 240vac to 24vdc transformer i do not have any option to use has an output that can get up to 40vac also i need to make the common line common to both ac and dc so i can not use a full bridge but can only use a single diode, so i loose half the wave and use a 1000uf cap to pull it up again, on the plus side i only need 120ma. SO using a 7824 in the usual arrangement i get my voltage but even with a heat sink i am getting 25 Deg C rise on the heat sink, due i think to the 7824 havig to drop 40-24 v so 16v. basically although this works i hate heat and consider the higher the temp the shorted life span the power supply will have.
SO any one got any idea's how i can make a simple power supply that will not get hot?
https://www.digikey.co.uk/product-d.../LD05-23B24WR2/2725-LD05-23B24WR2-ND/13968553

#### DickCappels

Joined Aug 21, 2008
9,508
It is hard to imagine that a 25° temperature rise will shorten the life of your power supply significantly. Check the datasheets for your components.

I just noticed that I can buy a 24V, 3A open frame power supply from my local distributor for about \$12 + shipping. Maybe buying a power supply is a better solution for your application. At least that way when it finally dies, it will be somebody else's fault

Seriously, in the development of many products, I would have been pleased with only a 25° temperature rise.

#### bexy

Joined Apr 2, 2020
22
thanks for the comments, the comment about the peak voltage made me think i may have use a cap that was to high and i was inadvertantly lifting it up higher, so i reduced it to 100uf, that helped a little bit. i am not sure what rating the transformer is apart from 24vac 30va, ther reason i am using this one is its the only small DIN mount version i can find, this is because all devices used in this system are to fit side by side using the same profile as a standard circuit breaker or multipes of it, for instance the power supply is a 3 mod so 3 times the width of a single phase circuit breaker. so unless i can find another i am stuck with it, yes there are 240vac to 24vdc DIN units out there but as they use a full bridge rectifier they can not be used with the systems as most sensors i work with use a 3 wire system, being power, common and output and if you connect these with any devices that use a full bridge some thing with go bang. now with the smaller cap i and down to 19 deg rise this is when it is open on the bench. when it is finished in an enclosure it will be higher, i have not tested that far yet. yes i know 25 deg rise is not a lot but in a 40 deg C day that would get up to 65 deg C and i always consider temperature to be the enemy. i have looked at all the comment so tomorrow i will checkout the specs on the other devices you guys mentioned. all idea,s repectfully examined

Last edited:

#### Irving

Joined Jan 30, 2016
3,187
but as they use a full bridge rectifier they can not be used with the systems as most sensors i work with use a 3 wire system, being power, common and output and if you connect these with any devices that use a full bridge some thing with go bang.
I've used DIN rail power supplies many times before and almost all sensors not on a current loop use a 3-wire setup of +/- power and output common to -ve rail. This is all common practice, with a common system ground tied to the -ve rail of all PSU (except those intended to provide + & - supplies or specialist - rails). Most, if not all, DIN AC PSU are isolated supplies though DC-DC converters may not be.

The fact it goes bang suggests a serious wiring error somewhere, or something is (mis-)using the common rail in a way probably not intended.

Rather than trying to work around this, might I suggest that some time working out why its going bang might reap dividends going forward, and possibly avoid a costly error further down the line.

#### Deleted member 115935

Joined Dec 31, 1969
0
thanks for the comments, the comment about the peak voltage made me think i may have use a cap that was to high and i was inadvertantly lifting it up higher, so i reduced it to 100uf, that helped a little bit. i am not sure what rating the transformer is apart from 24vac 30va, ther reason i am using this one is its the only small DIN mount version i can find, this is because all devices used in this system are to fit side by side using the same profile as a standard circuit breaker or multipes of it, for instance the power supply is a 3 mod so 3 times the width of a single phase circuit breaker. so unless i can find another i am stuck with it, yes there are 240vac to 24vdc DIN units out there but as they use a full bridge rectifier they can not be used with the systems as most sensors i work with use a 3 wire system, being power, common and output and if you connect these with any devices that use a full bridge some thing with go bang. now with the smaller cap i and down to 19 deg rise this is when it is open on the bench. when it is finished in an enclosure it will be higher, i have not tested that far yet. yes i know 25 deg rise is not a lot but in a 40 deg C day that would get up to 65 deg C and i always consider temperature to be the enemy. SO any one got any other ideas in how i can reduce it firther?
Why do you say / imply that a DCDC 250v AC to 24V will cause your system to go bang ?
Can you provide a circuit of how this PSU fits into your system and why you need to go for a 50 Hz transformer / LDO not a switcher

#### BobTPH

Joined Jun 5, 2013
6,070
as most sensors i work with use a 3 wire system, being power, common and output and if you connect these with any devices that use a full bridge some thing with go bang
Can you link to one of these sensors that goes bang when connected to a supply with s bridge rectifier?

Bob

#### bexy

Joined Apr 2, 2020
22
ok, its now off topic. i got the heat down to 20Deg C rise and i am waiting for some 317HV to arrive to see it i can get it lower. for those who went off track, yes i have been using 2, 3 and 4 wire sensors for over 30 years, although 4 wire sensors are rare in my industry now as most moves to 3 wire over 20 years ago this means that trying to connect a 4 wire device to a system that is full if 3 wire devices is silly. i have seen the damage connecting a 3 wire sensor to a 4 wire controller causes usually by blowing the bridge on the controller, also i have to be able to control 24vac relays and contactors, no not DC as thay are hard to get hold of and expensive. So AC has to be used for the switching while DC has to be used internally for the control devices and sensors to work. installing both AC and DC power supplies can cause miss match problems. SO not going to happen. what i am making is a basic control device that can get power from the common 24VAC transformer that is supplying all other items on the control circuit, yet it has to convert to DC so the opamps, mosfets and transistors can do there job. unfortunatly with these systems the transformers have to cope with delivering very low amps when not having to work to using most of there power when things happen so as we know the output voltage of the transformer is high when the load is low. the transformer that has to be used is basically the only DIN mount AC output transformer with a circuit breaker like profile available. i i found a better supply i would go with that instead.

#### Irving

Joined Jan 30, 2016
3,187
Ah, now it makes more sense. The easy solution is to use a bridge rectifier and smoothing capacitor and feed that to an isolated DC-DC converter to give you whatever output voltage needed. You can then safely connect the -ve side of the output to your common AC reference without any problem. Many of those units will convert 12 - 40v DC in to say 24V or lower at a range of wattages at 90%+ efficiency so no significant heat at all in many cases.

Last edited:

#### Irving

Joined Jan 30, 2016
3,187
Though I still don't understand why you didn't use an AC input DIN PSU with an isolated output like this one to start with?

I've used them many times with AC switching as is common in many industries and never blown anything up...