Author |
Message |
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Thu Mar 20, 2014 6:47 am Post subject:
Linear PSU question Subject description: (current rating) |
|
|
from time to time and here and there i read comments like " linear power supplies should be run at 60% of their actual current rating".
i think that 60% is near the best effectivity for linear PSUs, thus creating less waste (temperature).
besides this, is there any reason i should not use up more than 60% of the rated current when using a LM317 based linear PSU? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
diablojoy
Joined: Sep 07, 2008 Posts: 809 Location: melbourne australia
Audio files: 11
|
Posted: Fri Mar 21, 2014 12:30 am Post subject:
|
|
|
Hi matthias
60% is a really really rough guide to go by and you probably shouldn't go by it , there are too many variables in power supplies to actually say it's a hard and fast rule. it comes down to the weakest link which may not neccessarily be the regulators themselves, perhaps read osal's forum topic on the subject. it's very good
http://electro-music.com/forum/topic-51694.html
edit : oh- if what you were powering with the supply had a constant power requirement that was always absolutely stable then no you could probably go quite a bit higher however nothing we build is like that at all , perhaps it will help if you try to think of it as allowing for some headroom.
I will probably get flamed for that analogy but hey i couldn't think of a better way to express it at the moment _________________ In an infinite universe one might very well
ask where the hell am I
oh yeah thats right the land of OZ
as good an answer as any |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Fri Mar 21, 2014 1:52 am Post subject:
|
|
|
thanks for the answer, diablojoy.
however, what i dont' get is: the datasheet for the i.e. power one linear supplies tells me it would regulate to 0.05%. why shouldn't i go higher than 60% then (given that i could live with the higher power dissipation)? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Fri Mar 21, 2014 2:35 am Post subject:
|
|
|
isn"t this all about heat/efficiency.
the power one linear PSUs (12 & 15V outputs) are rated with 55% efficiency.
the HBB provides 1.7A at 12V, which is 20W. so when providing 1.7A the loss would be ~10W?
now the ambient temp might be as high as 40degC, and i don't want the PSU/regulator to get warmer than 70degC (max operating temp according datasheet), then i would have to use a heat sink of 3K/W? why not? the regulators are mounted to the chassis anyways, and it should be easy to mound an additional power heatsink, no? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Fri Mar 21, 2014 3:05 am Post subject:
|
|
|
halt.
the power one linear PSU datasheet says:
Operating Temperature: At 100% load max 50 °C
so at 40degC ambient temp i would need a 1K/W heat sink _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
brock
Joined: May 26, 2011 Posts: 112 Location: Canada
|
Posted: Fri Mar 21, 2014 11:42 pm Post subject:
|
|
|
I think you are interpreting the datasheet incorrectly for the Power One supply. The efficiency stated is the ratio of DC power out to AC power in so most of the power loss is probably due to the transformer and rectifiers, not power loss across the regulator.
The operating temperature spec means that you can operate it at 100% power at up to 50C. After that you have to start derating by drawing less current to prevent the thing from overheating.
If you are looking to design your own supply a great resource is the old National Semi Voltage Regulator Handbook.[url] https://archive.org/details/NationalSemiconductorVoltageRegulatorHandbook1980[/url] Pay attention to heat sinking. The cooler the circuit runs the better, and remember that natural convection cooling only works for relatively low temperature increases, maybe 10C.
The real trick with power supplies that generate any heat is to ensure there is airflow. Still air is a real problem for power supplies.
The only reason to run power supplies at less than the rated output is to extend life, and this is probably the basis for the 60% rule of thumb. Running at full power means they run hot. Heat reduces the operating life of all components and electrolytic caps usually are the most susceptible. As long as you understand this, run it full power and just check the ripple voltage on the filter caps under load once in a while and replace them when the ripple gets to an unacceptable level. You'll likely get years of life from most supplies before you need to replace anything. |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Tue Mar 25, 2014 4:02 am Post subject:
|
|
|
thank you very much for your replay, brock.
nevertheless, i still have some questions:
brock wrote: | I think you are interpreting the datasheet incorrectly for the Power One supply. The efficiency stated is the ratio of DC power out to AC power in so most of the power loss is probably due to the transformer and rectifiers, not power loss across the regulator. |
it is still loss, or waste, and this will be dispensed in form of heat, no?
Quote: | The operating temperature spec means that you can operate it at 100% power at up to 50C. After that you have to start derating by drawing less current to prevent the thing from overheating. |
wouldn't this mean that the 20W PSU would have to get away with about 10W dissipated as heat still? so the only way to keep it at 50C would be to use some kind of cooling.
or am i missing something fundamental here. i really don't know ad want to learn.
Quote: | The real trick with power supplies that generate any heat is to ensure there is airflow. Still air is a real problem for power supplies.
The only reason to run power supplies at less than the rated output is to extend life, and this is probably the basis for the 60% rule of thumb. Running at full power means they run hot. Heat reduces the operating life of all components and electrolytic caps usually are the most susceptible. As long as you understand this, run it full power and just check the ripple voltage on the filter caps under load once in a while and replace them when the ripple gets to an unacceptable level. You'll likely get years of life from most supplies before you need to replace anything. |
_________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Tue Mar 25, 2014 10:25 am Post subject:
|
|
|
fonik wrote: | wouldn't this mean that the 20W PSU would have to get away with about 10W dissipated as heat still? so the only way to keep it at 50C would be to use some kind of cooling. |
Yes it needs some sort of cooling, but the spec means that this can just be passive cooling ... as long as you do not put the PSU in a closed box.
When you do put it in a box, and have some ventilation holes there will be some airflow still to cool the device ... but less so than in free air ... so you may want to monitor the actual air temperature in the box indeed indeed.
When your room temperature is 20C, and the temperature in the box would be 40C that would mean that you'd still have 10C spare ... and so it could work up to a room temperature of 30C then.
Yo ushould measure the air temperature in the box tho, and not the temperature of the heatsinks .. those will be warmer. _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Tue Mar 25, 2014 3:39 pm Post subject:
|
|
|
Blue Hell wrote: | When your room temperature is 20C, and the temperature in the box would be 40C that would mean that you'd still have 10C spare ... and so it could work up to a room temperature of 30C then.
Yo ushould measure the air temperature in the box tho, and not the temperature of the heatsinks .. those will be warmer. |
i thought i could calculate it, but no. i have to try and test it in the box i use. thanks for the clarification, jan.
now, in most of todays modular cabinets is no ventilation at all. i see a lof of wooden(!) cabinets, powered by huge linear PSUs. obviously it works as long as i don't draw too much of current. metal boxes would be of advantage, i guess. they will transport some of the heat to the outside, no? _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Tue Mar 25, 2014 4:06 pm Post subject:
|
|
|
Drilling some holes below and above the heat source to have airflow would be better than a metal box even ... it is true that a metal box transfer heat out easier, but the temperature is quite low so radiation is not too efficient .. air flow is better, and passive is probably fine.
Calculation of the whole system would be almost impossible yeah ... for some simple systems its doable, like when you know thermal resistance and the heat flow is through solids it is doable, but when air is involved it is hard to come up with a good model .. and without a model .. no calculations.
You might find some heuristics tho .. when googling. _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
Osal
Joined: Aug 16, 2011 Posts: 147 Location: Here
|
Posted: Fri Mar 28, 2014 1:40 pm Post subject:
|
|
|
Hello Fonik.
Regarding the power dissipation, there is a great difference between a LM317 based linear power supply and a Power-One one.
The LM317 regulator has a junction max operating temperature of 125C and a junction to case thermal resistance up to 5CW.
In the other hand, the power transistors used by Power-One like the ST 2N3055 have a junction max operating temperature of 200C and a junction to case thermal resistance of 1.5 CW
This is a huge difference and heat sink requirements are different. _________________ electronic-sea.net |
|
Back to top
|
|
|
fonik
Joined: Jun 07, 2006 Posts: 3950 Location: Germany
Audio files: 23
|
Posted: Fri Mar 28, 2014 1:52 pm Post subject:
|
|
|
Osal wrote: | Hello Fonik.
Regarding the power dissipation, there is a great difference between a LM317 based linear power supply and a Power-One one.
The LM317 regulator has a junction max operating temperature of 125C and a junction to case thermal resistance up to 5CW.
In the other hand, the power transistors used by Power-One like the ST 2N3055 have a junction max operating temperature of 200C and a junction to case thermal resistance of 1.5 CW
This is a huge difference and heat sink requirements are different. |
oh, huge difference, indeed. thanks for the additional info. _________________
cheers,
matthias
____________
Big Boss at fonitronik
Tech Buddy at Random*Source |
|
Back to top
|
|
|
|