# Power Supply Output Capacity

Discussion in 'Electronic Projects Design/Ideas/Reviews' started by ADWSystems, Sep 18, 2017.

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
I have always understood a power supply will maintain its output voltage up until it reaches the output current capacity. That is to say a power supply with a 12 volt output and a 20 amp capacity will be able to maintain an output voltage up until the current draw reaches 20 amps. After 20 amps??? but that's fine it is now being used beyond spec.

Is this a correct understanding or have I missed something all these years?

2. ### JLNYActive Member

Joined:
Mar 5, 2011
Messages:
220
Likes:
35
Location:
New York, USA
Yes, you are correct. Most power supplies are constant voltage, so they will output at the same voltage regardless of current draw up to some maximum rated current. Above that, it might either lower the voltage to limit the current, latch and shut off completely, blow a fuse, or just keep going until it fails, depending on the type and construction of the supply.

You say you are using the power supply beyond its spec? by how much? Maybe there will be some amount of headroom in the design above 20A, but at that point all bets are off and you are basically relying on the good will of the designer. IIRC, most fuses have about 20% headroom above their rated value to prevent false blowing. Either way, it's possible that some component could start to overheat and fail. Probably not recommended.

Last edited: Sep 18, 2017

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
I am just putting out the foundation of my expectations to have them corroborated before delving deeper. I should be operating well within the power supply current capacity. The power supply in question is 12VDC 20A. With no load the voltage read 12.25V. With a 2 ohm load, the voltage out is 9.9V and the reading is 4.33A. With a 1 ohm load, the voltage out is 8.8V and the amperage is showing as 7.5A.

To check myself I also have a SOLA (ie., brand name, more trustworthy) 24VDC 1.3A power supply on the shelf. At power on (no load), the voltage is reading 24.1V. When connected to the 21 ohm load, the ammeter was reading 1.2A while the voltage dropped to 23.7V.

In light of the reading for the 12V power supply, how do you rate a power supply? Using a 12V nominal reading, at what voltage drop (due to current draw) is said to be the limit? The SOLA 24V power supply, it dropped 0.4V (1.6%) with a 1.2A (of 1.3A rating). But the 12V 20A supply, was reading 8.8V at 7.5A. Obviously I would not call this a 20A supply. If I continued to decrease the load (increase resistors), then I would need to find the load that produces a 0.2V drop in output voltage. Am I on the right track?

Last edited: Sep 18, 2017

Joined:
Jan 12, 1997
Messages:
-
Likes:
0

5. ### Nigel GoodwinSuper ModeratorMost Helpful Member

Joined:
Nov 17, 2003
Messages:
39,205
Likes:
640
Location:
Derbyshire, UK

A regulated supply should maintain the voltage within a reasonable degree as load increases, yours is obviously faulty as it drops so drastically under load.

• Like x 1
6. ### JLNYActive Member

Joined:
Mar 5, 2011
Messages:
220
Likes:
35
Location:
New York, USA
Hmm. Well, the 24V supply sounds about correct. the internal resistance of the supply/leads/ammeter etc. will probably result in some degree of voltage drop (0.4V/1.2A~=0.3Ohm series resistance). The 12V supply, however, doesn't sound right. The voltage drops you are seeing would be equivalent to about 0.5 Ohms series resistance, which might be okay in a 1.3A supply, but is way too high for a 20A supply. Are you using any kinds of leads with it which might be contributing to a high series resistance? I would be using pretty thick and well secured cables in line with a 20A supply.

Last edited: Sep 18, 2017

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
I am using the same cables for both tests, changing as little as possible between them.

The tests of the 20 amp supply are only being performed at 12A and 6A (1 and 2 ohm loads).

Last edited: Sep 18, 2017
8. ### KeepItSimpleStupidWell-Known MemberMost Helpful Member

Joined:
Oct 30, 2010
Messages:
9,911
Likes:
1,094
ONLINE
I really think your missing something. Two basic power supply topologies are CV, CV/CC and CV/CL where C?=constant; L=Limit and; ?C=Current.
In a Limiting situation, it's not regulated and CV could just be protected by a fuse. The Power supplies will have an SOA or Safe Operating Area.

There is a mechanism known as foldback current limiting: tps://en.wikipedia.org/wiki/Foldback_(power_supply_design)

You may find SOA or Safe Operating Area curves for power supplies. They may be in various forms. I-V and I-T are possibilites. http://www.tij.co.jp/jp/lit/an/slva766/slva766.pdf The SOA can be clipped at the corners too.

There are "power supplies" names "Voltage sources" and "current sources" or SMU's that operate in 4 quadrants. e.g. I+V._, I-V-, I-V+ and V+I-.

The HP DC power supply handbook http://literature.cdn.keysight.com/litweb/pdf/5952-4020.pdf is a really old reference, bit a very goos one.

I've read about a new design and just can;t place who with some very strange SOA curves.

Nothing is ever simple.

There are times that 3 resistors in series are not the same as a single resistor or times that mounting a SMT component edgewise is a different component because of parasitics.
Wires that wiggle in the earth's magnetic field generate a current. Moving wires generate a current. Most people don;t have to worry about those effects- I did.
There are circuits that can be altered if you touch the components.

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
The TI SOA document seems more premised on temperature effects and (to a lesser extent) DC-DC converters. An interesting read, though I'm not seeing the direct applicability here. The HP read is a good reference as it explains most of the acronyms you used prior. As I have three points of reference (no load, 1 ohm, 2 ohm) and both the voltage and current change. Any two points will form a straight line, so unfortunately my two data points will as well. But do they make sense? 2 ohm, 9.9V, 4.33A and 1 ohm, 8.8V 7.5A. 2 ohms should be 6 amps and 1 ohm 12 amps. 4.33A over 2 ohms is 8.66V, the voltage is reading 9.9V.

10. ### KeepItSimpleStupidWell-Known MemberMost Helpful Member

Joined:
Oct 30, 2010
Messages:
9,911
Likes:
1,094
ONLINE
i was answering the general question while you were talking about specifics.

Since models and datasheets were not mentioned, the only thing I can offer is:
1) Are the sense leads connected?

If i do this: =(9.9-1.2)/4.33 , I get about 2 ohms which suggests the sense leads are disconnected.

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
There isn't external sense on the power supply that I'm aware of. I have seen them on many Power-One supplies, but not on the Accopian supplies I have used, and there aren't on the Sola supply. This supply doesn't seem to have them external either.

What is the basis for you formula (9.9-1.2)/4.33? I know it is ohm's law, but what of the -1.2? How does that indicate sense is not connected?

12. ### KeepItSimpleStupidWell-Known MemberMost Helpful Member

Joined:
Oct 30, 2010
Messages:
9,911
Likes:
1,094
ONLINE
Usually, a diode is used to clamp the sense leads if they are open. I did subtract, but you may have to add. So,1.2 is basically two diode drops, one for + and 1 for minus.

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
This must be internal to the power supply. The only terminals on the outside are L,N, G for the AC and 3x V+ and 3x V-. Nothing that indicates "sense", as I have seen on the Power-One units. They have terminals labelled V+, S+, V-, and S-. Obvious enough for me.

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
I'm working to find a specification or manual for the power supply in question. In doing so I found this supply
http://www.jameco.com/z/HLG-240H-20...utput-Switching-LED-Power-Supply_2102166.html

and its associated data sheet
http://www.jameco.com/Jameco/Products/ProdDS/2115004.pdf

The 5th page is interesting as it discusses the CV+CC regions.

Which is opposed to this power supply
http://www.jameco.com/z/SP-240-12-M...e-Output-12-Volt-20-Amp-240-Watt_2101497.html

and data sheet
http://www.jameco.com/Jameco/Products/ProdDS/2101497.pdf

Which does not speak of the regions nor does it mention sense terminals (which would make me assume they are internally connected, hopefully).

15. ### JonSeaWell-Known Member

Joined:
Oct 1, 2012
Messages:
1,118
Likes:
90
Location:
Seattle, WA
Nigel summed this up pretty clearly. A regulated supply will put out the nominal voltage across the rated range. If it does not, it's either broken or a poor design.

In this article, AC-DC Power Supplies - Using Wall Warts, I examined the behavior of switching power supplies over the rated operating range. Although these supplies have lower power outputs than the supply you are talking about, it's reasonable to extrapolate these consistent results to the higher range. The pertinent information from the article that gives some numbers to Nigel's statement above is

"In contrast to the linear supplies, the switching power supplies tested maintain their voltage across the load range. In fact, from 0 to 100%, the supplies tested are within 3% of the target voltage."

This is shown by this plot.

Now you are clouding the situation even more by discussing LED power supplies that may be operated as a constant current supply. About the only time you want a CC supply is for powering LEDs. Most power supplies won't mention this in the data sheet because they are not designed to do this.

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
CC and CV power supplies were previously mentioned. All I am saying is that I found one and am noting the difference between the datasheets (where one lists the CV/CC regions and the other does not. No clouds.

17. ### KeepItSimpleStupidWell-Known MemberMost Helpful Member

Joined:
Oct 30, 2010
Messages:
9,911
Likes:
1,094
ONLINE
Two terminals for + and - are used respectively for the latter. It does not mean it has sense. I would expect the block diagram to show that. That is more like CV, CL where CL is not adjustable.

The first one is "application specific", driving LEDS.

Joined:
Mar 14, 2008
Messages:
711
Likes:
5
Location:
USA
I found many power supplies like the first, but only this one had the data sheet.

There must be a "sense" connection somewhere in order to regulate the voltage output (assuming CV). Sometimes it is user accessible sometimes (most times) not. On the Power-One supplies I have used, there are S+ and S- terminals you can connect to a load at a distance and have the power supply compensate for any voltage loss on the way there. Only I never had the occasion to go any distance to the load, so V+/S+ and V-/S- were shorted at the power supply terminals. If there are multiple + or - terminals, without indication to the contrary, I don't expect them to sense terminals. In the case of the linked units, the sense is not a user connectable point.

I am wondering if the power supply I have is not broken, but rather a unit with some sort of CV+CC feature (opposed to CV). Or it could be DOA.