1
\$\begingroup\$

My question is this: What is the convention, if any, for specifying a power supply's input voltage value when performing a load regulation test?

Background

A typical specification for DC power supplies is its input voltage range--e.g., \$115\,VAC\,\pm10\,\%\$ for line-powered "AC/DC converter" power supplies,

115 VAC -> [DC POWER SUPPLY] -> Regulated Volts DC Output (CV mode)
115 VAC -> [DC POWER SUPPLY] -> Regulated Amps DC Output (CC mode)

or perhaps for a DC/DC converter power supply the range of unregulated DC input voltages \$V_{in}\$ is specified:

10 VDC <= Vin <= 14 VDC    // nominal is 12 VDC

Vin -> [DC POWER SUPPLY] -> Regulated Volts DC Output (CV mode)
Vin -> [DC POWER SUPPLY] -> Regulated Amps DC Output (CC mode)

Assume I have a DC power supply operating in constant voltage (CV) mode whose output voltage is specified as \$5.0\,VDC\,\pm\,0.1\,VDC\$, i.e.,

4.9 VDC <= Vout <= 5.1 VDC

and whose output power is specified as 10 Watts--i.e., when operating in CV mode the supply's maximum output current is 2 ADC (and its minimum output current is implied as 0 ADC).

Now I want to perform a load regulation test to verify the supply's output voltage remains within its specified voltage window

4.9 VDC <= Vout <= 5.1 VDC

over the supply's full range of output currents--e.g.,

0 Amps <= Iout <= 2 ADC

What is the convention, if any, for specifying the supply's input voltage value when performing a load regulation test? Does one ordinarily specify/use the nominal input voltage for this test--i.e., 115 VAC or 12 VDC for the examples above--or does one perform the load regulation test twice, once at Vin(min) and once again at Vin(max):

// Load regulation test 1a: Vin = Vin(min)
115-10% VAC -> [DC POWER SUPPLY] -> Regulated Vout @ Imin, Imax
10 VDC -> [DC POWER SUPPLY] -> Regulated Vout @ Imin, Imax

// Load regulation test 1b: Vin = Vin(max)
115+10% VAC -> [DC POWER SUPPLY] -> Regulated Vout @ Imin, Imax
14 VDC -> [DC POWER SUPPLY] -> Regulated Vout @ Imin, Imax

Or perhaps the test is run thrice @ Vin(min), Vin(nominal), Vin(max)? Or "other"?

(n.b. I understand that a complete answer depends upon "other factors" to include time and cost trade-offs when specifying testing regimes--i.e., what degree of testing is necessary and sufficient for a given product? Nevertheless, I'm interested in hearing some opinions / conventions for selecting the input voltage value when performing a load regulation test.)

\$\endgroup\$
  • 1
    \$\begingroup\$ Unless otherwise specified I would use the nominal rated input voltage when measuring load regulation. To be thorough you could do it at min, nominal, and max Vin. To be more thorough you could do each test at min and max rated temperature too. There's no convention, but many manufacturers have standard validation tests that they run through. \$\endgroup\$ – John D Mar 7 at 0:04
2
\$\begingroup\$

While there are no standards, the range of parameters you test will depend on the type of testing you are doing:

Manufacturing test; here you would certainyl want to span the characteristics that give you the best possiblity of finding early failure devices.

  1. Test the full spectrum of Vin at full load
  2. Test startup and settle times for maximum and minimum loads at Vin minimum
  3. Some soak test (probably lowest Vin and max load) to find early failures (this could be minutes to hours depnding on the supply). It might also include IR monitoring for hotspots.

Inbound test; here you are expecting the manufacturer to have already performed a broad scale test and you want just to detect failed units as a sanity check.

  1. Startup settle at full load and minimal Vin
  2. Multiple power cycle at min and max load
\$\endgroup\$
1
\$\begingroup\$

There is no standard convention. It depends on the practice of the manufacturer and the end user. For example, a reputable manufacturer would be expected to test his equipment to the full extent of its specification. Thus he would conduct load regulation tests over the full specified range of inputs. However, he might do this only for an initial production run to ensure that the product design is robust. Cost considerations may limit the amount of testing after that. Some customers may ask for a full set of tests to be performed on all units but they would be expected to pay extra for that. Some customers may accept the manufacturer's specifications at face value and not conduct any incoming inspection tests on load regulation. Others may want to verify that the supply is working before installing it in the end user equipment. Even then, they may not do a full test but only at the nominal input voltage to minimize cost and time. This often depends on the end user. A commercial customer may not be willing to pay extra for the time and effort to conduct more testing. A defense customer, on the other hand, may be willing to pay extra to ensure that the equipment will work properly under all conditions. As you mention in your question, many factors are involved in these decisions; there are no standard conventions.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.