Best test load for SMPS?

[mkiijam]

I'm repairing a few SMPS modules that are combined into one "supply." I would like to run them at their specs which is around 20 amps. Would be nice to step up the load in 1-3 amp steps. Any ideas? 

[Konkedout]

What is the voltage?  Do you want cheap or do you want elegant?  I still use a klunky e-load that I designed and built maybe 28 years ago.   That is good for maybe 3V up to 60V.  I also made two resistor load banks that I use for high voltages.  Just a connection of 15 resistors.  I use alligator clips to select 1, 2, 4, 8 in parallel in any combination to give me 15 steps.

[moffy]

Suitable power resistors.

[jonpaul]

Power rehostat
wirewound power resistors

j

[Vovk_Z]

Power rehostat
wirewound power resistors

These days an electronic load may take less space and even cost less than powerful rheostats (if you already don't have them). And it is more suitable.

[pienari]

I use resistor wire.
Easy to use and adjust .

Can easily handle 40-80A currents.

https://www.reichelt.com/de/en/100-g-resistor-wire-1-50-mm-rd-100-1-50-p57218.html

[electr_peter]

Most DC electronic loads can do 20A or more. DIY DC load is easy to build if all you want is CC mode, just have to use several transistors with adequate heatsink. There are plenty ideas on this forum for that.

[Mechatrommer]

I'm repairing a few SMPS modules that are combined into one "supply." I would like to run them at their specs which is around 20 amps. Would be nice to step up the load in 1-3 amp steps. Any ideas? 

what wattage we are dealing here? 20? 50? 100? 500W? for quick snap, passive version maybe the answer without burning 100 bjt/mosfets to verify your 500W e-load design/circuit. https://www.eevblog.com/forum/beginners/4r-and-8r-100w-resistors-from-ali/msg4484695/#msg4484695 just need water if it starts to smoke... ymmv

[MarkT]

Not necessarily relevant, but in big test rigs for PSUs they use smart loads that recycle the power back into the mains supply that's powering the units under test.  Much cheaper and cooler.

[MathWizard]

I use resistor wire.
Easy to use and adjust .

Can easily handle 40-80A currents.

https://www.reichelt.com/de/en/100-g-resistor-wire-1-50-mm-rd-100-1-50-p57218.html

So do you unroll it a lot ? Otherwise wouldn't there be a ton of inductance ? I've seen video's of people with a roll of wire in a bucket of cold water, but I think computer PSU's hate inductive loads. I'm too lazy to work the inductance of coil like that right now.

I've talked about trying stuff like 12V 200W heaters as computer PSU load, but never spent the money to try it yet.

I bought a bunch of mosfet's that I tried on some 1/4" 1-ft steel square plates, and it was not very good. The steel could not conduct the heat fast enough to keep up with the 1 BJT.

One day sort of by by accident I ran about 35 or 70A through my BM269 DMM, trying to load test a computer PSU. It made my 200W resistor's winding glow red too. Some how the fuse never blew tho.

[Doctorandus_P]

The white square "cement" resistors are among the cheapest power resistors you can get.

They also have a high mass, are robust and can withstand high overloads for a short time.

It's also relatively simple to clamp a bunch of them between a few aluminum L- profiles and add a cooling fan for more dissipation without overheating.

I've also seen a long end of copper wire being used which is dunked in a bucket of water for cooling.

[pienari]

The white square "cement" resistors are among the cheapest power resistors you can get.

They also have a high mass, are robust and can withstand high overloads for a short time.

It's also relatively simple to clamp a bunch of them between a few aluminum L- profiles and add a cooling fan for more dissipation without overheating.

I've also seen a long end of copper wire being used which is dunked in a bucket of water for cooling.

Well it depens how many wires paraller.
Air cooling is not best option when wires get red hot.
I tested all my server psu:s 30-200A and never had any issues.

My friend have tested water cooling and 80A boils water .

[Mechatrommer]

My friend have tested water cooling and 80A boils water .

if thats not good, ask him to use oil...

[ejeffrey]

So do you unroll it a lot ? Otherwise wouldn't there be a ton of inductance ? I've seen video's of people with a roll of wire in a bucket of cold water, but I think computer PSU's hate inductive loads. I'm too lazy to work the inductance of coil like that right now.

A series inductance raises the load impedance at high frequency, and it's still got the resistance so it won't be high Q.  Generally that tends to increase the stability of most voltage regulators although that ignores the parasitic capacitance and self resonance of the spool.  The reactance in any case should be dominated by the power supply filter capacitors so a coil of resistor wire shouldn't upset it at all.

It's different for a lab supply which needs to operate in both CC and CV modes although a lab supply generally should be designed to be stable under as many reasonable loads as possible.

[ArdWar]

if thats not good, ask him to use oil...

Unless you really need that high temperature, steam is usually safer than whatever fume produced by hot oil. The phase change also work as safety feature limiting your max temperature as long as you don't let it run dry.

You can even reliably calibrate out wire tempco by deliberately running it at (or near) boiling point

So do you unroll it a lot ? Otherwise wouldn't there be a ton of inductance ?

You can always wound it bifilar.

[KrudyZ]

Hooking up a bunch of resistors is hardly a test for an SMPS.
The thing you really want to test is the step response and frequency stability which will require a dynamic load.
So the minimum I would add to a home brew would be a power MOSFET that can handle the current driven from a pulse generator.
This would allow you to set a base load with permanently connected resistors and a step load through the FET.
Commercial units allow you to do that with less hassle and they usually also include slew rate control.

[mkiijam]

The three supplies in question will be +/-18Vdc and a 5Vdc.

The thing that intrigues me, and sorta-kinda-always has is: Resistors and light bulbs seem such low DCR. I really can't wrap my head around the DCR of a 100W lightbulb might be like 150 Ohms. If I measure the DCR of a string of op amps that consume a 100Ws, is it really going to be that low?

[Mechatrommer]

if thats not good, ask him to use oil...

Unless you really need that high temperature, steam is usually safer than whatever fume produced by hot oil. The phase change also work as safety feature limiting your max temperature as long as you don't let it run dry.

ask you family whoever familiar with cooking with oil, is it not safe or what? phase transition will create bubbling around resistive components and multiply corrosive rate several magnitude. it will corrode in a matter of minutes. but oil is a bit messy to work with. ymmv.

[ArdWar]

For some reason I was thinking of mineral oil and petroleum oil instead of the much more common vegetable oil.

Well, in my defense mineral oil (and freaking organophosphates) are much more common in my work.