I’ve got this 48V 10.4A generic chinese switching power supply to power my mini lathe’s controller and stepper motors.

The two steppers draw when working just 1.5A max each. And the controller 0.5A.

This supply turns on its loud fan a couple of minutes after it’s on and never turns the fan off, even if the steppers/controller are in standby mode.

The supply’s vendor told me the fan is not temperature aware, so it stays on forever.

I wired a small temperature control circuit to turn the fan cooler on and off depending on the temperatures inside the supply.

But I noticed that the more it stays on, and when powering actively the steppers, this toroidal coil (don’t know its proper name) heats a lot. It slowly got to almost 120º C. In the picture I attached a temp sensor to this coil, to turn on and off the cooler.

But with a laser thermometer, i checked the temps of most components and nothing ever reached 50ºC, but this coil gets too hot.

The supply works normally, outputs it’s 48V properly, the lathe works normally.

Is it normal for this component get so hot?? Is there anything I can do to fix it?

Thanks for reading![enter image description here]1

The output choke measures 40mm diameter, aprox. 23 mm width, the hole measures 12mm aprox. The wire is 1mm thick and appears to be three wires wired in parallel. It’s green core is not one whole unit, but two half width kept together by the wiring.

enter image description here

  • \$\begingroup\$ Do you have a multimeter with current measure capability. Check the actual current as the toroid starts heating up. Anything over 100C is not OK IMO. I have had cheap supplies NOT deliver the claimed rated current. If this one isn't meeting spec, spend a few extra bucks and get one from a quality supplier like Digikey. \$\endgroup\$ – Randy Nuss Apr 20 at 0:14
  • 1
    \$\begingroup\$ Apparently the design uses wrong undersized inductor. The winding wires look too thin for a 10A output. And who knows what kind of ringing is there, so the core can dissipate too much if it is of wrong kind. Can you determine with measurements, which one gets heated, the core, or the coil? \$\endgroup\$ – Ale..chenski Apr 20 at 0:28
  • \$\begingroup\$ A temperature rise of 70 degrees C is too much. Of all the actual electronic components on the board, the transformer is the likeliest to survive -- but I don't consider it acceptable. I believe that you are discovering the difference between "inexpensive" and "cheap". \$\endgroup\$ – TimWescott Apr 20 at 0:31
  • \$\begingroup\$ @RandyNuss I do. Do you mean desoldering one of the legs of the toroidal and measuring the current or measuring the current on the output of the supply? Those values I wrote (from the steppers moving and in standby mode) were measured with the multimeter on the output of the supply. In my country there’s no digikey and it’s difficult to find from other suplliers, but I’ll try if there’s no way to fix it. Thanks! \$\endgroup\$ – Rodrigo Apr 20 at 2:36
  • \$\begingroup\$ @Ale..chenski someone told me something also about the thickness of these wires. I can try to check the temperature (although it’s difficult with the laser meter) if the core or the wires are hotter. A friend told me to rewire the coil with the same amount of “turns” but with double the wires in parallel with each other, to divide the current per wire. Would this be a solution? \$\endgroup\$ – Rodrigo Apr 20 at 2:38

The output choke is cooking .The copper should be cool because you are not heavily loading the PSU .I have seen this before on a Chinese made Japanese designed PSU .The choke looks like it is made of powdered iron which for marketing reasons is now called something else .The powdered iron becomes more lossey with temp and time and then it gets hotter and runs away .You could replace the output choke with something more efficient .

  • \$\begingroup\$ Thanks! I’m looking for replacements for this output choke, and I’ve found some for 10A. But how do I know the needed inductance? A higher value the better? Or it must be an exact value? Is there a way for me to measure this in this one? \$\endgroup\$ – Rodrigo Apr 20 at 11:00
  • \$\begingroup\$ Not entirely marketing reasons; the different names do refer to different chemical compositions, and many of them don't even have all that much iron. Sendust, a particularly common one, is mostly nickel if I remember right. \$\endgroup\$ – Hearth Apr 20 at 21:30

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.