Linear power supply & pumps

[jackjr]

Hi there,

Some time ago I hooked a small PC water-cooling pump to my lab power supply, a GW Instek GPD-3303.
The pump was one of these popular Eheim ones, 12V.

The power supply kept triggering the current limiter and after a few seconds stopped working altogether.
One of the internal fuse had blown up and I had to tear-down the entire PSU to replace it.

Any idea why a small pump did that? Connected to any other basic 12V switching PSU the pump works just fine.

[jwet]

I have the exact same power supply that is front and center on my bench- I love it.  Mine has an "S" suffix- I don't know what that means but mine has been flawless.

Motors very have high inrush current and significant inductance.  The inrush is kind of luck of draw thing depending on where the rotor is on startup.  This is a brushless motor and can likely draw some pretty healthy peak currents at strartup.   Running motors on a precision supply can be difficult but I'm surprised that the Instek couldn't handle it.  My best guess is when the supply kept going in and out of current limit and shutting off the output- the motor inductance was getting charged and discharged and the smart motor stuff was gettging confused.  This created big voltage spikes across the motor inductance on each interruption and it worked its way back to supply.  It might have been enough to break down the motor driver FET's.  This is a torture test for a power supply.  I imagine with all this up and down, the load went shorted and drew enough overcurrent fast enough to blow the fuse.  Fuses have a time dependent reaction time but at 10x or so at the current, they can blow very fast.  I guess the lesson is that if you're driving something "smart" and inductive, don't let the current limit hammer on it. 

[Circlotron]

Try putting a big capacitor across the power supply output. Say 4700uF.

[penfold]

[...]
Some time ago I hooked a small PC water-cooling pump to my lab power supply, a GW Instek GPD-3303.
[...]
One of the internal fuse had blown up and I had to tear-down the entire PSU to replace it.
Any idea why a small pump did that? Connected to any other basic 12V switching PSU the pump works just fine.

That is interesting, I have a similar model of PSU and have had similar kinds of experience. I've never really sat down to study its (the PSU's) characteristics in any detail, but I did find that it was quite sensitive to negative impedance loads (anything with a constant power feedback loop, if the voltage increases the current decreases, therefore, has negative incremental impedance), where it goes a bit marginally stable and very sensitive to step-changes in load. I also found it was a bit twitchy when it does a transformer tap change around 15V, often oscillates with anything other than a purely resistive load.

I've needed to change fuses a few times now. I kinda chalked it up to bad design, maybe a quality control issue.