DC power supply doesn't work with DC motor

[tjkolev]

Hello circuit wizards!

I have this 120VAC to 12VDC (rated at 5A) power supply that came with an Igloo cooler. I am also using it to power a (cheapo) 12VDC air pump (DC motor with impeller) to inflate/deflate our air mattresses.

The problem:  Switching on the air pump causes the motor to make ~ 1/4 turn, and then stops.

My workaround: I have added bullet type connectors on the air pump cable. I switch on the air pump. And then I keep connecting and disconnecting the bullet terminals on one of the wires progressively at a faster rate. The motor starts spinning faster, and after 5 - 6 of this it "catches on", I keep the terminals connected, and everything works like a charm. It feels like I am playing the role of a starter winding in an AC motor...

The need: If possible and practical, find some way so the above power supply and air pump work normally, without my workaround. I am thinking of adding something to the circuitry. Looks like the power supply isn't happy with inductive load.

What I've done:
* The air pump works normally when powered from my desktop power supply. Using 12 - 14V, it draws 3.1A.
* When I have it running off the Igloo power supply it draws 4.2A, and the motor spins faster. When the air pump can't start it draws 0.2A.
* The Igloo power supply shows 14V without load, and a nice, flat, unremarkable line on the oscilloscope.
* The air pump is glued/sealed. Through a grate I can see the two wires connected directly to the terminals on the back of the motor. There's no capacitor, or any anything else.

I am adding some photos. From what I can see, the three legged element with a heat sink on the low voltage side is a SBL2040 Schotkky diode. No clue how that works...

Cheers!
tjk

[Xena E]

What voltage do you measure across the supply when the motor won't start?

It's starting current surge is probably putting the PSU into fold back limit.

[inse]

You‘re going to need a bigger boat - err power supply

[Ice-Tea]

Inrush causing foldback? Try a 5 NTC...

[CaptDon]

Those inflator motors draw a huge amount of current!! I measured one that runs on 4 D cells and was surprised at the current draw. The inrush at start could hit 10 amps. The motor is wound with heavy wire and the motor needs to spin up to a high RPM before the back EMF reduces the large current draw. No wonder they kill batteries so quickly. That motor has to spin like crazy to get any meaningful air flow from such a small impeller.

[tjkolev]

I could not figure out how to determine the "inrush at start" current. On the bench power supply I could not detect any surge. Of course, the thing could be slow to react. This is what the oscilloscope caught when starting the air pump using the bench power supply:

I limited the current to 0.6A on the bench power supply, and the air pump could still start up (albeit spinning slowly). Voltage went down to 2.3V

What voltage do you measure across the supply when the motor won't start?

0.224V

And here's the oscilloscope view (not at the same scale as above) when the air pump would not start:


I am thinking I'll just get a bigger boat...
tjk

[Circlotron]

It's starting current surge is probably putting the PSU into fold back limit.

Yep. I'd try putting a very large capacitor across the output of the power supply, maybe something like 50,000 - 100,000 uF that can provide the starting current surge for the motor. Of course, that size capacitor might actually prevent the power supply from starting up because of it's own charging current  Maybe have the capacitor charge up through a resistor, then short out the resistor, then connect the air pump.

[perieanuo]

Inrush causing foldback? Try a 5 NTC...

have you tried this or putting a monostable relay for inserting ie a 12V light bulb (or even a 230V one) in series with the motor for the first 2-3 seconds?
it's obvious your psu is entering protected mode (folding someone said), the PC supply is just not so start-current-limited like the other one.
starting with low current until the motor gets spinning will do it for sure. imho, the ntc may not do it for sure, you gotta test one of those

[BrokenYugo]

I solved a similar problem by placing a power resistor across the power switch, making off the "start" position, then flip to on once up to speed to give the motor full power, unplug to shutdown. I think 1 ohm did it in my case, starting such a pump from a 5V supply. A NTC would probably be safer, as would automating the startup with a delayed relay so operator error can't hold it in the start state.

[Xena E]

It's starting current surge is probably putting the PSU into fold back limit.

Yep. I'd try putting a very large capacitor across the output of the power supply, maybe something like 50,000 - 100,000 uF that can provide the starting current surge for the motor. Of course, that size capacitor might actually prevent the power supply from starting up because of it's own charging current  Maybe have the capacitor charge up through a resistor, then short out the resistor, then connect the air pump.

Here:-

[tjkolev]

@Zena E,

Thank you for the schematics!  I didn't have the right components in my collection. I used a 180 ohm resistor, 2x 1600 uF 16V capacitors. My hi current diodes wouldn't go into the holes of the breadboard, so I used skinnier ones. But hey - it worked like a charm! The diodes got really hot, and one of the jumper wires started smoking...  I'll use proper components, and package them somehow on the PSU cable.

Thank you all for your time and suggestions!

Cheers!
tjk

[ledtester]

I wouldn't use a breadboard for a high-current application like this one. Besides, as you have noticed, the leads of the high-current diodes don't fit.

The diode leads are likely going to be too large even for prototyping boards that have copper pads.

Since this circuit only has a few components a quick and easy way to test it out is to solder the leads of the components together -- perhaps with help of thick solid wire -- creating a "free air" circuit without a supporting board. Once you get it working you can turn your attention to making a more permanent version of it.

[Xena E]

@Zena E,

Thank you for the schematics!  I didn't have the right components in my collection. I used a 180 ohm resistor, 2x 1600 uF 16V capacitors. My hi current diodes wouldn't go into the holes of the breadboard, so I used skinnier ones. But hey - it worked like a charm! The diodes got really hot, and one of the jumper wires started smoking...  I'll use proper components, and package them somehow on the PSU cable.

Thank you all for your time and suggestions!

Cheers!
tjk

Glad to help.

Like ledtester says about the construction its simple enough to sky wire.

Unfortunately you will need diodes that will handle the full current of the application. Perhaps one of the twin packaged T0 220 types bolted to a metal heatsink. Little 1 or 3 amp ones will burn out eventually.

The component values for the rest are basically what works for you, the resistance was suggested as a minimum, and the capacitor needs to be big enough to give the starting kick.

Regards,
Xena.

[rfengg]

Hi @Xena E,

Great idea.....do we need the series diode in line with the PS also though? If the series diode is replaced with a short, would it not work?

[Xena E]

Hi @Xena E,

Great idea.....do we need the series diode in line with the PS also though? If the series diode is replaced with a short, would it not work?

Good question!

It possibly would, I added it as I can't be sure how the OP power supply would react, without the diode it may still try to lead supplying the current, as the PS voltage will be ≈ 1V more than the capacitor and still go into fold back limiting, leaving it still unable to start the pump motor.

This way may not be strictly necessary but ensures results, perhaps the OP will report back if he tries it.