Powering a car wiper motor through a PC PSU?

[Chriss]

HI!
I'm thinking to use in one of my DIY project a wiper motor from an old car and power the motor through a PC 400W PSU.

As far as I know, the motor is using around 5A on the max rpm - without load.
I'm a bit afraid what about the inrush current of that motor?

My PC PSU what I have is rated:
5v - 36A
12v - 16A

The motor could maybe under load use 10A but what if the motor must start even if he is under some load?
The motor should transfer pellet in my pellet burner.

The motor inrush current could maybe go over 20A or so.
The PC PSU would maybe turn of if I'm correct? cos the overcurrent sensing circuit in the PSU.

So, what I was thinking is to start the motor every time through the 5v rail and after a few sec. switch the motor to 12v rail.
Maybe I could avoid some damage on the PSU or avoid the constantly turning on and off of the PSU.
The 5v rail can handle 36A...

Or is there any better idea how to mod the psu to meet the requirement?

I also checked some AC 220V TO DC 12V 40A PSU's but they are also just switching power supplies.
I'm not sure does they would do the job better then a modded pc psu...

Any help and idea wellcome.

Thank you.

[james_s]

Try it and see what happens, if the inrush causes a problem you could wire a NTC inrush limited in series with the motor.

[Chriss]

Can you advice pls what type of NTC could I put in series?

Thanks.

[akimpowerscr]

Why not buy an oldschool 12V used battery charger (transformer + bridge rectifier) ?

[Chriss]

That is also a possibility but now I have the mentioned psu and I have maybe 10 or 20 of them...

[james_s]

Just search for NTC inrush limiter and you should find lots of options. Have you tried just connecting the motor to the PSU? It might work, and if it doesn't it won't damage anything.

[akimpowerscr]

That is also a possibility but now I have the mentioned psu and I have maybe 10 or 20 of them...

Have you checked if you can draw 16A from the 12V output without load on the 5V output?

As far as I know, the regulation acts on the 5V rail and we cannot fully load the 12V output if there is no load on the 5V output

To be checked because I am not sure of it. 

[james_s]

Most modern PC power supplies have separately regulated outputs for 5V and 12V, modern motherboards and GPUs have onboard regulators and draw most of their power from 12V. It's worth checking but I don't expect this to be an issue.

[Chriss]

Tomorrow I will check and make some investigation and measurements.
I made some calculation about what type of NTC I should use. I come out with a result to use an NTC ~1 Ohm / 40 Joules if the max inrusch current is  ~18A and the input voltage 12V.

[Siwastaja]

You could bodge a relay to start the motor with 5V (say, for half a second) then switch quickly to 12V. Obviously make sure the relay can handle the current, the switching spark, and under no conditions short the 5V and 12V lines together. If the switchover takes, say, some 20ms or so, the motor inertia keeps the motor running during that time.

Although, putting any effort into such hacks, you wouldn't be too far off from just creating a proper PWM controller.

[akimpowerscr]

All these more or less messy adaptations (NTC, switching from 5 to 12V) hide an obvious reality: these computer power supplies were never designed for such (ab)use and they are not protected enough to power a dc motor.

[Circlotron]

Maybe put a very large capacitor across the +12V output to supply the momentary current surge.

100000uF 16V doesn't cost much.  https://au.element14.com/united-chemi-con/esmh160vsn104ma80t/aluminum-electrolytic-capacitor/dp/1682019

[Chriss]

A large cap would probably also pull out a huge amount of current from the psu whet it has to charge the cap...

But what if I would build a pwm controller which would every time when the moto is starting start from a voltage around 3v ap to 12v in a time laps of ~200mS?

[akimpowerscr]

Starting with the wrong option to save a few bucks doesn't seem like a smart decision to me.

A 10 or 15A battery charger without electronics, with transformer and diode bridge is a much better solution because:

- the transformer in these battery chargers are always a high short circuit impedance transformers, so will limit the motor starting current (= smoother starting torque not to damage the transmission)
- the transformer is very robust and can withstand heavy overloads for a few minutes
- the rectifier bridge can be changed and replaced by a bridge for higher current
- can remain constantly connected to the motor and only be supplied when the motor is needed.

Super simple, robust and reliable, why look for troubles ?

[Chriss]

Yes, you absolutely correct. I was just curious is there any option to make the proper mod on a pc psu to simply meet the requirement.
Because I have many psu's in my garage sitting and waiting to make something from them.
Because some of them are labeled with huge current rating like 20A on 12V etc but then it turns out they can not handle large current. Somebody would say yes they can but I have to use them in a proper way. That would also be a correct answer.
And that looks for me like a muscel man with huge muscels but without power.
But yes, the muscel man should also be used on a proper place, in show rooms and in front of mirrors.

Actually I didn't search for a psu how to power up my motor but I'm thinking about, how to reuse thinks I have already like a PC psu.

It is not a problem to buy a car battery charger or maybe to interchange the 12v wiper motor with a 220VAC motor and build the controller.

But thus makes more fun to reuse thinks laying around. :-)

[Siwastaja]

If you build a PWM controller, add current sensing to it: that way you'll create an automatically robust circuit which actually limits current and torque, keeps the motor running at high efficiency, without needing to think about any corner case assumptions such as "how long does it take to ramp the speed up" which fails whenever something special happens (like mechanical overload).

Current sensing can be as crude as a comparator detecting voltage drop over sense shunt resistor and inhibiting PWM for some time. For example, if you use a microcontroller like a simple Arduino board to provide PWM control, you can use a low-side current shunt, wire the output (preferably amplified) to the analog comparator pin of the microcontroller, to get a software interrupt each time current limit is exceeded; then you can quickly bring the PWM setpoint down in the ISR. This is a simple starting point and I have built robust motor controllers that way.

Yes, it will be more work, though, don't expect it to run after a day of work. But you'd learn how to design and build actual motor controllers, possibly a valuable skill.

[akimpowerscr]

But thus makes more fun to reuse thinks laying around. :-)

Your preoccupation with recovering instead of throwing away is very commendable, myself, I am participating in a "repair café" project which has the same goals.

https://en.wikipedia.org/wiki/Repair_caf%C3%A9#:~:text=A%20repair%20caf%C3%A9%20is%20a,by%20and%20for%20local%20residents.

But why not try to sell these power supplies if they are still working well?

NB: Dave also shares this same concern of not throwing anything away ..... here is the proof.



Good fun 

[m k]

But thus makes more fun to reuse thinks laying around. :-)

Since it's a pellet burner you probably have space.
Use more of them in parallel, use so many that who cares what the current is.
That could diminish the fun part considerable though.
Maybe not, you could use the power alternatively for a subwoofer based vaulting cracker.

Don't know, maybe just sell those PSUs and buy a simple charger.

[james_s]

All these more or less messy adaptations (NTC, switching from 5 to 12V) hide an obvious reality: these computer power supplies were never designed for such (ab)use and they are not protected enough to power a dc motor.

They're very well protected, hence the issue, they shut down instantly when even a momentary overload takes place, I have never managed to kill one though.

I prefer the hot swappable server power supplies. I use them to power my RC aircraft chargers, you can get hundreds of watts of clean DC power for peanuts. Usually they can be modified to run two or more in series for higher voltage, many larger chargers can accept 24V.