Getting additional voltage/power to pc water pump

[gijoe50000]

Hi, this is my first post here..
So basically the water pump on my pc is rated to run on 7v-12v but it's really noisy. It will run at 5v but the problem is that sometimes it won't START at 5v. It's a powerful pump @23w which is too much for the motherboard itself to handle so the manufacturers recommend using a molex from the psu instead.

What I WANT to do is to run it from the 5v molex but give it an additional input from the motherboard fan header to ramp it up when necessary or to give it a boost at startup. Through the fan header I can control the extra power using software so if the temperature sensors in the cpu detect high temperatures it will kick in the extra power.

The fan header on the motherboard is 12v and from what I gather they can handle a max of about 10w but I'd be happier with a max of about 6w - 7w . My thoughts are to put a diode and a resistor from the fan header to the pump in addition to the 5v molex from the power supply.

I've done a few electronics projects in the past but I'm not an expert by any means.

Will this work?
thoughts?
Recommendations?

 

[GPG]

If the pump will run on 5V with a Schottky power diode in series this would isolate the 5V from a higher voltage.There are other solutions,

 

[gijoe50000]

I think I have a few of those lying around alright.
I'm most worried about backflow to the motherboard header (12v) from the molex (5v). As the motherboard header can only handle ~10w.

 

[GPG]

Got a link to the pump? Also you could use the 12V from the psu connector and use the MB supply to switch it. (no load on the MB connection) Other possibilities include a switch mode psu and use the signal from the MB to ramp the voltage up or down.

 

[gijoe50000]

This is the pump it's supposed to run at 23w @12v from what I've gathered online.
http://www.alphacool.com/product_in...language=en&XTCsid=2crjiu8phas3jumja7uekdqa40

Yea, I'm considering all those ideas, but I'm fairly crap at designing circuits.
Just doing a bit of research there at the moment..

 

[BobK]

If you want to run the pump at lower power use 12V and PWM.

Bob

 

[gijoe50000]

If you want to run the pump at lower power use 12V and PWM.

Bob

Do you mean like an rc circuit?

 

[AnalogKid]

GPG is correct, the most simple way will require two diodes. The +5 V and +12 V diode-OR into the fan. This way the +12 V does not try to pull the +5 V up. This method works only if the +12 V source can be programmed down to 0 V (or at most 4 V) when in the unused state.

PWM is more complex than a simple R-C circuit. Rather than supply the fan with a steady 12 V, you chop the 12 V into a 12 V square wave. As the ratio of the on and off half-cycles varies, so does the average power to the fan.

ak

 

[gijoe50000]

GPG is correct, the most simple way will require two diodes. The +5 V and +12 V diode-OR into the fan. This way the +12 V does not try to pull the +5 V up. This method works only if the +12 V source can be programmed down to 0 V (or at most 4 V) when in the unused state.

PWM is more complex than a simple R-C circuit. Rather than supply the fan with a steady 12 V, you chop the 12 V into a 12 V square wave. As the ratio of the on and off half-cycles varies, so does the average power to the fan.

ak

Ah yes, that all makes sense alright. I know Speedfan tries to ramp up the power to the fan headers on startup to calibrate them but it doesn't ALWAYS do it for some reason, I should be able to find a way around it alright. Will a normal diode be able to handle the 23w ~ 2amps?

 

[AnalogKid]

It will if it is a 3 A diode.