Power a water pump through USB 3.0

[Peentje]

Hey guys, I'm looking into a next project and stumbled upon something I'm not sure of.
Theoretically, would it be possible to power a low power water pump through USB?
The USB 3.0 standard introduces a higher current limit of 900mA compared to 500mA for USB 2.0, this was done because more often than before external apparatus need more juice.
This is 900mA off single bus-powered ports at 5V, which gives me 4.5 Watts.
Using a Y-cable and combining to USB ports could double that.
To me this looks more realistic than before to power a smaller water pump through USB, compared to the previous standard.
I'd have to step up the voltage from 5 to 12V though, losing some efficiency.

It has to be through USB, so please don't come up with alternatives ;)
Any thoughts?

[Shoggy]

Keep in mind that a pump draws much more power when spinning up... I don't think that this will work on a USB port.

[Peentje]

Keep in mind that a pump draws much more power when spinning up... I don't think that this will work on a USB port.

Yeah, since it's a motor the starting current might be an issue I agree. But how much more current and if it will fail...
What about high-end USB 3.0 external hard drives (not ssd) powered by the usb port. They need to spin up also, taking more current at first.
So the question is rather if it's gonna exceed the max specs I think.

[Raja@ASUS]

I'd have to step up the voltage from 5 to 12V though, losing some efficiency.

It has to be through USB, so please don't come up with alternatives ;)
Any thoughts?


In addition to what has been said above, this is one of those gross concept errors that makes absolutely no sense. You are overcomplicating something to no real advantage or gain. You have numerable 12V sources within a PC, so why use 5V USB and step up?.

[Peentje]

In addition to what has been said above. This is one of those gross concept errors that makes absolutely no sense. You are overcomplicating something to no real advantage or gain. You have numerable 12V sources within a PC, so why use 5V USB and step up?.

Because it's a laptop, and it would be convenient that way. So I'm not overcomplicating, I'm running with the options I have.

[Raja@ASUS]

Because it's a laptop, and it would be convenient that way. So I'm not overcomplicating, I'm running with the options I have.

Watercooling a laptop is overcomplication in the first instance.

[Conumdrum]

Watercooling a laptop is overcomplication in the first instance.

Yep. +1 to that.

[Olly_K]

make that +2

[Peentje]

Hmm, I missed the part where this was a poll ;)
This is a simple inquiry wether or not it's possible...

[ARandomOWL]

Using the power from two ports would give you 9W. A 90% efficient boost converter would give you 8.1W to run the pump. In that case, as long as the pump will draw no more than 625mA @ 12V during startup, it would work.

Another thing. With USB 2.0, the default current limit is 100mA. A device must be registered on the bus and request more current to be able to draw upto 500mA. I don't know if this is the case for USB 3.0, you would have to check.

[Conumdrum]

Okay. NO. It's 5V, not 12V. Circuits don't like less voltage than they were designed for. Simple.

WC a laptop? Sure, have fun, report back.

[ARandomOWL]

Okay. NO. It's 5V, not 12V. Circuits don't like less voltage than they were designed for. Simple.

WC a laptop? Sure, have fun, report back.

What's wrong? Never hear of a boost converter (http://en.wikipedia.org/wiki/Boost_converter) before?

[Conumdrum]

Sure, add complexity/cost. And still lacking on current. You increase voltage the amps must increase on the supply side, it's how xformers work. Plus the loss of the transformer needed to up the voltage in the booster. The USB circuits can't handle the amps .Math says no workies for any pump I know of.



If the laptop is going watercooled, it becomes stationary anyway. Get a proper 12 VDC PSU.

[ARandomOWL]

Sure, add complexity/cost. And still lacking on current. You increase voltage the amps must increase on the supply side, it's how xformers work. Plus the loss of the transformer needed to up the voltage in the booster. The USB circuits can't handle the amps .Math says no workies for any pump I know of.



If the laptop is going watercooled, it becomes stationary anyway. Get a proper 12 VDC PSU.

I'm not sure where transformers come in here. Why must the current increase. Anyway, as I said before, if you have two 900mA 5V supplies available and a 90% efficient boost converter then you can run a 12V pump as long as it draws no more than 625mA. There are pumps that will operate within this range.

[Peentje]

I'm not sure where transformers come in here. Why must the current increase. Anyway, as I said before, if you have two 900mA 5V supplies available and a 90% efficient boost converter then you can run a 12V pump as long as it draws no more than 625mA. There are pumps that will operate within this range.

This was an example of a pump I was looking at:
Phobya DC12-220 12V Water Cooling Pump - (DC12-220)
Nominal voltage: DC 12V +/-10%
Starting voltage: 8V and above
Power consumption: 6.5W +/- 10%
Current draw: 0.62A +/-10%
Max. flow: 400 l/h +/-10%
105 Gallons/Hour
Pressure head: 2.2m +/-10%
Dimensions: 52 x 45 x 50mm
Lifespan: 50.000 hrs at 25°C
Noise emissions: 17.2 dB
Included: 1 × Phobya DC12-220 12Volt Pump
1 x Anti-vibration Pad
2x Brackets

It doesn't even have to run at 12V, starting voltage is 8V. The closer I can get to 12V the better of course.
620mA is the rated current when running at 12V in this instance.
There are probably other pumps out there, this is just an example.
The peak current to start the pump, that could be an issue, as it'll be higher I presume.



Another thing. With USB 2.0, the default current limit is 100mA. A device must be registered on the bus and request more current to be able to draw upto 500mA. I don't know if this is the case for USB 3.0, you would have to check.
How would one go about doing that? Registering a device on the bus.
I read this (it applied to USB 2.0 though): When you short the 2 data lines together, the device will then not attempt to transmit or receive data, but can draw up to 1.8A, if the supply can provide it.
Could that be a solution or workaround for registering the device? It's what is done for chargers that use usb.

[ARandomOWL]

You are right. The startup surge current will be higher than the stated 620mA.

I haven't heard of shorting the data lines. Do you have a reference link?

[Peentje]

You are right. The startup surge current will be higher than the stated 620mA.

I haven't heard of shorting the data lines. Do you have a reference link?

http://pinouts.ru/Slots/USB_pinout.shtml
The paragraph "Dedicated charger mode".

USB 2.0 and USB 3.0 Circuit Protection Considerations (http://www.te.com/content/dam/te/global/english/products/Circuit-Protection/knowledge-center/documents/ap-usb-2.0-3.0-cp-considerations-for-hard-disc-drives.pdf)
That is also an interesting article, it applies to hard disks though, but if you read the paragraph "Overcurrent protection considerations", there's some nice reading about the allowed values.

[ARandomOWL]

http://pinouts.ru/Slots/USB_pinout.shtml
The paragraph "Dedicated charger mode".

Unfortunately that is referring only to USB slaves and not hosts. If for example you have a plug in charger for your phone that connects to it's mini-USB connector, the plug-in charger should short the data lines so that the phone knows it can draw more power.

[Peentje]

Unfortunately that is referring only to USB slaves and not hosts. If for example you have a plug in charger for your phone that connects to it's mini-USB connector, the plug-in charger should short the data lines so that the phone knows it can draw more power.

Right that makes more sense indeed.

Here's an example of a 5V to 12V step-up converter circuit.
It's only up to 500mA though, to meet the usb 2.0 standard.
USB 5V to 12V DC-DC Step-Up Converter (http://www.circuit-projects.com/converter-circuits/usb-5v-to-12v-dc-dc-step-up-converter-by-lt1618.html)

If only there was a way to deal with the inrush current, but that is stuff I don't know enough about.

[ARandomOWL]

I'm afraid that's where my knowledge of the boost topology ends. I'm not sure if a soft start would limit the inrush current. Maybe something to look into.

[OC Maximus]

Watercooling a laptop is overcomplication in the first instance.


Yep. +1 to that.


make that +2
+3 :yepp: