Page 1 of 1

Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 6:32 pm
by Mettauk
I want to connect a small 5v 25x25x7mm fan without interfering with the GPIO pins so thought of soldering to a Test Point providing 3.3v as the fan does not need to run at full speed and I don't want to get into adding more circuitry. Although its going into a well vented case (https://www.thingiverse.com/thing:3725115) heating is a concern.

I wonder which Test Point to go for TP 4, 5, 11, 26, and 27 show up as 3.3v with TP 11 seeming like a good choice as it's close to the power but can anyone advise?
(Warranty voided I know)
Pi4bBackjpg.jpg
Pi4bBackjpg.jpg (143.67 KiB) Viewed 868 times

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 6:43 pm
by pcmanbob
I would recommend you do not to connect your fan to the 3.3v on your pi , I very much doubt it's rated to provide power for a fan.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 6:52 pm
by drgeoff
pcmanbob wrote:
Wed Jul 17, 2019 6:43 pm
I would recommend you do not to connect your fan to the 3.3v on your pi , I very much doubt it's rated to provide power for a fan.
+1

Use 5 volt and ground points, eg the underside of those pins on the GPIO header. If you then want to reduce fan speed/noise, experiment with various values of a series resistor.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 7:07 pm
by Mettauk
pcmanbob wrote:
Wed Jul 17, 2019 6:43 pm
... I very much doubt it's rated to provide power for a fan.
Fan is rated at 5v 0.2A so at 3.3v something like 0.13A Is that to much?

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 7:10 pm
by goodburner
Mettauk wrote:
Wed Jul 17, 2019 7:07 pm
pcmanbob wrote:
Wed Jul 17, 2019 6:43 pm
... I very much doubt it's rated to provide power for a fan.
Fan is rated at 5v 0.2A so at 3.3v something like 0.13A Is that to much?
I have a 5v 0.2A fan that worked perfectly at 3.3V, just a bit slower and a lot quieter. No resistors were involved.
However I consider my Raspberry Pi too deary to connect a fan directly to its GPIO pins, so this was running out of a arduino that was connected to the raspberry.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 7:14 pm
by drgeoff
Mettauk wrote:
Wed Jul 17, 2019 7:07 pm
pcmanbob wrote:
Wed Jul 17, 2019 6:43 pm
... I very much doubt it's rated to provide power for a fan.
Fan is rated at 5v 0.2A so at 3.3v something like 0.13A Is that to much?
I would not assume that Ohm's Law is applicable to the fan.

I do not know how much current is too much to take from the 3.3 volt rail. I prefer that you, rather than me, does the required experiment.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 7:46 pm
by wh7qq
Be sure to add a reverse diode (anode to positive) across the fan leads. It may already be built into the fan but they don't tell you. 1N4001 and the like. When the motor turns off, an inductive spike may occur that could damage the RPi.

Ohm's law definitely does not apply to motors...they are a partial inductive load, especially at turn on/off. Motors present a short when the rotor is stalled like at start-up so that may be sufficient to disrupt the RPi's start-up, particularly on the severely current limited 3.3V line.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 9:18 pm
by drgeoff
Modern small fans used to cool PC chips are almost always brushless. They have active circuitry switching current on and off in stationary coils. Whatever protection incorporated for that circuitry is more than adequate to also protect an external power rail.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 9:53 pm
by LTolledo
I too would advice against connecting the fan to the 3.3v test points... :!:

my usual advice is to tap the fan's power to the incoming 5v line (prior to RPi4B's internal protection circuity, so that will be TP1 for +5v and TP7 for the GND), and not from the GPIO header 5v line.

but hey! nobody here in the forum can (physically) stop you from insisting on doing what you like... just be prepared for any consequences... 8-)

Re: Pi4b cooling fan wirerd to TP No??

Posted: Wed Jul 17, 2019 11:01 pm
by drgeoff
LTolledo wrote:
Wed Jul 17, 2019 9:53 pm
my usual advice is to tap the fan's power to the incoming 5v line (prior to RPi4B's internal protection circuity, so that will be TP1 for +5v and TP7 for the GND), and not from the GPIO header 5v line.
@LTolledo
Perusal of https://www.raspberrypi.org/documentati ... educed.pdf will reveal to you that TP1 and the 5 volt pins of the GPIO header are electrically equivalent. Ditto for TP7 and ground pins of the GPIO header. There is no polyfuse, perfect diode or any other series protection and thus no "prior" or "after" distinction.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Thu Jul 18, 2019 1:39 am
by LTolledo
drgeoff wrote:
Wed Jul 17, 2019 11:01 pm
LTolledo wrote:
Wed Jul 17, 2019 9:53 pm
my usual advice is to tap the fan's power to the incoming 5v line (prior to RPi4B's internal protection circuity, so that will be TP1 for +5v and TP7 for the GND), and not from the GPIO header 5v line.
@LTolledo
Perusal of https://www.raspberrypi.org/documentati ... educed.pdf will reveal to you that TP1 and the 5 volt pins of the GPIO header are electrically equivalent. Ditto for TP7 and ground pins of the GPIO header. There is no polyfuse, perfect diode or any other series protection and thus no "prior" or "after" distinction.
very true that the TP1 and GPIO header pin2 and pin 4 for the +5V are electrically equivalent.
as well as TP7 and pin 9 and other ground pins on the header
I did checked the schematic before suggesting the TP numbers, I always do before making my fan connections on all RPi boards.

my point for selection these TP is that since its very near the power-in port (USB-C) the required current for the fan is localized as close as possible to the power-in port, and need not travel throughout the board all the way to the GPIO header.
reduces the current load to the foil paths.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Thu Jul 18, 2019 5:20 am
by pik33
I have a lot of my Pis - 3 ad 4 - cooled that way, a 3 cm fan attached to 3v3 on GPIO. No side effects discovered so far.

Re: Pi4b cooling fan wirerd to TP No??

Posted: Thu Jul 18, 2019 8:32 am
by drgeoff
LTolledo wrote:
Thu Jul 18, 2019 1:39 am
drgeoff wrote:
Wed Jul 17, 2019 11:01 pm
Perusal of https://www.raspberrypi.org/documentati ... educed.pdf will reveal to you that TP1 and the 5 volt pins of the GPIO header are electrically equivalent. Ditto for TP7 and ground pins of the GPIO header. There is no polyfuse, perfect diode or any other series protection and thus no "prior" or "after" distinction.
very true that the TP1 and GPIO header pin2 and pin 4 for the +5V are electrically equivalent.
as well as TP7 and pin 9 and other ground pins on the header
I did checked the schematic before suggesting the TP numbers, I always do before making my fan connections on all RPi boards.

my point for selection these TP is that since its very near the power-in port (USB-C) the required current for the fan is localized as close as possible to the power-in port, and need not travel throughout the board all the way to the GPIO header.
reduces the current load to the foil paths.
@LTolledo
You overlook the fact that if a RPi4B is powered via a PoE HAT, all the current for the RPi itself and any power drawing USB peripherals is flowing through those foil paths. Assuming they have been designed to handle that without risk, the much lower current of a small fan is nothing to worry about.