Fan at 100% all the time, how to troubleshoot


 

JMcMillon

New member
As the title says, my fan runs at 100% even when it suppose to be idling.

Some background. My first smoke went great (everything worked as it should), but I realized that I had installed the 4 way switch backwards. I de-soldered it and did a second smoke. That's when I realized the fan was stuck at 100% all of the time. I have since replaced the four way switch, but am still having the same problem.

My digital multi-meter shows ~12 volts at the RJ45 connector between the ground and "blow" pins.

With the LCD up, IC4 (MCP1700) shows 5 v on the "top" pin, 3.3 v on the "left" pin and 0 v on the "right" pin.

Any ideas on how to further trouble shoot?

Would a picture be useful?

Thanks in advance for any help from this fine forum!
 
Could be a solder bridge somewhere at the MOSFET.

Post some clear, high resolution pictures.
 
For whatever reason, I thought MCP1700 was a MOSFET, so ignore my previous measurements, they are just fine for a voltage regulator.
Close up of one side:

IMG_3994.JPG


and of the other:

IMG_3995.JPG
 
The solder mask on the board should prevent this from bridging, but if you look at the blue circle, it looks like there's a solder bridge on the MOSFET.

With the unit unplugged, perform a continuity test between the yellow arrows. They should not be connected.

EIC1W3w.jpg
 
Last edited:
The solder mask on the board should prevent this from bridging, but if you look at the blue circle, it looks like there's a solder bridge on the MOSFET.

With the unit unplugged, perform a continuity test between the yellow arrows. They should not be connected.

EIC1W3w.jpg

No continuity between the two yellow arrows.

Any other points to try?
 
The large Q3 MOSFET (FQU11P06TU) is controlled by the smaller Q1 MOSFET (BS170) above it.

So either there's a problem with the ATMEGA sending a permanent signal to Q1, which in turn is activating Q3, or there's problem with Q3 itself.

Edit: Since the BS170 is an N type, if it's shorted it, it's also going to tie the gate leg of Q3 to ground, which will keep activate Q3 and allow the fan to blow.

Check for a short on the 2 pins in the red box.

qzsVUyql.jpg
 
Last edited:
Also, helpful when looking for shorts, is to get some 90-99 isopropyl alcohol and a toothbrush and give it a good cleaning to get all the flux and any thing that could cause a short.

Also, taking a solder wick around places like the RJ45 jack, where you have small close solder points where things can bridge very easy. Just take the wick and a soldering iron and just touch in between solder pads. After you may need to retouch with some solder if you remove to much. This technique has helped me track down a problems fairly quickly.
 
Apologize for the delay to responding to this thread, work has pulled me away from home for a while.

Steve_M no continuity between the points in red when not powered on. When powered on, there is continuity.

John Bostwick I have given everything a good cleaning and still get fan running at 100%.

Any other ideas?
 
With my DMM in diode checker mode, I placed the red on the drain and black on source of the Q3 MOSFET FQU11P06TU, and get a voltage of .52.

According to limited internet research, this indicates a blown MOSFET, correct? If so, any other parts I should check for replacing?

Thanks!
 
Aloha!
PIN 5 on the ATMega is where the control signal for the blower comes out. If you look at that pin, the trace from it goes across the board right under the tactile switch on its way to the BS170 MOSFET. Since you said the last thing you did before this problem cropped up was change the switch, my guess is something is going on under that switch with this trace from Pin5 of your ATMega.... Something shorting out on that trace under/around the switch which is turning the BS170 on.

Perhaps in the area where I have pointed the arrow in the pic below? Looks like a blob of something there? At any rate, look into what could be shorting out on that trace under the switch..
short
 
Last edited:

 

Back
Top