Fan issues on new build


 

Pat Laf

New member
I know there are many of these posts out there but I haven't seen one with a solution to my issue. The problem I have is even at 0% in manual mode, ~8.8v is being output to the fan. I've gone through this post https://tvwbb.com/showthread.php?70400-2nd-Build-Fan-running-100&p=775338&viewfull=1#post775338 and confirmed that i don't have a short anywhere.

My fan settings are: Fan output Voltage on above 0% min 0% max 100% startup max 100%.

I've looked at my board and I can't find a short anywhere. I've attached photos in case someone has better eyes for spotting that kind of stuff. The RJ45 contacts look a bit messy because of the poking and prodding with my multi-meter. Any help is appreciated!


https://goo.gl/photos/GesHn6hfPJzhgUhi6
https://goo.gl/photos/xFLvspFakzx5GLpX8
 
There are bunch of threads about troubleshooting out there, this is a pretty common problem. It never is the same root cause though, and because there are literally 100 solder points that can cause this, there's no easy way to provide a list of instructions to find the one joint causing a problem.

Step 0) Do not have an output cable connected during testing, as wiring errors in the cable can confuse results. Always use the GND pad on the barrel jack (lower left hand corner) as your multimeter GND. The easiest place (for me) to check the blower output voltage is not at the RJ45 jack itself, but rather if you start at the word SERVO on the jack, go to the right to the first pin you encounter. That's the capacitor on the blower output and it is in the open so it is easy to test without shorting another pin.

Step 1) Change fan output to PULSE mode for troubleshooting. This takes the feedback circuit out of the loop to determine if the problem is in trying to control the voltage versus blind output. We want blind output to troubleshoot.

Step 2) In pulse mode, set to manual 0% and 100% output. Does the output voltage change?

Gonna reuse this drawing


Step 3) At manual 0% output, pin 4, 6, and 7 should be 12V. At manual 100% output, pin 4 and 6 should be low (around 0V-1V) and 7 should stay 12V.
 
Hi Bryan, thanks for the reply.

With the fan on pulse mode, here are my readings:

0% pins 4,6 -> 8.6v pin 7 -> 12,1v
100% pins 4,6 -> 9.1v

There's something wrong somewhere.... just not sure where to look. I can see that the BLWR pin on the Arduino does range from 0.04v with fan at 0% and 3.3v with fan at 100%.

Thanks!
 
Q1 is bleeding current through R14 to control the voltage on the gate of Q3, when there is no current through Q1 the voltage on the gate should be 12V, at about 5.5ma the voltage should be 0V.
You have verified that the voltage which controls Q1 is sweeping properly between 0-3.3v at the atmega, are you seeing the same at the gate of Q1? If so, Q1 is getting biased properly, so suspect Q1 being bad, or R14 being wrong value, or a short or bad solder joint somewhere. With Q3 in circuit you cannot rule out an issue caused by Q3.
What you are reporting on pin 4,6 is not showing proper bias for Q3, at 8.6-9.1V the blower should be running at roughly 25% and not vary much from 0-100%. If you remove Q1 pins 4,6 should go to 12V and the blower should turn off, if you tap a 2.2K resistor from Pin 4,6 to ground with Q1 removed the blower should go from off to about 50%, if you tap pin 4-6 to gnd the blower should blow full blast. If that happens that verifies Q3 and the blower driver part of the circuit is working.
I would look at the value of R14 and perhaps swap Q1, while q1 is out of circuit do the same tests you did above and see what the voltage is on pin 4,6?... and do the test on Q3 to see if you get 50% with 2.2K to ground and 100% blower speed with pin 4,6 shorted to ground to verify Q3 is functioning properly.
 
Last edited:
As Ralph said, I'd check the voltage at Q1's center pin (the pin closest to the front of the board) at 0% and 100%. You should see it match the BLW pin. Also verify that the right pin has continuity to GND, and the left pin is connected to #4.

Again copying Ralph, the next step is to remove Q1. Without Q1, the blower should not run at all. If you short the left and right holes of Q1, then the fan should run full speed. If that all checks out, then there's a good chance Q1 is toast. MOSFETs can be sensitive to ESD in a way that can permanently damage them if they get a little shock while being handled so perhaps that is what we are looking at. I'm not sure what the measurements would be on a broken FET, but if I measure resistance to GND (with the multimeter's black/negative probe on GND, this is important) I get 125kohm from the left pin to GND, 0.7Mohm from the center pin to GND, and <1ohm from the right pin to GND. This is consistent across all my HeaterMeters I have here so it might be useful in diagnosing.
 
Thanks for the help guys! I got it working now. Followed your advice, determined that Q1 was bad. Ripped out Q2 as I read somewhere that's for the buzzer and put it in place of Q1 and everything is normal now. Looks like I'll be ordering a replacement BS170.
 
Glad you got it working!... get a couple bs170's when you order, they're cheap enough and good to have around...
 
A ha! Well that is good news, but strange that the MOSFET was blown. If you can't find BS170s cheap, then you can also use 2N7000. The 2N7000 are identical except their pins are reversed so the part would just need to be installed rotated 180 degrees. Radio Shack used to carry these if Canada still has Radio Shacks.
 
All our Radio Shacks were converted to "The Source" a little while back. Not sure if they still carry things like this though, it's more like a small best buy now. I've ordered a bunch off of ebay for 1$. It'll take them a while to get here but that's not a problem. I'll just rely on the alarm on my phone until they come in. Thanks again for the help guys.
 

 

Back
Top