Troubleshooting - Fanr running at 100%


 

Henrik B

New member
Troubleshooting - Fan running at 100%

Finished my first HM build, v.4.3. Not having soldered for over 20 years, I'm very happy with the result! Installation and booting up and accessing webGUI via my rPi3 works great, thermocouple and temp probes are all working, however the fan is running at 100% directly form start up. I have switched to pulse, running the fan manually but I can't get the fan to run as intended.

I have read through this thread:
https://tvwbb.com/showthread.php?70400-2nd-Build-Fan-running-100

With the following measurements, for pulse mode:

I have thoroughly checked the board for shorts or poor solder fil but it looks good to me, have also cleaned the board from flux and dirt.

RalphTrimble's measurements as per post #5:

Q1 to pin 5 ATMEga reads 0V regardless of fan speed (in pulse)
Q3 to Q1 reads 0V regardless of speed
This is the same with pin 5 bent out of position.

I did not follow where to measure Ralph's other post with the schematic picture, but I do have 12V on one of the Q1's legs

I did the continuity test as Bryan explained and got the correct result.

Unfortunately, my knowledge about the logic behind on how the board is created and supposed to work is limited, but I suspect that the problem lies in volt regulation somewhere, as the fan seems to be getting 12V all the time.

Any ideas where to look next?
 
Last edited:
If you have pin 5 pulled out of the socket and it is not sweeping from 0 to 3.3v you have some sort of logic problem. Either the HM config settings are wrong for the fan or the ATMega isn't programed properly, or the ATMega is hosed.
I would check the HM config, set fan on above 0%, min=0%, max=100%, startup max=100%. If that doesnt do it then try re-flashing the ATMega.
 
The sittings are set to above 0%, min 0%, max 100%, start up max 100%, fan is still running contineously at 100%. Where do I re-flash the ATMega? I have tried the AVR firmware through the WebGUI, tested all online 4.x versions available with no change. I kept the fan connected while flashing the firmware and it ran full speed even when stoping and re-booting the unit.
 
Sounds like you have the fan connected to the wrong wire on the CAT5 jack, or you have something wrong in the blower circuit. Though reporting Pin5 of the ATMega is not changing (even when out of circuit) is confusing and improper, 0v on that pin should turn the blower off, so it appears to be locked in the off state while you are reporting the blower is locked in the on state. Meaning, the blower should be off if Pin5 has 0v.
I would check continuity from your blower + wire to the BLOW pin on the CAT5 jack of the hm to confirm you have the right wire, and check Q1 to be sure it is a BS170, and make sure you do not have any shorts at Q1 and Q3, and make sure R14 is proper value and soldered well on both sides of the board.
 
I also find this difficult, looked through a lot of threads yesterday and still no clue!

Got a brand new ATMega and re-wrote the SD card with a newly downloaded imsge, but with the same disapoibting result, so I agree thet the ATMegas is not the problem.

Cheched the blower wiering, black has continuity with ground and red with blow. Switching these kills the fan at any % point.

Have checked the soldering for Q1, Q3, R14 and nothing looks out of the ordinary, will upload som pictures. I took of Q1 yesterday to check for broken legs, all are good and can confirm that it is a BS170.

Checking the voltage on the Q1 and Q3 I get the following, measuring to ground
Q1:
S: 0V @ 0%, 0V @ 100%
G: 1.3V @ 0%, 1.3V @ 100%
D: 1.5V @ 0% 1,5V @ 100%

Q3:
S: 12.2V @ 0% 12.2V @ 100%
D: 12V @ 0% 12.2V @ 100%
G: 1.5V @ 0% 1.5V @ 100%

Above measurements are taken with out the fan connected to the board and the ATMega pin 5 disconnected.

What can I measure on the R14 to determine if it works as it should.
 
Pit temp has been set close to ambient, but I run the fan manually so the pit temp shouldnt impact the fan. However it runs at 100% regardless if the pit temp i set above or below the temp the prove reads.
 
I would assume you set the fan manually to 0% and 100%, so temp settings should not be an issue.
Post pics, cause what you report is odd... 0v into Q1 should have the blower off yet you say you have 0v and blower is full. Q3 according to your chart IS biased for the blower to be running nearly full blast, so that part makes sense, but what you have going on with Q1 is a mystery. Maybe try another BS170? Could have burned it while soldering or something.
 
Yes fan speed is set to 0 and 100 manually.

I was also considering the idea that either Q1 or Q3 is damaged in som way. Unfortunately I don't have any extra lying arround so I will have to order them online. What would happen if I switch Q1 with either Q2 or Q5? How would I know doing this that th BS170 in Q1 has been fried? Or could I as a first step remove Q1 and try the HM without it all together?
 
You can take the BS170 from the buzzer, your buzzer just wont work until you replace it. IDK the transistor number, you can either look it up on the schematic or perhaps you can follow the traces on the board to identify it.
If you remove Q1, while it is out test the voltage on the holes where it installs, particularly the center one. it should sweep from 0-3.3v as the HM goes from 0-100%. If it doesnt then there is an issue (measure 3.3v somewhere else to make sure your 3.3v supply is good and meter is working) It's possible Q1 is dead and shoring out the control voltage, so things might look right after you remove it.
 
Thanks Ralph for all the support, I hope you can keep guiding me to find the solution to my problem. It's not fixed yet.

Removed the BS170 in Q1 and measured the holes, left reads 12.2V at 0 and 100%, center and ground both read 0V at both 0 and 100%.

Measured 3.3V on the RPi so the 3.3V generation on the board works...

Will try to upload som pictures of my board, I have a feeling that I have messed it up somehow, despite that I have checked and double checked components and the soldering.
 
center pin of Q1 to gnd should read 0v and 0% and 3.3v at 100%, so definitely a problem there. Does the center pin show a short to ground using continuity test? Is the blower off now with Q1 removed?
 
No continuity between center and gnd or left and gnd.

Fan is not running with Q1 removed.

The engineer in me, tells me this points to component error and not a short on the board. However as I'm chemical engineer and not electronics, my conclusions in this are not alwasys worth that mouch...

Ralph, you mentioned the R14 capacitor earlier, could it be this, or Q5? Orare both these OK as the board does generate 3.3V?
 
I'm confused, R14 is a resistor, not a capacitor. It is the component that connects the bias voltage for Q3. It should be 2.2K.

Q5 is the transistor that is the servo booster, so it is in the servo circuit, not your issue.

IC4 is the 3.3v regulator, if you have 3.3v on your board it is working.

As for your blower problem, the situation is still a bit confusing. Since the blower turns off with Q1 removed that does point to an issue with Q1, however, since the center leg of Q1 does not sweep from 0v to 3.3v when you go from 0% to 100% that points to a problem with the ATMega. It is not impossible that a burnt Q1 killed the ATMega, or two. If the center leg of Q1 does not go from 0-3.3v your HM will never work.

If your Q1 came out intact you can test it with your multimeter, should read like two diodes, if you see dead shorts in both directions she's dead.

You can steal the BS170 from the Q2 position, that is the buzzer driver, with that one missing everything will work but the buzzer wont buzz. Make sure not to over heat it coming out or going in though.

Since your blower is off now with Q1 out, and you have 0v on the center hole for Q1, if you put a good BS170 in there your blower should remain off. If you get that far then you have made some progress, but again, you will not have a working HM until the center pin on Q1 goes from 0-3.3v when your HM output goes from 0-100%. So your second problem to troubleshoot is why pin5 on your ATMega is not doing what it should.

Somewhere on this forum is a voltage chart showing what you should expect where. I suggest you find it and test the voltages on the ATMega, particularly all the places where 3.3v appears. If all those voltages check out, and your blower stays off with a good Q1 in place, unfortunately it looks like your ATMega may be toast. (again)
 
Last edited:
Somewhere on this forum is a voltage chart showing what you should expect where. I suggest you find it and test the voltages on the ATMega, particularly all the places where 3.3v appears. If all those voltages check out, and your blower stays off with a good Q1 in place, unfortunately it looks like your ATMega may be toast. (again)

It's here https://heatermeter.com/devel/pcb/hm-4.3/HeaterMeter433BaseV.png

iweXOmy.png
 
Got the BS170 out from Q2, measured both the Q1 and the Q2 with my multimeter giving different results, what should it read?

Oh well solderd the Q2 MOSFET in the Q1 position, assembled the boards, powered up, hooked up fan... I give up! Full speed at 0% #*£*?!?!$€% OK obviously that coponent is fried as well, took it of the board and broke of two legs, I give up again!!!

Tomorrow I will order a new set of compoents and a PCB for the base board and start from scratch, I'm having withdrawls from my BBQ adiction and need a new controller as my CyberQ is dead. But I will not give up on the existing board, I want to find what is causing this issue.

Obviously I need to order two extra BS170 and an extra ATMega. Should I get anyother extra components while I'm at it?

Thanks for the support so far, I will let you know the progress with the new board and measurements on sick one when I have the energy and components to go at it.
 
What you may want to try now that you've broken the legs off your MOSFET is to remove Q1 completely. Without Q1, the blower should never run regardless of output %. If it still runs full speed, then check the legs of the Q3 (big power mosfet). You shouldn't get continuity between them at all. Looking from the top (Boxy the Robot graphic side), you should also measure ~2200 ohms resistance between the left pin and the right pin.

Lastly, if you need to order anything from the HeaterMeter at heatermeter@capnbry.net and I can send you a coupon code for half off anything you need.
 
Got the components I ordered yesterday and mounted new MOSFETS in Q1 and Q2, installed a new ATMega, assambled the boards and started it up. To the moment of truth connected the fan, changed to manual mode and finally there is controlled air flow from the fan.

Now I just need to make an adapter for the fan and hook it up to my Kamado, I can already smell the smoke and taste the ribs.

Thanks for the help with the trouble shooting, you were on it right away with a faulty MOSFET and thanks to your help I have my HeaterMeter up and running!
 

 

Back
Top