Failure with new build


 
C

Christopher Vance

New member
Good evening everyone. I received a 22.5' WSM for Christmas, and HeaterMeter is the natural companion. I ordered the kit and everything was going along swimmingly until I got to the LCD assembly. I installed it backwards (LCD screen was facing the same side as the components). I was furious and desoldered it using a sucker, braid, and a heat-gun and was able to remove it and get it installed in the proper orientation. When I powered up the board, the backlight did not come on, but I figured whatever, I am pairing it with a RPI so I could deal with the LCD issue later. So I wired it up to my RPI and the webUI came up successfully and was communicating with the HeaterMeter board.

I skipped a resistor for the pit probe so it was showing some odd temperature maybe ~300F even though there wasn't a probe connected, so I "fixed" it by adding the appropriate resistor, but somehow in the process of trying to add more solder to the LCD (to try to fix the backlight) and adding the resistor to the pit probe, the RPI lost its communication to the HeaterMeter board. I don't know what I did.

[EDIT]
Progress: the backlight of the LCD is not working, but after fiddling with the contrast dial, I am able to read the screen. It appears that the HeaterMeter is working with the exception of the LCD backlight and communication with the RPI. The webUI is still reporting failed communication.

I thought I somehow corrupted the software on the AVR chip, so I went into the webUI to reflash the software, but I get the following error: AVR fuses ERROR

I attached photos to the post below as well as a zip archive with less compressed images at the following url: https://drive.google.com/file/d/0Bwj5rc-OCgMMVFozVnhMZkhTSmM/view?usp=sharing. A live gallery here: http://imgur.com/a/zwi2x#0.

The individual photos:
http://i.imgur.com/U9Ixc4S.jpg
http://i.imgur.com/ZEPyRwC.jpg
http://i.imgur.com/gWPhRpW.jpg
http://i.imgur.com/7XV99lA.jpg

Thanks in advance for your suggestions and have a nice evening.
 
Last edited:
Christopher Vance said:
... but somehow in the process of trying to add more solder to the LCD (to try to fix the backlight) and adding the resistor to the pit probe, the RPI lost its communication to the HeaterMeter board. I don't know what I did...
Click to expand...

I have to start out by asking, assuming, you did not have power applied to the HM when you were trying soldering trying to get the LCD backlight to come on? Cause that could be a whole other can of worms.. Hopefully not...

First couple things come to mind are the backlight jumper, it looks like you have that in place. Second, there is an LCD BACKLIGHT setting in the HM Config that can turn the light out if set to 0%. The LCD si controlled by the shift register IC, looks like you have light solder on that chip so perhaps do a little solder touch-up on that one. The LCD and rPi both run on the +5VDC, you can see it marked on both their headers which pin is +5, check that for starters, also check pin 15 (BK) for 5VDC, that is the backlight power pin. On the rPi issue, you might pull the ATMega out of the socket, check that all pins are straight, while it out retouch the solder for its socket. Make sure you insert back with all legs in place and seat the chip properly. Sometimes just re-seating the chip can solve a problem...
 
Another thing to try is go here and check the voltages on your board. Scroll down the post and see the circuit board with voltage designations. Might help to find your problem. That is what I had to do on my first build for my vers. 4.0 HM. Helped solve the problems and that HM has been working for 1 1/2 yrs. Make sure you have a good soldering station with a fine point tip, solder wick for solder removal, solder flux with dispenser (helps in removing solder also). We have all gone through a learning curve but it is all worth the effort. Good Luck
 
Progress update

@RalphTrimble: You assume correctly... I did not power the HeaterMeter when I was soldering :), I also ensured that the backlight setting is at 100%.

Regarding troubleshooting the LCD power pins, pin 15 is providing 5v. I added a bit of solder to the shift IC so it looks good now, but no joy on the backlight. Regarding the RPI, I yanked out a spare (who doesn't have a few laying around the house) and voila it works. I wondered what caused my RPI to lose communication with the HM board, so I haven't tried it on 12v power yet, I'm not sure I'm ready to damage a 2nd RPI's header. I noticed that when powering the first RPI from 12v, it got extremely warm, e.g. uncomfortable to touch, but probably not burn. It would work for a while, but then stop responding. Do you have any input as to why that might be occurring? Perhaps herein lies the root of my problem.

Thanks again!
 
Last edited:
Test all voltages on the HM board without the RPI connected. Bryan has posted this picture of the PCB with all voltages.

hm-424.png
 
Do yourself a favor and load up the schematic diagram and start tracing down these circuits to verify what is right and wrong..... Pin 15 is the +5v for the LCD, pin 16 is the ground, which is controlled by the ATMega via transistor Q4. So you could have an issue with Q4.... I assume when the LCD is full bright pin 16 should be a direct short to gnd, so if you gnd pin 16 (without shorting anything else out!) the LCD backlight should come on full bright. If it does not then you prob broke a trace on the LCD board for the backlight when you desoldered it.... cause you would have verified the +5V and gnd for the backlight on Pin 15/16...

On the rPi, the rPi is running on +5V, a well regulated +5VDC should be present on pin 24 & 26 on the rPi header (on the HM board). From what I can see the pins on the rPi header should read between gnd and 5v, nothing higher should be present... So if the rPi is heating up then perhaps power the HM without the rPi connected and measure the voltage on each pin of the rPi header and look for anything above +5v.... There are 6 pins on the rPi header connected to the ATMega, a few gnds and power, that's it... A solder bridge somewhere or high/unregulated +5v on the rPi header is really the only thing that would make the rPi overheat....
 
@RaphTrimble and @Peter F: I spent some time working on the traces on the board and, because my multimeter is a POS and not being able to tell about bouncing between 3.3v and 0v @ 1hz, voltages looked pretty good on the board. Following Ralph's advice on shorting ground and LCD pin 16 resulting in nothing, I tried shorting LCD pin 16 and the positive leg of the LED and I got backlight. I clearly damaged some trace, as you suspected, and it appears that it was pin 15, so I wired up jumper wire and got the backlight working (it's even controllable from the LCD brightness setting on the HeaterMeter).

After spending an inordinate amount of time checking voltages on the RPI header, things looked good, so I popped another RPI on there and it's working now. So I bench tested it for a few hours and ran into another problem, the RPI goes offline after a couple of hours when connected to 12v (not sure about 5v yet so I will try that next). The HeaterMeter board stays alive throughout the evening but no idea why the RPI is going offline. I'm going to leave it hooked up to a monitor and see if I can grab some debug messages from it or something. I've got a wireless adapter on the way as well so looks like I've got a lot of variables to control! Hopefully I can report back with success.

Do you have any advice as to why the RPI is responsive for a couple of hours but then goes dark?

Thanks!
 
What rPi version are you using, and what wifi dongle? After your rPi drops offline for a while if you reboot do you see a blank space on the graph when it was offline or do the temp continue to get logged (indicating the rPi and HM had been talking while the wifi connection was dead)

On the backlight I am having a hard time following you... If you look at the section of the schematic below, you see pin 15 on the LCD header says BK and it is tapped to +5v through R7 which is a zero ohm jumper wire, so that is the backlight power source. Pin 16 says BKLT_GND and it connects to Q4, which connects to GND and the ATMega. I have to assume Q4 is acting like a valve regulating "how much ground" pin 16 sees... So with +5VDC on Pin 15 and GND on Pin 16 your backlight should be at full brightness... If it is not then there must be a trace busted on the display board.

BK.jpg
 
Sorry, I mustn't have been clear, right now I am using the wired Ethernet for debugging and the RPI pops offline on wired, I will watch to see if the temperature logging remains and report back. In the meantime the backlight issue is solved, it was most definitely a damaged trace on the LCD PCB, but it is now solved, thanks for the wiring diagram which was very helpful in diagnosing that problem. My only remaining issue is to determine why the RPI goes offline so I will try to reproduce the problem and let you know what it could be.

Thanks.
 
I can't believe I'm just now finding this thread. :) As I mentioned in a thread of my own, I also had a colossal brain fart and put the display on the wrong side. I also have a display that fails to back-light. The solder pad for the very right-most pin on the end opposite of the GND pin was damaged when I separated the LCD from the HM board. Is it likely that this is the reason I have no back-light, and how should I approach cobbling a remedy? Pardon me if I've already overlooked the answer, I'm viewing and responding on a very tiny screen.
 
Yah, answer is right above....
Pin15 is the +5VDC power for the backlight, Pin16 is the ground for the backlight. The ground is controlled by the ATMega via Q4 to adjust the brightness of the backlight. If traces to either of those two pins are damaged (either on the HM board or display board) that will cause the backlight to fail. Get a magnifier glass and inspect the traces, check for continuity with a multimeter, find the break and then install a jumper wire to fix it....
 
displaysolderpad63x.jpg
[/URL][/IMG]

As you can see, the solder pad on the B- pin is no longer. I inadvertently tore it from the board when I un-did my ale fueled assembly error. Did that now missing trace just run over to the R0 jumper solder pad? If so, I guess my fix should be easy, jes?
 
PIN 15 comes from the R0 jumper
PIN 16 comes from Q4.

Looks like PIN 16 is the one with the bad trace? Or maybe I am mistaken. At any rate, YES, just run a jumper from the pin over to where the trace used to connect and you should be good to go... (unless there are damaged traces on the display board as well)
 
You must log in or register to reply here.

 

Back
Top