Connection to HM not available


 
M

Michael Pitra

New member
Hello. First of all, I want to thank the whole community for supporting this great project. I've ordert all parts including the PCBs and soldered everything, including LCD. Powering the RPi works, the SD card boots up. I can connect to my home WiFi network correctly and see the web pages.

But when the HM board is connected, the communication doesn't work correctly - it seems that the serial communication to the ATmega is not working as expected. In the log I can see that the "avrupdate" command failed. Trying it by hand (from the Configuration screens, using the hm.hex file) doesn't work as well. LCD seems to display the top line with blocks, so the shift register and the LCD components seem to function properly.

When powering on, the LED2 and LED3 light up for a short time, so this seems to be ok too. The LED1 does nothing at this time.

The funny thing: I tried some measuring with the multi meter while powered on, and when connecting GND to the 3rd pin of the ATmega (counted from top, left row), the buzzer seem to make some noise. Connecting to the 4th and 5th pin is even weirder: the LCD backlight starts to light up, sometimes it comes to flashing with long intervals, sometimes with long intervals. After playing around with pin 4 and 5 some more, suddenly the "No Pit probe" text line appeared on the LCD. So I thought, the "avrupdate" command worked in the background, and looked on the web pages again: still no communication to the ATmega. Sometimes also all LEDs startet to light up. Somehow everything is working, maybe there's a bad soldered item somewhere, but I can't figure out where to look first.

Can you help me? If there's a need for pictures of my soldered work, just tell me.
 
As an addition: I have the first version of the RPi's with the 256MB on board, there's a pin connector called "P2" on it, that interferes with the buzzer on the HM board. I had to bend two pins for attaching the HM board, but this should have nothing to do with my problems. Just wanted to mention it, maybe this could be a worthy information when added to the HM assembly page.
 
Buzzing when touching Pin 3 (Pin 12) is normal. Your probe is enough to charge the gate that drives the buzzer amplifier. Pin 4 (Pin 11) is the LCD backlight and Pin 5 (Pin 10) is the microcontroller clock so you're basically shorting out its clock source and it is rebooting over and over. That's really not something you want to be doing.

So it sounds like it is working sometimes, which is very strange and probably means you have a loose connection somewhere. LED1 should be the one that comes on most reliably because it is actually driven directly instead of through a buffer. If that's not blinking once when you first plug in, something is definitely not right. I'd also advise disconnecting from the Rpi while you debug. If the LCD backlight is coming on when you plug in the HeaterMeter, you've got a good AVR flash so having the Rpi connected just complicates things and you can short something when probing that fries the Pi.
 
What version HM board are you running? There was one version of HMv4.1 board that had an incomplete trace for LED that prevented it from working......
 
RalphTrimble said:
What version HM board are you running?
Click to expand...

It's an 4.2.4 board, so this should not be the case.

I'm pretty unsure how to start the debugging - are there any schematics on voltage levels I could use? I found some in the forum, but they were for older versions of the board.

Bryan Mayland said:
So it sounds like it is working sometimes, which is very strange and probably means you have a loose connection somewhere. LED1 should be the one that comes on most reliably because it is actually driven directly instead of through a buffer. If that's not blinking once when you first plug in, something is definitely not right. I'd also advise disconnecting from the Rpi while you debug. If the LCD backlight is coming on when you plug in the HeaterMeter, you've got a good AVR flash so having the Rpi connected just complicates things and you can short something when probing that fries the Pi.
Click to expand...

Thanks for the advice debugging without the Pi attached. I only saw LED1 blinking for a fraction of a second once, when plugged the power in. But 2 & 3 are lit up most of the time. LCD backlight is not always on when plugging in, so there is indeed something wrong, but I don't know where to look, all soldered items look ok so far, even under magnifiers (but I'm not that experienced in seeing loose connections).
 
Bryan Mayland said:
Yup, first post of the HeaterMeter 4.2.4 thread
Click to expand...

Thanks for sharing. I debugged all pins and found no error, except for the pins where the bounce should happen. Those have either 0V, 3,3V or 0,9V, but steady. If I follow the paths the voltage levels stay the same. I mentioned one funny thing: while measuring for example the Pi connector, Pin marked as "MIS", I unfortunately touched the GND pin of the Pit connector with my little finger. Then the voltage bounced between 0 and something above 3, was not easy to see on the multimeter. When removing the little finger from the GND pin, the voltage settled to 0,9V. The same thing happened on the FTDI connector on TX / RX pins and on some other pins, I assume they are connected through the PCB.

I also checked all resistors, the color code is correct on every item. So I don't have any idea where to look further. Is there any advice left for me?
 
No resonator, no Heater Meter, that's your (A) problem... I broke one off one time and the HM went from working fine to dead like yours....
 
You must log in or register to reply here.

 

Back
Top