Linkmeter AVR fuses ERROR


 
A

Alexey G

New member
I have soldered the board, but when I turn it on and go to LinkMeter, I get the following error:

Stopping LinkMeter OK

LinkMeter platform is BCM2708
Loading SPI modules...
AVR fuses ERROR

Starting LinkMeter OK

Attached are the pictures of the boards, and I can't seem to find the problem. I highlighted one pin in particular that doesn't have continuity with ground, which I thought it should. I checked the MOS<>MIS connections from the rPi to Atmega, and SCK, and RESET, and all have continuity.

Any advice?
https://www.dropbox.com/s/w68rmhdhdrhqcfi/IMG_4241.JPG copy.jpg
https://www.dropbox.com/s/qxmwy11tx30utd2/IMG_4240.JPG copy.jpg

IMG_4241.JPG%20copy.jpg

IMG_4240.JPG%20copy.jpg
 
That pin you've circled is just connected to SVO on the microcontroller I believe, it's the servo level shifter. Did you try measuring the voltage at RST on the microcontroller? That should be 3.3V, along with VCC and AVC. Also verify (with no power) that you have no continuity between the MOS/MIS/SCK/G25/3V3/GND pins, like test one to each of the others and repeat for all pins.
 
Bryan Mayland said:
That pin you've circled is just connected to SVO on the microcontroller I believe, it's the servo level shifter. Did you try measuring the voltage at RST on the microcontroller? That should be 3.3V, along with VCC and AVC. Also verify (with no power) that you have no continuity between the MOS/MIS/SCK/G25/3V3/GND pins, like test one to each of the others and repeat for all pins.
Click to expand...

Thank you. I measured as follows:

RST= 2.8V
VCC=3.3V
AVC=3.3V

There's not continuity between MOS/MIS/SCK/G25/3V3/GND pins, however strangely enough there also isn't continuity between the two 3V3 pins on the rPi header. Any ideas?
 
The 3V3 pins on the RPi header aren't connected on the HeaterMeter board. You'll probably see 3.3V coming in on the pins but that's because they're connected on the Pi PCB. We generate our own 3.3V with a linear regulator to try and reduce the noise on the analog side.

I just checked one of my boards and I also have 2.8V at the RST pin so that's good too (I forgot that the pin pulls some current so it is below the 3.3V). I've been looking over your PCB for anything that looks wrong and it's actually a very good soldering job so nothing stands out as being wrong. What about checking the continuity from the atmega's GND and AGN pins to the power jack's GND? Maybe try checking continuity from the underside of the Pi PCB to the HeaterMeter PCB on those MOS/MIS/SCK/G25 pins? I'm really at a loss as to why it isn't working.
 
I would first check for shorts on adjacent pins of the rPi header (with the rPi removed), if you dont see any shorts then I would suggest you pull the ATMega from the socket then reflow the solder joints on the entire board, shouldn't take too long. Don't flow a ton of fresh solder, just enough to melt the solder that is in place, wiggle the leg a of the component a bit with the soldering iron to encourage solder to flow to the other side of the board. Then use a soft brush and some isopropyl alcohol and scrub the solder side of the board. After it dries install the ATMega, connect the rPi and try again. Sometimes using the "shotgun" method is easier than tracking down the specific point of fault.
 
Last edited:
Sorry to be spamming but been trying and getting nowhere. One last piece of info is that I haven't soldered the LEDs yet, because I was waiting for the case to print. Not sure if this would impact anything? I doubt it.

At this point I might just have to buy another board and resolder everything...
 
Alexey G said:
Is there a way to test the atmega or ceramic oscillator? Maybe it's one of those two?
Click to expand...
I can't think of any way to test the atmega without having something else to put it in. The oscillator you linked seems ok, and before the ATmega is programmed the oscillator isn't even used (the chip runs at 1MHz by default off the internal clock). It is possible on an earlier boot, the fuses were set to "external clock source" and now the oscillator is coming into play, but there's no way to test that at all without a higher end scope. The LEDs and in fact the whole LCD/Button board doesn't need to be present at all for everything to work properly. In fact, to reduce the complexity of the parts connected, I'd recommend doing all testing without the LCD/Button board connected at all.

If you've been testing without the LCD/Button board, I am sorry I still don't have any other direction to point you-- it is a little baffling. If you want to mail it to me to take a look at I could do that for $10 bucks to cover cost of time and return shipping? But no guarantee I'll be able to bring it to life. If you're interested email me at heatermeter@capnbry.net.
 
Is there another RasPi you can borrow to test with? I had a RasPi that had a messed up SPI bus that wasn't able to program the atmega chip.
 
You must log in or register to reply here.

 

Back
Top