No Communication Issues


 
M

MKader

TVWBB Member
Hello All, I started building my first HeaterMeter this weekend. I got it started up and all the software installed. However upon connection (5V to Pi) I could not see any temp data.
I checked and I had no communication. I tried to update the firmware but got a fuse error (I beleive)

Based on Bryans advice from another thread I tested the folowing (supposed to be 3.3V)

GND to VCC - .89 V
GND to "Above M" - .123V
GND to "Above Last e" - .89V


I also tested the following

GND to "2 Above M" - .89V
GND to "2 Above Last e" - .89V
GND to 3v3 pin on pi - 3.3V
GND to 5v pins on pi - 4.8v
GND to VCC (on ATMEGA) - 89V

Obviously this is wrong, I've checked all the solder connections and see no obvious issues. I have posted my board here to see if anyone finds an obvious issue.

Thanks for any help in advance, I hope I can get it figured out and have an awesome HeaterMeter.

-MK
 
If you're seeing GND to VCC as anything but 3.3V you can stop right there. Check your MCP1700-33 and make sure you have it inserted properly and that you didn't insert a BS170 instead. A few people have swapped these two parts now because the bags may have been mislabeled in the mouser project when you ordered. Get up close and the part number is lasered into the case on the front.
 
I just checked and I can clearly see that the BS170s are in the right position. Its hard to read the MCP1700-33 but unless its just the wrong component it was the only water tower I had left :)

Grabbed a flash light, it says
1700
3302e
TO (something)
3238k6

Solder looks fine to me there as well.
 
Last edited:
Looks like you've got the right 3.3v regulator there... Are those measurements are while powering the boards from the rPi? Seems maybe you have a solder bridge shorting things out or a diode or cap installed in the wrong orientation on the HM board perhaps?
 
RalphTrimble said:
Looks like you've got the right 3.3v regulator there... Are those measurements are while powering the boards from the rPi? Seems maybe you have a solder bridge shorting things out or a diode or cap installed in the wrong orientation on the HM board perhaps?
Click to expand...


That is from powering it through the Raspberry Pi. I took Bryan's advice and did it through that first. When it didn't work I decided putting 12 V to the system may not be the best idea.

I just went back through the entire assembly process and i couldn't find anything that was obviously wrong.

Here is a picture of the entire board

IMG_20140818_162045300_HDR.jpg


Here is a couple pictures closer.

IMG_20140818_211123372_HDR.jpg


IMG_20140818_211131850_HDR.jpg


Here is a photo from the top

IMG_20140818_211908117_HDR.jpg
 
Last edited:
Just for the record, I get this when trying to update the firmware.

Code: 
Connecting to capnbry.net (IP stuff taken out in case it matters)

hm.hex                23% |*******                        | 15702   0:00:03 ETA
hm.hex               100% |*******************************| 66209   0:00:00 ETA
Stopping LinkMeter OK

LinkMeter platform is BCM2708
AVR fuses ERROR

Starting LinkMeter OK

Also, I am attempting to do this without a screen. Because of this, I did not install the potentiometer or the "0 ohm capaciter" or the display. I did install the 8 bit shift register (I think that's what it's called) but I don't think it is necessary w/o the screen.
 
I can't see the diodes in your pic so can't verify the orientation on that, and the pics go blurry when I zoom so I can't say for sure, but there seems to be a bit of stray solder around the board, particularly in the area between IC4 and the ATMega?

Also, the 3.3v regulator used to be optional, now I think it is mandatory? That is because the rPi was fed 5v from the HM and pushed back 3.3v to the HM (making the 3.3v HM regulator optional unless it is run without the rPi). With 3.3v regulator being mandatory now I'm not sure if that means the rPi is no longer feeding back 3.3v to the HM??? Bryan would have to answer that?
 
Last edited:
RalphTrimble said:
I can't see the diodes in your pic so can't verify the orientation on that, and the pics go blurry when I zoom so I can't say for sure, but there seems to be a bit of stray solder around the board, particularly in the area between IC4 and the ATMega?
Click to expand...

I looked over the area. the only place that could be touching is the 2 below the last e in HeaterMeter. I'm 95% sure those are not touching either,but I think those are be connected in the board anyways.
 
You are right, a pain to de-solder even for an experienced tech.
If I may offer some advice, having done this on many occasions, the safest (to preserve the PCB pads) way to remove this part is to cut it off.
Importantly you will need a pair of fine tipped cutters, and close to the IC body, cut each leg. Eventually it will come away, and you are left with the legs still soldered to the board. Then it is a matter of heating each pin, and extracting the leg with tweezers. Once they are done, go back and use a solder sucker or braid on each pad to remove the solder in the hole.
Be careful not to overheat them, as the pads can lift from the board.

This does destroy the IC, but does give best chance for PCB survival.

The alternative is to de-solder all pins, and try to remove as one hit - this does work, but it very tricky and more often will end in catastrophe!
 
Well darn, don't know how I missed that. Actually IIRC I had it in and it fell out and I wasn't careful when I put it back in. Is the shift register necessary for operation without an LCD and LEDs?

If it is I'll likely be ordering a new one to make the install easier.

Thanks,
-MK
 
MKader said:
Well darn, don't know how I missed that. Actually IIRC I had it in and it fell out and I wasn't careful when I put it back in. Is the shift register necessary for operation without an LCD and LEDs?

If it is I'll likely be ordering a new one to make the install easier.

Thanks,
-MK
Click to expand...

Looking at the schematic here, it looks like it's just for the LCD and LED #2 & #3
 
Bryan Mayland said:
Nope. Nor is the 0 ohm resistor or the closest BS170 to the LCD nor the 390 ohm resistors by the LEDs.
Click to expand...

THanks, I'm assuming it could still cause the issue?

Also will things work if some of those are installed? I was wondering if that would leave something open and possibly cause noise.
 
It's a one way street, data comes out of the ATmega and into those parts. Leaving them vacant doesn't do anything that having them connected doesn't do, probably less noise because there's no CMOS switching or LCD backlight. However, as you've found, you can't put pieces you don't want to use in backwards because that totally can **** stuff up.
 
Bryan Mayland said:
It's a one way street, data comes out of the ATmega and into those parts. Leaving them vacant doesn't do anything that having them connected doesn't do, probably less noise because there's no CMOS switching or LCD backlight. However, as you've found, you can't put pieces you don't want to use in backwards because that totally can **** stuff up.
Click to expand...

Ha Ha, yes, that is certainly correct.

I took Dave's advice and just snipped it out. I checked the VCC after and it is 3.3V! I checked the other 2 voltages as well and I got 3.3 on one but on the switching one I got 0 and never got 3.3.

I powered it up through the Pi again and I have communication! I set up the probes and everything seems right!

I think I'm probably going to go ahead and order the display and LEDs. I already see some use cases for the heatermeter I didn't originally intend, where having the display would be nice.

Thanks to everyone for the help!. I'm stoked to have a working HM.
 
You must log in or register to reply here.

 

Back
Top