New build - No power


 

Dominick A

TVWBB Fan
I just completed the HM portion of the build, less the 4 way tactile button that I have on order. I mated the HM with the rPi, installed the openWRT, asked the wife for a drum roll, plugged in a microUSB cord into the Pi and.......nothing.

I unmated the HM from the rPi, plugged the power back in and it lit up. Also plugged in to my network via ethernet and could see it on my network. However, when the two units were mated, no lights flashed on the rPi, could not see it on my network, and the LCD did not light up.

Where can I start to troubleshoot my HM build? Other than the button, I'm pretty certain all parts are installed correctly. Solder looks ok, but this was my first soldering job, so what do I know, really!? Are there points on the HM PCB that I can test with my multimeter to see where I may have gone wrong?

Any help, as basic as possible, would be extremely appreciated!
 
If the rPi boots on its own but not when mated to the HM that sounds like you have something wrong on the HM board. Either a part backward, or a solder bridge somewhere, something like that which would drag the voltage down and prevent the rPi from booting.
Post some pics of the board front and back and maybe someone will spot your problem.
 
to me it looks like this area wasn't soldered, might just be lighting.
M5emz6o.png
[/IMG]
 
I'd measure the resistance between the VCC and GND on the FTDI connector and make sure you don't have a short in the 3.3V side, then measure it between the +5V at the ethernet jack and GND to check for a short in the 5V side. They should be reasonably high, i.e. not just a few ohms. If they are shorts then you "just" need to look over your board and find a solder bridge. If they're ok then check all your little black water towers (BS170 / MCP1700) to make sure you have the right parts in there and you don't have two swapped.
 
I'd measure the resistance between the VCC and GND on the FTDI connector and make sure you don't have a short in the 3.3V side, then measure it between the +5V at the ethernet jack and GND to check for a short in the 5V side. They should be reasonably high, i.e. not just a few ohms. If they are shorts then you "just" need to look over your board and find a solder bridge. If they're ok then check all your little black water towers (BS170 / MCP1700) to make sure you have the right parts in there and you don't have two swapped.

Bryan - would I measure using the 12VDC/1A power brick plugged into the HM power jack? Just want to make sure I don't kill anything.

Dave - I'll double check those joints tonight. They definitely got some solder, but maybe not enough? They do stand out though in color compared to the rest of my board.

Thanks guys, I'll report back this evening with an update, much appreciated!
 
I don't spot any components installed backward, though you should check to be sure you put the MCP1700-33 over there by the ATMega and not a BS170 MOSFET. Those parts look just the same except for the label on them.
I would take some isopropyl alcohol and a soft brush (like a toothbrush) and scrub the board a bit to get rid of all that char and flux first so you can see what is going on there. Then inspect for solder joints that may be bridged or component legs that may be touching.
As Bryan stated, with NO POWER applied to the board measure DC resistance from gnd to the +5V line, and the 3.3v line, you should have a high resistance in both cases. If you read a low resistance you have a short somewhere or a wrong/damaged component installed somewhere. If you read high resistance to gnd on both the +5V and +3.3V then I would power the HM board with the 12V supply (WITHOUT the rPi connected) and see what you get. Measure the +5V and +3.3V, post back the findings if you dont get it working.
 
https://lh3.googleusercontent.com/-...AAAAAAAArQ/pz8lkLbMWuc/w346-h231/IMG_2970.png

This looks like a "cold solder joint" it can happen if the joint gets bumped or moved while the solder is still molten. The result can amazingly produce a non-conductive connection and can be a nightmare to troubleshoot. They usually have a frosty or crystalline appearance instead of being shiny like a good solder joint. Ideally you would remove the solder with solder wick and replace with new solder, but a lot of times can get by with touching your iron to the joint until it is molten then being careful to not move while it cools.

I have not posted a lot of photos in forums...how do you insert a decent sized photo and have it where it can be viewed without clicking a link??
 
Last edited:
The thing is, he said the rPi powers up on its own but is dead when mated with the HM board. Assuming he had the rPi mated to the HM properly that would indicate some sort of short or bad/wrong component somewhere on the HM board that is dragging the voltage down and preventing the rPi from booting. A cold solder joint might make the HM board fail to boot but it shouldn't hinder the rPi from booting.

EDIT: One other thought, is it possible the USB power supply you are using to power the rPi craps out with the extra load of the HM? Were you using a USB cable connected to a computer or USB wall wart charger to power the rPi?

Another Edit: Also make sure that nothing on the rPi is physically touching the HM board when you mate them, like shorting the display header pins or whatever... Try mating them slightly less than fully seated, throw a piece of paper between the boards temporarily to prevent them from touching. Someone recently posted about having a power problem and this was the issue.
 
Last edited:
Bryan - would I measure using the 12VDC/1A power brick plugged into the HM power jack? Just want to make sure I don't kill anything.
Nope, test it with nothing plugged in at all. The way you say the Pi doesn't turn on at all makes it sound like you've got dead short somewhere.
 
If I'm using one of these. If this isn't delivering enough, what is recommended instead of plugging in the 12V barrel jack for the first boot?
 
Do you have a multimeter?
As suggested, set the meter to measure resistance, measure from GROUND to a 5V pin, and from GROUND to a 3.3V pin (without the rPi or any power connected to the HM board) You should see a very high resistance in both instances, if you are reading a low resistance from GND to 12V, 5V or 3.3V that indicates you have a short somewhere on the board.

I just pulled a HM board and did this resistance test.... Orient the HM board so you are looking at the board like you would when using it (button facing you with LCD on top) To the right of the button, between the button and the ATMege is the 3.3V regulator. It is the three pins that make a triangle there. With no power or rPi connected to the HM board put your meter in resistance mode and put the black lead on the lower right pin (GND) of the 3.3v regulator, then put the red meter lead on the TOP (5V input) pin, the value will likely fluctuate as caps charge but should settle somewhere around 5K. Now move the red meter lead to the pin on the LEFT (3.3V output), it will likely fluctuate a bit but should be way up around 1.5M resistance. You can also measure resistance across the power jack 12V and ground, should measure a couple Meg.

If you measure similar resistance on your board then I would power the HM with the 12V supply (without the rPi connected) and see what happens. If you measure much lower resistance on any of those pins then I would not apply power to the board in any fashion until you clear up whatever is shorting things out.
 
Last edited:
Do you have a multimeter?
As suggested, set the meter to measure resistance, measure from GROUND to a 5V pin, and from GROUND to a 3.3V pin (without the rPi or any power connected to the HM board) You should see a very high resistance in both instances, if you are reading a low resistance from GND to 12V, 5V or 3.3V that indicates you have a short somewhere on the board.

I just pulled a HM board and did this resistance test.... Orient the HM board so you are looking at the board like you would when using it (button facing you with LCD on top) To the right of the button, between the button and the ATMege is the 3.3V regulator. It is the three pins that make a triangle there. With no power or rPi connected to the HM board put your meter in resistance mode and put the black lead on the lower right pin (GND) of the 3.3v regulator, then put the red meter lead on the TOP (5V input) pin, the value will likely fluctuate as caps charge but should settle somewhere around 5K. Now move the red meter lead to the pin on the LEFT (3.3V output), it will likely fluctuate a bit but should be way up around 1.5M resistance. You can also measure resistance across the power jack 12V and ground, should measure a couple Meg.

If you measure similar resistance on your board then I would power the HM with the 12V supply (without the rPi connected) and see what happens. If you measure much lower resistance on any of those pins then I would not apply power to the board in any fashion until you clear up whatever is shorting things out.


The top pin registered 0.00. The bottom left pin registered 1.68. I also measured at the Ethernet jack and got 0.00, while the power jack next to it measured 1.47. The output at the FTDI connector measured 1.68. Does this indicate anything specific to look at, or just that I likely have a solder bridge somewhere? Upon closer look, I think I do have a couple joints that could be candidates for a de/re solder.
 
Last edited:
I agree, what are you measuring? 0, 1.68 and 1..47 are useless numbers unless you tell us the units and what you are measuring? 1.47Meg Ohms? 1.47 Ohms, 1.47Volts???

If I had to try to make sense of what little info you posted I would guess those measurements are Mega Ohms? It seems like you are saying you read a dead short on the 5V input leg of the 3.3v regulator, but 1.68M Ohm on the 3.3v output?

There is a lot of assumption going on there because the info you posted was not at all clear. However, if my assumptions are correct it seems you have a dead short on the 5V line somewhere.
 
Last edited:
What exactly where you measuring? Ohms, Volts, Amps, gravity?

1.47 gigawatts!!

Seriously though, I believed I was measuring ohms. I'm learning this as a I go. My multimeter was set to 20K Ohms on the dial. Not sure if that's even right for this test, but since I didn't get a 1 showing it was overloaded, I figured I was in the right neighborhood, or at least proving that nothing was flowing through? I got my multi-meter some time ago from my father-in-law. I think it was free at Harbor Freight. I'm reading reviews now that it might be garbage. If that's the case, any recommendations on a good multimeter?

Back to the Back to the Future reference.....of all the things for that movie to come true, it's likely going to be the Cubs winning the World Series this year. Not saying that's more ridiculous than time travel or hover boards, but I digress.
 
OK, well then perhaps my assumptions are correct. If you measured in resistance mode from the right pin (or any gnd on the board) to the top pin of the 3.3v regulator (which is the +5v input pin) and got 0 Ohms that means you have a dead short on the 5VDC line somewhere, which you will have to locate and correct before things start to work.
 
OK, well then perhaps my assumptions are correct. If you measured in resistance mode from the right pin (or any gnd on the board) to the top pin of the 3.3v regulator (which is the +5v input pin) and got 0 Ohms that means you have a dead short on the 5VDC line somewhere, which you will have to locate and correct before things start to work.

Got it. Sorry for the confusion. I was testing later than I wanted to last night and need to remember to give ALL of the possible info I can since I'm new to this game. I'll be reworking some joints tonight and let you know what happens. Thanks again!
 
Like I said previously, it is impossible for a cold solder joint to create a short circuit... IF you have 0 Ohms between gnd and the 5VDC line then you are not looking for a cold solder joint, you are looking for a short circuit in the form of a SOLDER BRIDGE (solder connecting two points together that should NOT be connected together), an improper component installed somewhere (wrong value or installed backward) or a blown component somewhere that is internally shorted.

I have looked at the pics of the board and it seems the transistors, chips, diodes and caps are installed in the proper orientation, so count that out. Next thing you want to do is read the label on the 3.3v regulator and make sure it says MCP1700-33 on it and NOT BS170. The BS170 is a MOSFET TRANSISTOR that looks identical to the 3.3v regulator, you wouldn't be the first guy who got them mixed up. After that you need to clean up the board to get rid of all that flux and char and try to find the solder bridge that is shorting things out.
 

 

Back
Top