Project stopped working

I have a project that started on an Arduino, moved to breadboard, then to custom PCB. Everything worked great, for a few weeks, then the custom board stopped working. I should note that I'm very new to electronics, I'm more of a programmer. I have measured with a mult-meter and power is going where it should go. I have hooked up a logic probe to the FTDI cable and see DTX raise, RXI transmits 3 pulses, then nothing. As far as I know I do not have a way to test the capacitors or crystal. Can anyone give me any tips or pointers on how to diagnose what has failed?

I should note, I forgot to put on the caps that regulate the power when I connected the board to the brick I was using. No excuse for that, all I can say is I was verifying the board worked while connected to the USB cable (via the FTDI cable), and got carried away. (There are even holes and traces for them on the custom PCB!) I have no idea if that caused the failure of the board.

Any help on how to diagnose is greatly appreciated!

Chad

Any help on how to diagnose is greatly appreciated!

Try checking the value of the resistor and make sure that your transistor is not encabulating.

OK, just pulling your leg ;) Can't really help you out unless you post a schematic (and preferably a layout too) of your custom board.

-- The Ruggeduino: compatible with Arduino UNO, 24V operation, all I/O's fused and protected

Just to make sure, if your custom board is using an external power supply of more than 7-9V, then check that there is no heat damage done to the circuit of the main chip.

Im doubtful that the problem is due to the capacitor or the oscillator as these things are build to last.

Also please specify more details about the board so that it would be easier to assist you.

Hope this Helps! :)

@RuggedCircuits
LOLz… 8) :stuck_out_tongue:

mehtaatur - How do you check for heat damage?

I guess I was thinking there would be some general diagnostics people perform when looking at a working board that stopped working.

I don’t have the schematic view laid out, but can do so if it would be easier to read. I have attached a breadboard view. I also have the PCB view layed out, as that’s what I etched from. I’m really not good at making it look nice. The top board is based on breadboard Arduino projects I found on the net. The bottom board connects to some relays, a radio transmitter, and radio receiver. The rest of the connections are to read the state of some off board magnetic switches.


Again, thanks!

  • Chad

I don't have the schematic view laid out, but can do so if it would be easier to read.

Well it is easier than a physical layout, because that is impossible. I notice a total lack of decoupling capacitors, which doesn't do a circuit any favors. http://www.thebox.myzen.co.uk/Tutorial/De-coupling.html

I guess I was thinking there would be some general diagnostics people perform when looking at a working board that stopped working.

First of all you need to know if your circuit is not breaking any of the rules and if it destroyed some component. If that is fine then you need test equipment, like an oscilloscope to test the circuit and see what bits are working and what bits are not. There is a bit you can do to narrow things down, like can the board upload a sketch, can it make an LED flash and so on.

Grumpy_Mike:
Well it is easier than a physical layout, because that is impossible.

I wasn’t sure. I had seen a lot of breadboard views posted. Unfortunately, I think I’m really bad at laying out a schematic. It’s probably something you get better at over time. I attached my attempt.

Grumpy_Mike:
I notice a total lack of decoupling capacitors, which doesn’t do a circuit any favors.

Thanks for the link! I’ll read up.

Grumpy_Mike:
First of all you need to know if your circuit is not breaking any of the rules and if it destroyed some component.

I can’t tell if any of the components are broken. I don’t have a good way to test anything but power and resistors.

Grumpy_Mike:
If that is fine then you need test equipment, like an oscilloscope to test the circuit and see what bits are working and what bits are not.

Time to expand the toolbox! :slight_smile: Other than testing the crystal, anything the oscilloscope would test the logic probe wouldn’t?

Grumpy_Mike:
There is a bit you can do to narrow things down, like can the board upload a sketch, can it make an LED flash and so on.

I can’t get the ATMEGA chip to talk to my computer over the FTDI connection. I didn’t use a socket, so I have to de-solder the chip to see if it works in an actual Arduino. My thought process was (If I add the FTDI connection to my PCB, no need to use a socket, I can just re-program by directly connecting to the board.)… Lesson learned…

Thanks for all of your help!

Chad

Almost forgot. The relays I’m using are wired differently than the component in Fritzing, so they look like they are wired wrong on the schematic.

And I attached the PCB view, for the physical layout.

  • Chad

Please when printing to image use a higher resolution. That or print to an XPS document. We will need the schematic including part numbers especially the relays, I think you may have dropped the ball so to speak on the supression diodes which may have popped the driver transistors and possibly the chip depending on how the transistors failed, if they failed at all. From what I can tell the outside world connects directly to the chip without any sort of clamping or series resistors. One good ground strike near by and there will be a crater in the middle of that IC.

That is all nice...Do you have a PHYSICAL picture of your fail board ? I would like to see it... your PCB that is.

Edit :

That PCB picture give me a headache. In the PCB picture, The positive ( +5 V) are not connected, also the gnd. I know, it is hard to do your own PCB, from a schematic to a PCB.

His relay driver circuitry is bizzar, clamping diodes are ineffective where placed.

@Grumpy_Mike - That link explains a lot, thanks! I had found other research about adding capacitors to the circuit, however, never really understood it before that explanation.

Thanks again for the help! The response on this site is simply fantastic! I'm learning a lot, should have asked questions before building, as obviously I missed a few things during my research while creating this.

I have updated the schematic posted before to include some missing labels on resistors and capacitors. Here is the details on the parts:

TLP-434 - Radio transmitter (http://www.sparkfun.com/products/10534) RWS-371 - Radio receiver (http://www.sparkfun.com/products/10533) ATmega - It’s actually a 328 LM7805 - voltage regulator (http://www.fairchildsemi.com/ds/LM/LM7805.pdf) Crystal - 16 Mhz (http://www.sparkfun.com/products/536) Relay (2 of them) - 5V - QUAZ -SS - 105D - 1033 (http://pdf1.alldatasheet.com/datasheet-pdf/view/226046/MACOM/QUAZ-SS-105D.html) The second page shows the wiring on the relays I used, which is why the schematic looks wrong. Diode (2 of them) - IN4001 - Radio shack Transistor (2 of them) - PN2222A Resistors - Most are labeled on the schematic D1R and D2R are 1k, not sure why that doesn't show on the schematic. Capacitors - As labeled, however, I forgot to actually put the 10uf caps on that go before and after the voltage rectifier.

I have attached 3 files total.

1 - Updated schematic

2 - picture of top of board

3 - picture of bottom of board

So, if someone gave you this board to diagnose, what would you do next? By the way, when the board was first put together it "worked" in so far as the sensors sensed open and closed, the relays switched when told to, and the radios sent and received. Blowing a component because the circuit was designed wrong is a good guess, but I'm trying to figure out which component(s). Assuming I can find and replace the broken part/parts, would adding in the de-coupling capacitors be enough to fix the overall design? I can add in the 10uF caps before and after the rectifier, and I could add in a .1uF cap directly on the VCC and GND pins of the AT without starting over. I actually have several holes for caps between power and ground after the rectifier that I put there incase I needed them to filter out noise interfering with the radios. Turns out the radios where working just fine for my needs, so I never added them. I had the 10uF caps in the breadboard, just forgot solder them to the PCB...

  • Chad

I can't seem to attach images anymore. The browser just hangs. Is there some other way I can upload them?

  • Chad

FailedBoard-sch2.png|19409x10835