I'm making a remote start system for my generator which I use to power my camper. I built my development unit on my workbench using an Uno and a breadboard. I got all the kinks worked out, and then I made my first prototype. For space saving, the prototype replaced the Uno with a Nano. I tried it out last weekend and had some issues.
the DS18B20 temp sensor didn't work (i haven't had time to look into this, probably bad wiring by me)
the generator battery was dead after sitting overnight (it's an 8 amp-hour battery)
the nano and the receiver module are dead
the plastic project box is slightly melted near the Nano
So everything worked from Fri night to Sat night. When I woke up Sun and hit the start button on the remote, nothing happened. I walked outside to the generator and tried the stock key and nothing happened. The battery was dead. I had to manually start it with the pull cord. After the battery charged, the stock electric start system was working again, but my remote system never did. I shut off the power to the remote to preserve the battery, and the battery never died again. When I got home, I swapped in a new Nano and Rx module and everything worked again. I'm trying to diagnose what happened before I progress with any more testing.
So it seems the dead battery, fried Nano and Rx module, and the melted box are all related. The first picture below shows how the components were located in the box. In the bottom middle right next to the relay block is the Rx module. The second picture is a better view of how the box melted a little near the Nano. The box wasn't near any heat source.
Onto the diagnosis. Obviously, an 8 amp-hour battery draining in about 8 hours is a bad sign. I didn't even think that was possible with this system. I figured something would burn up and short out first. My Nano is powered directly from the 12V battery on the generator, so there's a chance some ripples got to it. If so, I can see it resetting itself, but I wouldn't think it would die. I don't have a scope, so I'm not sure how bad the 12V supply is. This is my first experience with Arduino, so I'm still learning. I'm buying a scope now and reading about voltage regulators (Practical Electronics for Inventors is a big help). Could ripples and spikes have done this much damage? Either way, I will clean up the power supply but I'm not certain that was the cause this weekend. How would that have drained the battery? I'm not using any sleep modes for the Nano. I haven't figured how to do that yet, but I plan to.
The generator 12V goes into the Vin pin on the Nano. The 5V pin is used to supply the temp sensor, relay block, and the Rx module which all take very little power. There is also a servo to actuate the choke, but it gets its power through a UBEC connected to the battery 12V. The UBEC (first pic, to the right of the relay block) power is run through one of the relays. I didn't want the servo to always have power. I only supply power to it when the servo is needed during the starting sequence.
So how did the Nano presumably draw enough current to melt the plastic box and take out the Rx module? Or what else happened?
My schematic is kinda funky and done in powerpoint. I'm still learning how to use appropriate software. I'll try to clean it up and get it posted, but for now, here are the basics:
12V power from generator ---- through switch ---- Arduino Vin
12V power from generator ---- UBEC to convert to 5V ---- through relay block ---- to servo
12V ground from generator ---- everything
Arduino 5V pin ---- Rx Vcc
---- Relay block Vcc
---- temp probe
---- pull up resistor for temp probe
Digtal Arduino pins (input) ---- 5 from Rx module
---- 1 from switch on generator used to control idle speed
(output) ---- 4 to relay block
---- 1 to control servo position
I've found a decent oscilloscope on Craigslist. I hope to pick it up tomorrow. I haven't touched one in 15 years, so it might take me some time to get a reading on the 12V supply from the generator.
Is there anything I should inspect on the fried Nano that would give insight to how this failed?
"The generator 12V goes into the Vin pin on the Nano. The 5V pin is used to supply the temp sensor, relay block, and the Rx module which all take very little power. There is also a servo to actuate the choke, but it gets its power through a UBEC connected to the battery 12V. The UBEC (first pic, to the right of the relay block) power is run through one of the relays. I didn't want the servo to always have power. I only supply power to it when the servo is needed during the starting sequence."
I'd put money on it the Voltage Regulator becoming fried (Voltage Flux?), and then that could lead to the processor getting a full 12v (Causing the heat you see and damage to the box)
Get a 5v linear regulator, and possibly a heatsink for it (12v+ out to the Regulator, 5v out to VIN on the board)
cjdelphi:
I'd put money on it the Voltage Regulator becoming fried (Voltage Flux?), and then that could lead to the processor getting a full 12v (Causing the heat you see and damage to the box)
Get a 5v linear regulator, and possibly a heatsink for it (12v+ out to the Regulator, 5v out to VIN on the board)
That's the step I couldn't see. I couldn't connect the dots between voltage flux and overheating. Thanks. I'll do some reading about your theory and how to test it. I really want to know what happened rather than have a really good idea of what happened.
The remote system was last working at night when I used it shut off the generator. It didn't work the next morning. There shouldn't be any voltage flux during the night when nothing is happening. And the remote was working when it was used to shut off the generator. So most likely, the failure happened during the generator shut down. That makes sense. I'm sure the voltage gets more "fluxy" during the shut down process.
So how does this make the Rx module die? It has 7 wires between it and the Arduino: 5 digital inputs, 5V, and Gnd. Could the Voltage Regulator on the Arduino fry in such a way that it would send the full 12V to the Rx module?
It's possible it indeed did work but once the voltage regulator overheats and dies, all the ones i've ever had all supplied the chip with a full 12v ... for example if it was supplying at times over 12v (what kind of readings do you get, is it a stable voltage?)
All I have now is my multimeter. It gives a reading of 12.2V - 12.6V, depending on whether the generator is running or not. Once I get a scope, I can do a better job of looking for spikes (I think, right?).
This is the UBEC I'm using to supply the servo. It would be easy to change some wiring and drive both the Arduino and the servo from this. My concern is that it's a switching regulator instead of linear. Would this provide the same protection as the 5v linear regulator you recommended?