Hello Folks, I have a pump, solenoid valve and a Arduino uno connected to same 12v line. the uno is part of a water flow meter. when the pump or solenoid is turned on, it creates a voltage spike on the line and uno freaks out. either it shows garbage data on ssd1309 or just hangs. I have tried to use 12v dc-dc iso power supply from Ali express (link below) but it does not stop the spike. when I power the water flow meter from a external battery bank, it works perfect. so I put a battery bank between the main 12v supply and the water flow meter but I had the same issue. Please suggest how can I stop this from happening.

[https://www.aliexpress.com/item/400...//www.aliexpress.com/item/4001069352009.html)

 

Are you using a single point ground at the 12V source where the uno and pump/solenoid have separate ground lines to that source?

Does the uno and pump/solenoid have decoupling capacitors directly across their supply connections to their local ground?

Post a diagram of exactly how the power and grounds are wired to all the components with all the decoupling capacitors.

 

Are you using a single point ground at the 12V source where the uno and pump/solenoid have separate ground lines to that source?

Does the uno and pump/solenoid have decoupling capacitors directly across their supply connections to their local ground?

Post a diagram of exactly how the power and grounds are wired to all the components with all the decoupling capacitors.

There are no capacitors. I am very new to electronics so I don’t know where they would connect or what capacity I would need. Everything shares the same ground and positive connection.

 

There are no capacitors. I am very new to electronics so I don’t know where they would connect or what capacity I would need. Everything shares the same ground and positive connection.

Thus it's not surprising that you are having problems.

So you need to wire it as I stated in my post with proper decoupling across all the devices.
Look up "single point ground" and "circuit decoupling" for further details.

In particular you don't want any current from the solenoid/pump to go through the ground/power wires from the supply that go to the uno.

 

Thus it's not surprising that you are having problems.

So you need to wire it as I stated in my post with proper decoupling across all the devices.
Look up "single point ground" and "circuit decoupling" for further details.

In particular you don't want any current from the solenoid/pump to go through the ground/power wires from the supply that go to the uno.

I did some research. So I have this installed in my 4wd. Does it mean that both my pump and the solenoid need to have ground at different points on the chasis? Along with that I need to install a decoupling capacitor on both pump and solenoid? How do I figure the capacity of the capacitor that I need? Thanks

 

Hi user
The 0V/Common lines should be connected as this edited drawing.
Place a 100nF and a 100uF across each device +V and 0V, close to the device.
E

 

Hi user
The 0V/Common lines should be connected as this edited drawing.
Place a 100nF and a 100uF across each device +V and 0V, close to the device.
E

So 100nF goes on the positive and 100uF goes on the negative of each device?

 

HI use,
No, check this sketch.
E

 

The circuit shown in post #8 will still provide fairly effective coupling of the spikes if the positive supply line s shared between the three devices. The filtering as shown for the arduino is fine, but really it should not be sharing the same positive supply line with either of the power devices. The good news is that the processor does not draw much current and so that separate power wire does not need to be a heavy wire like the pump requires.

 

The circuit shown in post #8 will still provide fairly effective coupling of the spikes if the positive supply line s shared between the three devices. The filtering as shown for the arduino is fine, but really it should not be sharing the same positive supply line with either of the power devices. The good news is that the processor does not draw much current and so that separate power wire does not need to be a heavy wire like the pump requires.

You mean run a separate wire from arduino to the battery? I also have a 12v fridge that is unaffected by the spike. Is it because they have protection built in against such voltage fluctuations?

 

Yes, a separate wire because the current passing thru a wire's resistance generates a small voltage that affects the other things fed by that same wire. Mostly those are spikes in the negative direction, except for sometimes. Negative spikes can cause problems without doing damage, and that makes analysis more difficult.

 

You mean run a separate wire from arduino to the battery?

Also a separate ground line, since spikes due to the ground resistance can also upset the arduino.

 

Also a separate ground line, since spikes due to the ground resistance can also upset the arduino.

Thanks for the explanation. Would the alternator be causing spikes on the circuit too when the car turns on and off?

 

Would the alternator be causing spikes on the circuit too when the car turns on and off?

Not likely.
But the starter will if the circuit is powered when the starter is energized.

 

Not likely.
But the starter will if the circuit is powered when the starter is energized.

At least in some systems, if the voltage regulator starts adjusting the alternator field current while the engine is cranking, the alternator could deliver a higher voltage when the starter load is removed. The result was often called "LOad Dump Transient".
I think that problem has been addressed more recently, but certainly it was a concern with earlier alternator systems.