LCD Screen Random Characters

Hey, so my problem is that I get random characters from the LCD Screen when my water pumps starts.
I tried to add already a 10nF capacitor in parallel with a diode(1N4007) to the water pump rails.
The relay module has already a flyback diode...
I've tried to isolate already the pump somehow and the wires but still it interferes with my hardware.
I have a separate 5V power supply for the relay and the pump.
The problem appears only when the pump starts.
Any advice would help!
KiCad diagram on how I wired the pump:


What I use:

Try 100nF instead of 10nF.

Try a large cap like 1000uF in addition to the 100nF.

Try using MOSFET instead of relay, like IRLZ44

What is the power supply?

HW626 5V step up connected to a 3.7V Li-Ion battery.
Tomorrow I will try to add a larger cap. Funny part is that it used to work before with this capacitor but I rewired everything and now this problem appears again.
I've seen somewhere that it will help if I add a capacitor around LCD Screen 5V and GND rails, but I didn't test it yet.

Some details from AliExpress:


Shows caps on input and output of the module.

What is the coil/winding resistance of the pump?

Good question to be honest, I calculated 0.25 ohms to be honest.

If the pump winding resistance was 0.25 ohms, the current at startup would be 20 amps, and probably 4 amps when running. That's 10x to 40x the max current your HW626 can supply. So I doubt that's correct. How did you calculate it?

R=V/I, I used Ohm Law, so R=4.5V measured/0.2 amps. It gives me 22.5 ohms. I'm sorry if I calculate them wrong.

Ah yes, the pump datasheet says 220mA, I missed that because it was buried in the description!

But that is the normal operating current. At startup, it will be ~5x higher, just for a fraction of a second. That's more than your HW626 can provide, so some large caps will definitely be needed.

On top of the 220mA for the pump, your relay will probably draw 60~80mA. Using a MOSFET instead would require almost zero current compared to a relay.

So, I should try by adding first a 100nF capacitor, would that be enough?
The relays draws about 70mA, I know I should change it with a MOSFET, but first I want to try adding a bigger capacitor and see how it works.
And If I can't get such a bigger capacitor, can I connect 2 in parallel?

We don't know exactly what is causing the problem with your LCD. It could be a voltage drop when the pump starts. It would be inference from the relay contacts or the motor in the pump. It could be both.

Increasing the 10nF to 100uF should help with the interference, if that's what the problem is, but it won't do much to help with the voltage drop. That's what the much bigger caps should help with. Yes, you can connect them in parallel. Don't forget that the large caps will be electrolytic type, and these must be connected with the correct polarity, or they will be damaged, and may even explode!

1 Like

Thank you for your advices! I will try to find a very big capacitor, and connect it. Tomorrow I will comeback with updates and see how it went.

every coils in constant current circuit must be blocked by a diode.

1 Like

Connect it on the +5V side of the relay (COM) to help with inrush current.

1 Like

I should add a diode somewhere else then?

I will try this too, I can get some capacitors today to try that. Thank you!

Additional informations if it helps:
This is how I wired the supply for the pump and relay:(Ignore that switch, I rewired it)
The 5V and GND goes to a breadboard, from where I power the water pump. I tried it without the breadboard, but makes no difference, that's why I'm using it.

Fritzing

KiCAD

Photo of capacitor and diode:

relay is a coil

So I should add a diode to the relay?

better is.

Update, I'm using a 3300 uf capacitor, and problem dissappear, only appears 1/10 times.
Tomorrow I will add a 10k one and make the problem go away for good.
Thanks for all the advices!