Microcontroller resets when activing the relay for boiler

Hello, I made a revision of a board that used for controlling a boiler via a relay. There is an RF Reciever module on it and purpose of the revision was change the RF Reciever module. So I have not changed anything else(power layer, relay circuit, microcontroller decoupling etc). Before the revision board was working fine however now We are dealing with a problem for some boiler models. When the board received the predefined command Relay make a short circuit for COM and NO pins so boiler starts working. At this moment not for all boilers but for some of them the board starts to reset itself. I measured the voltage and current get through from relay and it is 24V 8mA. Without any load, the board works fine. The problem occurs just when the relay connected to the boiler. As I said the previous board works fine for all boilers. I am using HJR-3FF-S-Z-4 relay. You can find the datasheet from here I also share the schematic of relay circuit and layout. Schematics are same. However, there is a little bit difference in layouts. For the previous board Ground polygon does not contain D4 diode but in my new design I don't know why I did that but Ground polygon contains the D4. Does it make this much difference or should I care something else?
Thanks a lot for reading.

try scraping away the trace and see if it makes a difference if that's the only change. Is it possible the diode is reversed? Just trying to play what if....

Hi,
Have you tried a 0.1uF capacitor from the NO to COM of the relay to act as a suppressor?

Tom... :slight_smile:

TomGeorge:
Hi,
Have you tried a 0.1uF capacitor from the NO to COM of the relay to act as a suppressor?

Tom... :slight_smile:

Thanks a lot for your answer. No, I have not tried it yet but I am gonna try it for sure. Capacitor should be ceramic or electrolytic ?

Hi,
Ceramic or poly, doesn't matter. Don't use an electrolytic, to high an ESR.

Tom...

But why would it change from version to version

wolframore:
try scraping away the trace and see if it makes a difference if that's the only change. Is it possible the diode is reversed? Just trying to play what if....

Thanks a lot for your answer. I will scrap there. I checked the diode and it placed correctly.

wolframore:
But why would it change from version to version

I was thinking the same thing. The previous board was working fine with all boilers so suppressor might prevent this. However, most likely the clearance of 5V net is the reason for this problem. It is quite close to ground plane and there are two sharp corners there. Because it is the only different thing. There are 4000 products in the market now :frowning: .

that's a lot of scraping :smiley: if that's what's it... did they do an electrical test on your board before delivery???

wolframore:
that's a lot of scraping :smiley: if that's what's it... did they do an electrical test on your board before delivery???

Petty expensive. But a good insurance policy. Our board house calls that the "flying monkey test".

Paul