Rain sensor reports absurd numbers

[Liniik]

When I was testing the rain sensor before mounting it on the roof, values seemed fine. But now when we have rain, it's reporting apocalyptic numbers like 80 mm/min. The only change to the config from you is I put pulldown instead of pullup resistor because for whatever reason it needed that - without pulldown it was behaving like no resistor at all, reporting erroneous values and stuff, wiring should be the same. Do you have any idea where could be the problem? Did you have this sort of troubles?
Here's the config:

- platform: pulse_counter
    pin:
      # Don't forget to add a pulling resistor, see README
      number: GPIO33
      mode: INPUT_PULLDOWN
    unit_of_measurement: 'mm'
    name: "uhrn srazek"
    icon: 'mdi:weather-rainy'
    id: rain_gauge
    internal: true
    count_mode:
      rising_edge: DISABLE
      falling_edge: INCREMENT
    internal_filter: 13us
    update_interval: 60s
    filters:
      # Each 0.011" (0.2794mm) of rain causes one momentary contact closure
      - multiply: 0.2794
    accuracy_decimals: 4

[geejayem99]

How stable is the mounting? I do get some occasional false readings in windy weather as the pole it is mounted on is a bit flexible. The wind buffeting it around can cause the mechanism to flip. Maybe try increasing the debounce filter from 13us to something higher. I would think 15ms would be fine as it normally wouldn't flip that rapidly even in high rain conditions

[Liniik]

@geejayem99 Mounting is very stable. Mine aren't occasional false readings, mine are literally detection of tens of switches per minute which I don't think are really happening. I played with some intervals in the filter but then during literal downpour it registered maybe one or two switches... It looks to me then like when I change the filter to use_pcnt: false and any value, the detection is broken completely. When the weather improves, I'll have to check it on the roof. And maybe also consult esphome folks.

[jjiik]

can you rule out there is a problem of glitches when toggling, causing false triggers?

[Liniik]

@jjiik I definitely will once the weather gets better (hopefully two days at most) so I can climb on the roof. I'll then check the contact there and play with it a bit

[Liniik]

So I checked the reed switch on the roof - I can confirm there were no glitches when toggling, the bucket stayed on one side and the numbers went up like crazy (e.g. 200 switches per minute). No matter what I tried (pullup/pulldown, trying different filter values), I couldn't get correct numbers. I checked the reed switch directly with multimeter and separate magnet and it looked like it's switching correctly... I have it connected with the wind part of the station on one board which is then connected to the esp. Wind is working fine. So the only two things could be wrong is the hardware part or connections on the board.. Or do you have any other idea?