ORP Dosing Delay on Startup Possible?

[FlaMike]

Is it possible to set an ORP dosing delay when the filter pump starts like we can for the pH dosing? I changed the delay for pH to 5 minutes because it appears to take a little while for the probe readings to stabilize when the filter pump starts each morning. The filter pump runs at 1750 RPM which is about 17 GPM.

While the pH delay was counting down I noticed the salt cell turned on for a second or two and then turned off. This cycle repeated multiple times until the pH delay expired and the acid pump turned on (thereby turning off the salt cell until the pH dose completed). By the time the relatively brief pH dose completed, the ORP reading had stabilized & no additional chlorine was needed.

I observed this behavior for the first time this morning (I changed the pH delay yesterday). I did not have a capture running and because enough pool water has filled the flow cell, I cannot do so before tomorrow. That being said, I will try to capture the behavior tomorrow if the overnight probe readings drop sufficiently. Of course, if there is a way to delay the initial chlorine dose, I'll use that option.

Thank you!

[tagyoureit]

A few thoughts.

First, the pH and ORP probes are designed to stay wet all the time. If your flow cell is draining when the pool is off, I would add one (or maybe two) one way valves so the probes are never exposed to air. You may still get a variance at startup but the introduction of air bubbles into the cell at startup is probably as much a culprit, if not more, than the actual water readings.

Second, in REM you can smooth out the readings over n number of samplings which would make the readings less erratic.

Finally, if the cell turns on or off for a few seconds here or there it isn't really much of a concern. The start delays are more to make sure that flow has been established and we aren't pooling chemicals. You can set this with pH and chlorine dispensers because they don't know about the flow. The salt cells all have their own flow sensor so it isn't a concern as much.

[tagyoureit]

A bit of a spike/drop at the beginning (especially <+/-40) could easily be accounted for by stagnant water in the pipes sitting overnight, for example.

Other than your utility company trying to screw with you, I don't see anything concerning here. njsPC will adjust to any anomalous readings once they recover, and not that you have mostly normal readings there won't be any major impacts to the pool operation or chemical dosing.

You might want to put a delay in the HA startup (if possible), so that it can connect to njsPC after it's loaded up. Or you could use the npm run start:cached to load njsPC without recompiling the TypeScript code (you'll just need to re-compile it when you pull) which might help it load before HA. But again, I'm 👉 at the utility company for creating these issues.

[FlaMike]

The variation in probe values on start up do not concern me. I understand the water in the flow cell & pipes is not representative of the pool water when it's been stagnant for hours on end. In addition, the magnitude of the spike/dip is small compared to the "true" reading. Finally, a little extra acid or chlorine at start up isn't significant, I agree.

When I first built Nixie I did try to use the cached flag. However, I was not able to get it to work successfully with PM2. If I recall correctly, and it's been a while, the system just hung & never started. If you could point me to instructions on how to use the cached flag with PM2 correctly, I will be happy to make that modification. I must admit, I get tired waiting for the system to download & rebuild when I'm at the pad & do a hard reboot.

Ain't it fun when the power company or folks doing construction on your mainline mess with you?

Thanks!!!

[tagyoureit]

PM2 Wiki instructions

You would edit each app's startup:

    {
      "name": "njsPC",
      "script": "npm",
      "args": [
        "run", // new
        "start:cached" // changed
      ],

This matches each applications package.json startup scripts:

  "scripts": {
    "start": "npm run build && node dist/app.js",
    "start:cached": "node dist/app.js",
    "build": "tsc"
  },

Finally, you would run and save the new ecosystem.config.js per the initial instructions.

pm2 start ecosystem.config.js
pm2 save

[FlaMike]

Thank you for posting the changes to the startup scripts. In my earlier attempts to use the cache option I don't think I had the "run" line you commented as the new line--I only did the "startL cached" line. Probably why it didn't work.

It will likely be a couple of days until I can make the changes, but I will give them the old college try & be happy to report my success!

Thanks & have a super rest of the day!

[FlaMike]

OK, as it turned out I had a few minutes to play, so I made the cached suggestions you made to ecosystem.config.js file & went through the pm2 autostart setup process and it worked!!! My earlier attempt did not work for because I did not inlcude the "run", instruction line. Not sure if I just missed something in the documentation or it's missing in the pm2 Wiki, but all good now.

Thank you!!!!