Dosing oddity after firmware update

[BeanAnimal]

Hello -

I have a SW replacement setup with doser 2.2 Maxi to run several times a day.

It ran for months as expected. with 3 or 4 doses day at varied volumes (based on KH director and ReefBot lab usage)
I updated firmware a day ago on the dosers and P4 to latest versions.

I also adjusted the doses down from 200ml and 300ml to two doses per day at 125ml (reefbot offline).

Now I am getting oddball dosage amounts again. Notice that instead of 125 I got 124.8 and then 125.2
I am (again) very concerned here that I can't trust the dosing schedule or actual throughput or reporting- especially for critical doses like Alk, Ca, Mag, KH director, etc.

See image below - prior dosages were proper, but the new 125ml dosage is not.

Why is there an adjustment being applied????

Settings

@Gaël @Vinny@GHLUSA can either of you offer some insight?

I think this is a pretty serious issue that keeps cropping up in one form or another,

Math or rounding issue?
Specification is .1ml

240ml a minute is 4ml per second.
31 seconds is exactly 124ml
32 seconds is exactly 128ml
31.25 seconds is exactly 125ml
.025 seconds is exactly .1ml

There is no oddball rounding issue here unless the the .025 seconds is being rounded to .03 seconds and that changes the advertised or effective resolution from .1ml to .12ml

It then follows that
125ml desired dose can't be resolved to .12ml resolution and not only that but it would appear that error is further compound by rounding the .12ml to .2ml !!

But - it gets worse, because we are now undershooting or overshooting by .2ml which is a .4ml total range for a 125ml desired dose... nowhere close to .1ml resolution. This appears to be rather significant!


Thus 124.8ml or 125.2ml being reported (or dosed and reported?)

The question then becomes is the rounding issue a charting/reporting issue only and the weird "adjustment" just on paper and not really happening (I.E. the reporting math and the stepper math are not the same) and the dose WAS actually 125ml ?

Or
Are the dosages being mangled by the rounding and the P4 trying to adjust the following dose to compensate?

 

[Gaël]

Hello,
As already mentioned, the official support channel is the ticket system.
It's really not ideal to ask ours developers outside the ticket system to come here and analyze your issues.
It's also very time-consuming for me to copy your posts into the ticket system, and multiplying support channels is a source of error.
Gaël

 

[BeanAnimal]

Hello,
As already mentioned, the official support channel is the ticket system.
It's really not ideal to ask ours developers outside the ticket system to come here and analyze your issues.
It's also very time-consuming for me to copy your posts into the ticket system, and multiplying support channels is a source of error.
Gaël

I am always grateful for your responses and the support that you Vinny and others have provided since I bought into this ecosystem and can certainly officially open a ticket -

I do adamantly believe that public transparent conversation benefits everyone involved. It keeps vendors honest and on their toes and allows customers to provide open feedback and collaboration on issues that they may be facing as well or even otherwise unaware of (and affected by) or have workarounds for or input on.

I am not (by any means) posting in an attempt to tarnish GHL. At the same time I don't feel comfortable with everything being hidden behind a closed ticketing system (for the reasons mentioned above). For that reason, I started a public conversation and tagged you folks.

I will get that ticket opened.

Thanks again for the response!

 

[Matthias Gross]

I think when you need more accuracy (0.1ml accuracy) then you should reduce the pump speed.
Switching on/off (ramping the motor) always takes some time and may vary a tiny bit, but when you use a slower speed than the ratio between ramping and constant run gets better.

 

[BeanAnimal]

I think when you need more accuracy (0.1ml accuracy) then you should reduce the pump speed.
Switching on/off (ramping the motor) always takes some time and may vary a tiny bit, but when you use a slower speed than the ratio between ramping and constant run gets better.

Hi Matthias

Thank you for the response.

- In this case, accuracy doesn’t matter at all.

The results are odd and not expected though.

I am just more concerned about what is going on with the actual dose amounts and reporting As it relates to any dose, not this particular schedule and volume.

I think I found the answer as it relates to rounding, but not sure if that affects other areas or doses too. I would like to understand what is happening.

 

[Matthias Gross]

I think "adjustment" may be here a misleading comment in the app, we will have a look into that.
The bottomline is that we see here a deviation to the desired amount of 0.16%.
If you want to improve that you need to reduce the pump speed.

We didn't change anything in this part of the software since months.