• 19.07.2025, 19:53
  • Register
  • Login

  • You are not logged in.

 

Thermistor calibration globally - including built-in sensors (High Flow Next)

Teasuti

Junior Member

Thermistor calibration globally - including built-in sensors (High Flow Next)

Tuesday, August 6th 2024, 12:57pm

Hi all,
Newbie here.
I see that Aqua Computer devices measure temperatures with a two digit precision. What's the best way to calibrate the probes so they are equal? I'd like to calculate absorbed and dissipated power, so the high precision is welcome, but the probes also should be calibrated to each other if I want meaningful data.
I'll have 5 external probes and an internal one in the High Flow Next (4 probes on an Octo, 1 external sensor on the flow sensor). For the absorbed power the setup is going to be like: High Flow Next -> VGA -> HF Next's external probe -> CPU -> Octo's external probe.
I plan to utilize the HF Next's power measurement feature with the automatic offset for the VGA, if this won't undermine the manual calculation on the CPU. The requirement would be that the two external probes wouldn't change their offset.
So my question is two fold:
1. How is the best to equalize the probes that all of them should give me the same reading in a controlled environment. What's the best way to create such a controlled environment with simple tools? Normally people use 0°C and 100°C water to calibrate industrial equipment, but surely that wouldn't work with Aqua Computer products, right? Can the thermistors be submerged in water? The very least I'd like them to be relatively equal to each other, even if they're not absolutely correct.
2. In the High Flow Next's logic, which sensor gets calibrated to the other? Or which one is the master or slave? Is the external sensor that follows the internal or the other way around (automatic offset)?
Thank you!

This post has been edited 2 times, last edit by "Teasuti" (Aug 6th 2024, 1:03pm)

sebastian

Administrator

Tuesday, August 6th 2024, 1:03pm

The easiest way is to adjust the offset of all sensors so that they all display the same values.
To do this, the flow rate should be as high as possible and the load on the PC should be minimal.

The adjustment should best be made at the preferred water temperature, e.g. at 30°C. This also compensates for any non-linear deviations between the sensors.

Teasuti

Junior Member

Tuesday, August 6th 2024, 1:07pm

Hi,
thanks for the reply!
I was thinking maybe I could calibrate them on the desk or something. Out of the computer, in a glass of water or mounted on something very heat conductive and environmentally insulated.
This method would leave out the High Flow Next, but I kinda hope that it would calibrate itself to its external probe I'd set up manually. Wet dreams?

sebastian

Administrator

Tuesday, August 6th 2024, 1:12pm

Alternatively, you can put all the sensors in a box, leave it for 1 hour and then adjust the offsets.
Ideally, the sensors should be isolated from all external influences.
However, the sensors should then only be used at the input at which they have been calibrated.

Teasuti

Junior Member

Tuesday, August 6th 2024, 1:12pm

Also, would the wire length affect the readings on the 10k probes much? I was thinking that I'd shorten them in their final arrangement.
Thank you!

sebastian

Administrator

Tuesday, August 6th 2024, 1:14pm

Quoted from "Teasuti"

wire length affect the readings on the 10k probes much?

minimal, but I would adjust it to the final configuration.

Teasuti

Junior Member

Tuesday, August 6th 2024, 1:17pm

Oh, I like the box idea! I could perhaps use something like an oven to create a controlled environment? At a low temperature of course, such as 30°C.

This only leaves me with the question about the logic of the automatic offset in the HF Next. Which sensor is the slave? The internal or external?

sebastian

Administrator

Tuesday, August 6th 2024, 1:22pm

Quoted from "Teasuti"

At a low temperature of course, such as 30°C.
maybe a head bed in a 3d printer.

but i would then want to have an additional mass for heat distribution.
i don't think it will be accurate. normal ovens have an accuracy of +-5 °c at best.

Teasuti

Junior Member

Tuesday, August 6th 2024, 1:25pm

It doesn't have to be accurate in absolute measurements, they just need to be equal to each other for relative measurements. I will only ever use them for delta-T calculations.
Do you know about the master/slave configuration in the HF Next's automatic offset? ^^

Similar threads