Which impedance, low or high, from those offered by the Xiaomi S400 scale should I use?

[rubendediegomartinez]

Hello,
I have another question about the Xiaomi S400 scale. It's related to the two impedances it offers: "low," obtained at 50kHz, and "high," obtained at 250kHz.

My problem is that I don't know which one to use for this integration. I believe the "high" impedance is the correct one, since according to Xiaomi's information, it's used for "Measuring intracellular fluid and measuring cellular components of muscle, fat, bone, and organs," while the "low" impedance is used for "Measuring extracellular fluid and measuring water in blood and tissue fluids."

Can someone confirm this hypothesis? Thank you very much.

[dckiller51]

You need to use the lower impedance. Do you have any other information about the Xiaomi S400 besides the two impedances?

[rubendediegomartinez]

You need to use the lower impedance. Do you have any other information about the Xiaomi S400 besides the two impedances?

Unfortunately, I don't have any more information, although I'll try to find it: I bought an S400 ;-)
Your kind reply seems inconsistent with the purpose of the data Xiaomi provides, but I'll obviously take your advice.

Thank you very much for your work.

[dckiller51]

Can you give me your configuration: your weight, low impedance, high impedance, height, age? Then the expected results.
You can send it to me by email if you don't want to provide your data here.

[rubendediegomartinez]

I have no problem with sending you the data but your e-mail is? I hope you understand Spanish. BR El vie, 17 abr 2026 a las 11:06, dckiller51 ***@***.***>) escribió:

…

*dckiller51* left a comment (dckiller51/bodymiscale#343) <#343 (comment)> Can you give me your configuration: your weight, low impedance, high impedance, height, age? Then the expected results. You can send it to me by email if you don't want to provide your data here. — Reply to this email directly, view it on GitHub <#343 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ATYKMFSIQRYGYRWZIDYTR5T4WHXZPAVCNFSM6AAAAACXW2OBEOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DENRWG4YDMMZSG4> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[DJaeger]

@rubendediegomartinez
If you want to use it with a single impedance sensor you need to use the sensor with the numerically larger value which is the 50 kHz sensor.

You should read this section of the README for the S400 which describes the naming convention mismatch:
https://github.com/dckiller51/bodymiscale/blob/main/README_S400_UPGRADE.md#impedance-naming-convention-bluetooth-s400
or the corresponding comment in the code:
https://github.com/dckiller51/bodymiscale/blob/main/custom_components/bodymiscale/metrics/impedance.py#L75-L84

[dckiller51]

Thank you for these technical details.

To handle this, I have implemented a robust guard in BodyMiScale's code. Based on BIA physics, the 50 kHz impedance (low frequency) is always numerically greater than the 250 kHz one (high frequency). My code now uses max(low, high) for the 50 kHz value and min(low, high) for the 250 kHz value to guarantee correct results regardless of the naming.

However, to resolve this naming confusion at the source, I have opened an issue on the upstream library (xiaomi-ble): here.

This will ensure that the entities are correctly named (by frequency rather than numerical value) in future updates.