If you have a electronic balance then simply weighing will do.

D2O is 1/18 or 5% heavier than water. If you have 1L of water it weighs 1000g. 1% of D2O will add another 0.2 g. Most scales can measure 1mg precision if im not wrong.

Heres a basic scale:

https://www.fishersci.com/shop/products/mettler-toledo-f1ac-basic-weight-sets-range-1mg-1kg-without-calibration-certificate/01911700

If you want ratio of DHO then the same can also be done, you just need to solve a few simultaneous equations:

a(16)+b(17)+c(18) = weight

a+b+c = 1L

GC-MS can be used for this, see our manuscript on this for more details.

http://doi.org/10.1016/j.jasms.2006.02.008

I have always wanted to build a specialized nuclear magnetic resonance setup to do this.  The hydrogen and deuterium signals are quite distinct and well studied  With calibration, I always expected one could make the device quite small and inexpensive.  Low fields are fine, just over sample.  Kind of an elegant approach.  A bit overkill, but that is the fun of it.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2735574/ --  Solvent Signal is a NMR Concentration Reference

https://en.wikipedia.org/wiki/Gyromagnetic_ratio

https://en.wikipedia.org/wiki/Proton_magnetometer