Since you are here you probably are interested somehow in astronomy. And you probably already are aware of terms like Doppler effect or radial velocity. In ancient times, when I started to read about astronomy, spectroscopy was an area reserved for professional astronomers only. But things changed. Currently astronomy amateurs use sensitive CCD/CMOS sensors and good quality optics with computer controlled mounts. A few models of commercial spectroscopes are available for amateurs (DADOS, Alpy). Doing ATM spectrometer is also an option, and if you have access to 3D printer, then LowSpec device may be your choice. Once you have such device, then amateur measurements of star radial velocity are now among the things you can perform.
The heart of LowSpec spectroscope is reflective grating. Depending on the grating density the effective resolution varies. For 300 l/mm grating LowsSpec provides resolution about 600. But with 1800 l/mm grating you may already reach for resolution 5000-6000. Assuming, that we can determine spectrum line position with 1/10 resolution accuracy, then for 300 l/mm grating you may expect to measure radial velocity with accuracy about 50 km/s. And for 1800 l/mm grating this accuracy may be about 5 km/s. But such measurements requires much care to achieve reliable results. More information on this is present in the literature, for example in Practical Aspects of Astronomical Spectroscopy by Richard Walker.
Since my LowSpec currently has 600 l/mm grating for the first tests I have chosen high velocity stars. After some research I have found a good candidate: BD+37 1458 star (242 km/s). And in the neighbourhood there is much slower HD43094 star (38 km/s). These stars are located in the edge of Milky Way in constellation of Auriga.
Here are the captured spectra of both stars. Plus Relco SC480 starter bulb spectrum (light grey plot). When you take a closer look to the hydrogen alpha line at 6563Å, then you may notice that they are shifted a little bit. This is due to the Doppler effect, and this offset is the function of star radial velocity. But in the same time you may notice, that deep absorption line at about 6880Å is the same for both plots. Because this is Fraunhoffer B line from atmospheric oxygen – it does not come from any of these stars.
Now we can zoom the spectra and measure exact position of hydrogen alpha lines in both plots. It is pretty easy task in BASSProject software, but the projects needs to be calibrated with calibration lamp (or reference star in the same field of view, because moving telescope position affects spectrum line position due to bending or tilting).
Grey plot is from Relco SC480 reference spectrum. Green plot is more shifted to red, so it comes from faster BD star. I have made several measurements of each line and averaged them. Then basing on the measure offset I calculated actual radial velocity in km/s. Last step was applying the velocity correction caused by Earth orbital and rotational motions.
Actual values in SIMBAD database are 242 and 38 km/s. So the error is pretty large, but for 600 l/mm grating that gives resolution about 1400 it is expected. You may imagine, that QHY163M camera pixel size is 3.8um, and if the camera will tilt or bend for just one pixel between measurements, it will result with the error of 30 km/s (!!). Also signal to noise ratio is quite important, because in noisy spectrum line position measurement accuracy is lower.
Nevertheless I am pretty happy about the achieved result. 1800 l/mm grating is already ordered, so I expect to get much better accuracy then in measurements of radial velocity.
You will find more of LowSpec captured spectra at this dedicated page.