You can use the Syn control to set the driver's notion of its position to an arbitrary value. So if you know that you finish up one session at position 450, and don't touch the focus, next time you fire up your system you can Sync to 450 and go from there.

Those of you who suspect a defective Waveshare HAT might want to try using the Motor B controller.
If I am not mistaken, to select which motor driver is used edit the file wmh_focuser.cpp
Edit line 91

@ Ihoujin

The defect that I found on motor controller A did exist also on controller B, I checked both. The faulty board had a board revision 2.1, the new board did not show a revision at all, so I suspect a design change or a faulty production lot ...

@ Kevin,

I was out in the garden this night and tried the waveshare focuser: it worked like a charm! Great stuff!

Excellent, glad to hear it!
 

I thought I should share what I have done. So here is one of the focusers I have created based on the Waveshare Stepper HAT.

Very nice!  What stepper did you use?  I ended up using a larger stepper (NEMA-17) with a belt/loop under tension.  I really like the smaller stepper and worm gear arrangement you used.  And for this purpose, a larger gear ratio seems better since it enables finer control and decreases the required torque to move the focuser.

It's a tiny stepper motor I found on Aliexpress with a built in 50:1 gear box. The worm ratio is 120:1. Together it is a very high gear reduction. The position resolution is insane. A command for 100 steps barely has a noticeable effect on a bahtinov pattern. I had to use an initial step size of around 800 for the autofocus to get a good V curve.

I am thinking of selling my design as a kit.

That is supercool! 
I had tried a similar design initially, but there was too much torque,  I could not make it work.  I used a different design instead, but it's more bulky. 

 

Yours is much more elegant.
Can you post the instructions and the STL files for building it?

Jo
 

This is the motor I used. It is a minor miracle it actually has enough torque. But it manages with the very high gear reduction. It can get hot if it is run too much. And there was a one time issue when I left power to the HAT on while the Pi was shutdown. Thus my eNABL pin pull-up resistor bodge to prevent another meltdown.
www.aliexpress.com/item/32848283821.html...1.0.0.301c4c4dVCiy0M

Thanks for the link.  It looks like I was on the right track with this eBay listing I found

ebay.us/mzsqFl

Thanks, Andrew!  

I was actually wondering whether you would be willing to share your 3D files for the worm gear.  Designing a worm gear in Blender is complex and time consuming, having the final STL file would go a long way.
I don't think I will replace the focuser I already have on my WhiteCat, it is working perfectly, just as well as yours is, but I would like to adopt this to some manual Nikkor lenses that I also use for wide-field imaging.  Torque does not have to be so high for those, they move more easily.  The ring around the lens will have a different diameter, though, but I can scale the worm gear accordingly in Slic3r.

Thanks for considering,

Jo