Did you also find that the model is not fully defined by the drawing?
I often wonder if instructors assign incomplete "drawings" to test their student's ability to improvise. Or if the instructors are incapable of producing a complete drawing. I suspect the latter.
On the plus side it makes it easier to make the model. I stopped caring if my guesses and assumptions are correct. I put in some effort, but if the drawing is wrong or incomplete the model will be the same.
Simple part, if done is the correct sequence. I used the multi-view drawing in the referenced question - 2 missing dimensions.
I just noticed my mistake with the square left side that should be round! An easy fix.
Was that the significant difference?
With that change, I only have 1 missing dimension.
Inventor. Some dimensions and details were missing. Had to guess at some of the detail.
 Updated per Steen Winther's findings. (No dimensions missing.)
The fix is only easy if you don't use a fillet. Under the conical boss is some kind of undefined surface, possibly a spherical cap.
The first time I made this model, I used a weird approach that worked perfectly. The second time, I used a traditional extruded outline technique + cuts and fillets that first gave the wrong geometry and then numerous errors due to knife edges and sliver faces. Fun & Games.
I assumed a spherical cap on the top and bottom.
Yes, I think you're correct. I assumed that the curved cylindrical surface transitioned into a flat face with missing specs, but it is probably the large vertical hole that gives this illusion?!? Must try again :-)
The 17.9 mm dimension references an edge that is only defined after the 8.2 mm cylindrical cut is made, making the model even more tricky.
But with the assumption that the left side (cut by A-A) is a cylinder with spherical caps, the model is fully defined by the drawing.
Actually, if you look at my drawing, the 8mm radius on the bottom is undefined on the original. I assumed 8mm but it could be anything.
If you assume that the left side (cut by A-A) is a cylinder with spherical caps (disregarding the conical boss), then the radius is exactly 5.13 mm, and the drawing is fully defined. :-)
I see where you got that now. I concur.