Thanks Micha a lot!!!
I need to do simulation of 2D free network. All points are datum points. I propose use of instrument 3" accuracy for directions, and 2+2ppm for distances. Approximate coordinates are:

S1 5442.34 3926.15
S2 5503.94 3951.27
S3 5563.70 3920.66
S4 5513.12 3899.66
S5 5452.25 3880.17

Input file for observations - directions is:
S1 S2 0.00000
S1 S5 0.00000
S2 S1 0.00000
S2 S4 0.00000
S2 S3 0.00000
S3 S2 0.00000
S3 S4 0.00000
S4 S5 0.00000
S4 S2 0.00000
S4 S3 0.00000
S5 S1 0.00000
S5 S4 0.00000

Input file for observations - distances is:
S1 S2 0.0000
S1 S5 0.0000
S2 S1 0.0000
S2 S4 0.0000
S2 S3 0.0000
S3 S2 0.0000
S3 S4 0.0000
S4 S5 0.0000
S4 S2 0.0000
S4 S3 0.0000
S5 S1 0.0000
S5 S4 0.0000

Results that i get in Excel are:
PointID A [mm] B [mm] angle myi mxi
S1 0.94 0.77 344.3853168 0.41 0.34
S2 0.74 0.70 8.394335191 0.32 0.31
S3 1.12 0.47 352.0845818 0.48 0.21
S4 0.72 0.58 330.685026 0.30 0.27
S5 1.00 0.75 340.6293756 0.43 0.34

A and B and angle are parametars od ellipses and myi and mxi are a-posteriori uncertainty for Y and X component.
I also get 0 results for a-posteriori uncertainty for directions, so I dont know what I missed.
I will attach via google drive JAG3D project and video what I did, can you give me an e-mail for sharing these files?

Thanks in advance!
Zaklina