Diagonal rod length impact on Z dimension

Hello everyone,
I've always considered that the diagonal rod length only modifies the XY dimensions while the Z dimensions is impact with the steps/mm. While calibrating the diagonal rod lenght, I've notice that my Z dimensions is not accurate anymore.
Can someone confirm (or not) that the diagonal rod length only modifies the XY dimensions ?
NB : I want to have it as close as possible before using the M579 command.
Thank you !
Axel. 
what factor compensation did you use to calibrate the delta?

@AV13091 said in Diagonal rod length impact on Z dimension:
I've always considered that the diagonal rod length only modifies the XY dimensions while the Z dimensions is impact with the steps/mm.
That's correct.
While calibrating the diagonal rod lenght, I've notice that my Z dimensions is not accurate anymore.
I can't think of a reason for that. How far off is the Z measurement?

@Veti
Thanks for your answer,
That's the idea, i don't use any compensation tools, i'm targeting an exact value (as close as possible) before using any compensations. 
@AV13091
i am talking about
https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer
section Setting up the bed.g file
which factor of those are you using.
if you have not run the delta calibration, then you really have to. 
@dc42
Thanks for your answer,
it's not that huge (49.5mm instead of 50mm). What bothers me is that I had an exact 50mm before troubleshooting the diagonal rod.
I wasn't sure about its impact on the Z dimension as I always took for granted that the diagonal rod only affects the X and Y dimensions.Anyway, now that i'm sure about this, I will take care of the rest (my belts are quite old and could be too loose).
Again, thanks for your help.
Axel.

@Veti
Okay, I understand now, I'm using a classic S6 factor :).
I will for sure run a delta calibration but I would like to approach the real values first (effector's tilt, and other problems can give bad results).