@bar … if you mean “loose” connector, are you talking about those at the end of the cables? I’ve checked them all, again and again … there are all sitting proper in their sockets
1 Like
I say let’s go with @madgrizzle’s excellent plan for now. I will start thinking up reasons that measurement might be coming out wrong and we can explore them if using a manual measurement doesn’t work.
@bar … fair enough! Will give it a go now and will report 
1 Like
@madgrizzle @bar … one thing I forgot to explain/describe:
When I did the calibration, I extended 3 times 990mm, so a total of 2970mm, which I hardly could put on the right motor sprockets 3 o’clock position, as it was real tight already … retracting 2 times 10mm was almost tight and the measure function took only a fraction of a second … which is inline with my real motor disctance of around 2948mm …
I just got a call from a friend who’s dropping a LEICA laser meter in one hour to allow me a more accurat measure of the motor disctance … will try to make a small gadget to improve accuracy!!!
1 Like
@bar @madgrizzle … I’ve build a quick and dirty gadget to assist me in using the laser measure to get a proper result …
as already anticipated the distance of motors in my case is 2958mm …
this still leaves the question, why the measured motor distance is so much off center, than the reality! 
tomorrow I’m going to do the calibration entering the real distance in the options and skipping the chain measure procedure … and will do a small test cut to see accuracy! keep you posted …
1 Like
I did finalize the calibration using the laser measured distance of motors, but skiped the test cut pattern (yet but will do later/tomorrow| don’t want to get into trouble with my neighborhood) and moved sled to center of the waste board …
it’s “almost” center … I measured it out … it’s exactly 5 mm of center to the right and to the bottom!
It this result within the tollerance?
thank you for this information! So distance of motors (outside of the square edge) here is at 2991.5mm (measured with tape measure where once fixed with tape), substracting 40mm, 20mm for each motor, results in 2951.5mm as the distance of motors for the calibration. Correct?
The triangular calibration is a joke. I have reviewed the code, and it is not much better than a random number generator. The people who have had success have done so by repeating the calibration over and over; like a any random process, they eventually arrive at an acceptable
calibration by luck. You are better off skipping the calibration altogether. Just run with the measured parameters, and you will be better off.
2 Likes
@Joshua… You are not talking about the whole calibration procedure, but the “measure the chain lenght” part only! Right?
1 Like
I think that’s a bit of an extreme characterization.
I’d personally use the hand measured distance between motors over the machine measured. As for the rest, I think you are better off calibrating them than not. There is a much improved calibration process and very likely improved kinematics model (holey calibration) but it’s not incorporated into the current branch of ground control. You have to use someone else’s version or shift to webcontrol.
I’ve redone the measurement of the distance of motors, just to be sure!
Therefore, I’ve clambed a piece of wood to the outer edge of each motor…
… and used the laser meter to measure the distance between booth verticals.
I’ve redone it several times, and most of the measure resulted in 2992mm.
Furthermore I doublechecked the measure from outer edge of motor and center sprocket, which is indeed 20mm…
Hence, the distcance of motors in my case is 2952, which I have used for a full calibration incl. the test cut pattern.
We will now cut some small items to see accuracy…
1 Like
I am talking about the part where you perform the test cuts and click Calibrate. If it is better than a random number generator, it is not by much.
Truly, I am not trying to be negative. I have seen threads like this over and over. It is the same story repeating, and yet the Maslow community hasen’t recognized the root cause, the triangular calibration.
All the measurement steps are ok, but we could question their value of we prefer the hand measurements.
I’ve been looking forward to your next update. Were you able to get good results from your calibration after all your work with the distance measurements?
-Jeff
Hey Jeff, thanks for your follow up on this thread! Yes, indeed we got way better results after the “recalibration” … But there’s still inaccuracy in the left/right border area! That’s why we decided to do change our frame as recommended by MetalMaslow, which is under way!
There is an arc (arcing towards the middle of the vertical space) in the left/right that is to be expected, as I understand it, where the accuracy suffers. There is a script that will show you where to expect it.
I hope to compensate for it by cutting smaller boards that stay in the “sweet spot” of the accuracy area, but that changes the overall usefulness of the Maslow. It’s all trade offs. I don’t think you’ll get consistent accuracy over the entire 1220mm x 2440mm area. There is another arc (left to right) near the top center, I believe.
-Jeff
Which script you are talking about?
Boris, I’m afraid I can’t seem to find it at the moment. There have been various examples of the output posted in the forums, I think it came from VPlotter maybe. @dlang can you point us in the right direction for the script to generate the areas on the sides and top/bottom where accuracy may suffer?
-Jeff
Do you mean this:
Not that, there is a graphical view of the work area that shows where accuracy will suffer that I believe is generated from the calibration data on your specific machine.