Retrofitting a 1986 Maho MH400E

Tittle changed back to original on request of original poster.
Please do not change the subject tittle when replying.
Thank you.

Thanks Tommy,

I have worked through troubleshooting, and have the machine up and running again.

The EXE's are not powered by the Indramat power rails. The Indramat received 2x 90VAC (X and Y) and 2x 140VAC (Z) from a transformer straight off the mains. The X axis EXE still uses an original lane of the old three axis EXE board, and thus needs both 5VDC and 12VDC supplies, which are powered downstream of the main isolation transformer.  X and Y are on 601D EXE's which only need 5VDC, on the same PSU as the Z.

It seems like a combination of issues.
1/ I had an addition relay so LinuxCNC can also control the Z axis brake, which in the original set up was only controlled by the Indramat. The consequence is that the motor was fighting the brake briefly on every start up. Not sure if that has caused further issues.
2/ Errors in setting Ferror verse Ferrormin, Leaving Ferrormin too tight.

Axis    fError.    fErrorMin
X.           0.5           0.2
Y            0.5           0.2
Z             20             1

The 20 was obviously an oversight which I forgot to set after finishing pid tuning.
Question: What would be a typical value for fError and fErrorMin for a machine like set with 2400mm/min max rapids? Is 0.2 and o.5 really that tight?

3/ Not adding any intergral in the pid tuning, once I had P, FF1 and FF2 set. There is a constant offset in steady state motion.

4/ Suspect that using the 220 weight way oil recommended in the MAHO manual might be too thick in winter.

I have removed that 19K2 relay, removing linuxCNC from the brake control and returning it solely to the Indramat.
I have opened up the fError and fErrorMin values, and am now working to find sensible values.
Just playing around to find the right value of I for each axis.
Cheers,
Mark


 

The Maho is back up and running again, but I still dont really know what the FError and MinFError "should" be for such a machine.


Thanks for your help.

You could use Halscope to see what the F-error actually is, then set the limit to some arbitrary factor of that. It wants to be high enough not to trip when there isn't a problem, but to trip if something actually is stuck.

Greetings Mark!
First of all, excellent work on the retrofit - this thread will end up being a literal goldmine for me. 
Last week I also purchased a 1986 MH400E and the goal is to convert the machine to LinuxCNC in the coming few months. 

Somewhere in this thread I saw that you have a 'collection of notes' for your conversion - is there any way to get these from you? I can't seem to find an email-adress available on your YT Channel as previously stated. 

I haven't quite had the time to work through all 180 pages, but I do have a quick question: 
How is the gear-changing generally handled? Do you have to command an exact speed value, or will LinuxCNC pick the correct gear for whatever arbitrary speed you command? 
Also, will the gear-change happen automatically whenever a speed is commanded, or is there specific codes that need to be ran (I would love to use the F360 CAM and I wonder if linuxcnc will handle selecting the speed automatically, or if I need to fiddle around with the post-processor?) 

Thanks and very best regards from Germany
Daniel

 

Hi Daniel,
Shame this forum has no DM function. I think you need to be logged into YT to see the email address on the main page. Otherwise I have the same user name on Zerspannungbude, which has a DM function.

I put all I learnt about the gearbox here. 
github.com/jin-eld/mh400e-linuxcnc/wiki
This Github has the Comp my mate wrote to control it.

Yes, the comp rounds the commanded speed to the nearest lower gearbox ratio. It reacts automatic to the "S" code in gcode. Of course, speed change is inhibited with the spindle turning.

I added a G4 P5 pause to my post processor at every speed change, as it can take that long for the gearbox to finish changing ratio.
Cheers,
Mark