Fehlemann Picomax 100 Retrofit

 
last the 4th axis is a bit more trouble, I needed a while to figure out how to disassemble it:

here I transferred the hole pattern from the mounting plate the to new spacer:

and the new servo can be mounted:

but here it ended while widening the bore of the new coupler to fit the new servo shaft I scrapped the part:

with this I decided to stop for now and wat of a new coupler to arrive. When It's here I will continue and start ti figure out How I will route the wired for the new servos. I don't think I will us the same paths as the old ones, cars the connectors won't fit.


 

With the new shaft coupling arrives, I was able to fit it to the A-Axis:

Next is to figure out how to protect servo an brake from chips, the old housing won't fit anymore.
Maybe I'll simply print something and cover it with Aluminium if needed.

I rerouted the Z-Axis wire, cause there was no way I could route it through the old duct ways, so I made a small cut in the casting and printed a cable guide. Maybe I' make a cover in the Machinehead later to protect the cables from a ripping belt.

Last was some cleaning before preparing to route the X-, Y- and A-Axis cable through the machine base:

And I wonder if I should use the encoder signals from the servos. Having glas scales I didn't need them.
But maybe they give me a finer resolution 0.0016 mm vs 0.005 mm for X and Y, for A I need to use the new servo anyway and Z I actually don't know the difference, cause I don't know the gear reduction.
 

Lets find out if someone is reading here...

I've got the servo drives setup an did some first test. The encoders will give me a better resolution than the old glas scales. so I connected the encoder output from the drives to the mesa cards.
So far so good.

now I've got a problem, the first drive is just oscillating when controlled by linuxcnc
the drives work with +- 10V

My hal setting:

setp hm2_7i97.0.pwmgen.00.output-type 2 #PWM up on pin0, #PWM down on pin1
setp hm2_7i97.0.pwmgen.00.offset-mode 0 # offset mode so 50% = 0
setp hm2_7i97.0.pwmgen.00.scale [JOINT_0]OUTPUT_SCALE

# PID tuning params
DEADBAND = 0.000015
P = 100
I = 0
D = 0
FF0 = 0
FF1 = 1
FF2 = 0
BIAS = 0

now I disconnected the motor and when I turn the the spindle, can look at the output:

pwmgen.00.value -50 -> 50mv on drive

pwmgen.00.value -100 -> 9750mv on drive

pwmgen.00.value 100 -> -9750mv on drive

is this OK / normal O do I have something misconfigured here?

Your PWM setup is not correct for the 7I97, these are the correct settings:

# ---PWM Generator signals/setup---
setp   hm2_7i97.0.pwmgen.00.output-type 1 #PWM pin0
setp   hm2_7i97.0.pwmgen.00.offset-mode 1 # offset mode so 50% = 0

 

Thanks I‘ll give it a try.
Made this Hal a while ago, and now not sure if I modified my lathe file or used another base to start.

Scale is not finally set, cause I don’t know it yet, after I switched from glasscales to encoder feedback.

But wrong direction of scales usually results in a runaway axis…

I measured all the axis and corrected all input and output scale parameters, so they should be right.
and PCW was right... All my other hal files use the same settings he mentioned, so I adjusted them.
but Ist didn't fix my problem in total.
It's still oscillating, but I can jog now.

Hi, I know this tuning tutorial and used it before, my problem is.
My servo system is oscillating right away.
Even with 0.00001 Deadband and P at 10.

When using a potentiometer for as analog input the servo runs smooth ans stable.

Sot it something that comes in interaction with linuxcnc.
Maybe the PID from the drive and linuxcnc don't work well together.

The oscillation range is +- 40 increments,
but there is something else odd,
the DRO is running at a different scale than the encoder position.

 

Are the encoders rotary or linear encoders?
Is the analog out scale set for full speed (in machine units per second) at +10V?
(if not, the units for FF1 and P are unknown)

I would get the encoder and analog scaling right,
FF1 right and then worry about fine tuning