8i20 setup current limits etc.

Awesome progress today!
z and x axis both working,  ready to be connected to the slideways. The "iq" settings work flawlessly. 
I found this morning that a small amount of water must have gotten into the bottom of the x axis servo motor during its ordeal. The brass flecks in the brake had corroded, causing it to lock to the plates inside it. A quick disassemble and cleanup fixed it and revealed that it was indeed 24vdc, not 90vdc. It's printed on the plate that you can read when you separate it from the motor.
going to reassemble the spindle on Thursday,  - kluberlube and very expensive new bearings. - high precision anular contact. 5 of them 130mm od. And I need to hook up the turret to ladder and the limit switches.

Just borrowing the topic.. Issue with the settings of the 8i20. How all of them should be set if i want the maximum amps out of the thing?

I currently have original fanuc servo (10s, 8i20 should be plenty for it) on the Z-axis, and i can't seem to get enough power from it with the 8i20. Have been fiddling couple of hours with all the settings, searched the forum but haven't found any real answers for my questions. Best result I got was that the servo lifted the Z-axis just enough so that i could have removed the wood block from under it so it would drop from its own weight..

Some of the setting i have tried (nvmaxcurrent also tried with 3000, and tried to set the current maxlim 30000 etc and plenty of other combos without great success). Just would like to know, what settings for max 30a output, and maybe how to restrict that if necessary (should adjust the amps only on current-maxlim/minlim/scalemax?)

INI 

[JOINT_2]
TYPE = LINEAR
HOME = 0.0
FERROR = 10.0
MIN_FERROR = 8.0
MAX_VELOCITY = 25.0
MAX_ACCELERATION = 750.0
P = 160
I = 20
D = 10
FF0 = 0
FF1 = 0
FF2 = 0.01
BIAS = 0
DEADBAND = 0.000015
MAX_OUTPUT = 3000
ENCODER_SCALE = -1000
OUTPUT_SCALE = 3276700
OUTPUT_MIN_LIMIT = -3276700
OUTPUT_MAX_LIMIT = 3276700


 HAL
 setp hm2_7i97.0.8i20.0.3.nvmaxcurrent 30000 
setp hm2_7i97.0.8i20.0.1.current-scalemax 25
setp hm2_7i97.0.8i20.0.1.current-maxlim 25
setp hm2_7i97.0.8i20.0.1.current-minlim -25


Something else I am missing maybe? I have 315vdc bus voltage.

nvmaxcurrent would be 3000 for 30 A
(note, you should only set nvparameters once)

PCW wrote:

nvmaxcurrent would be 3000 for 30 A
(note, you should only set nvparameters once)

To elaborate, this means that once the value has been programmed, and "stuck" you should remove the nvmaxlimit from the HAL file or you will eventually wear out the nvram. 

NV parameters are programmed when the sserial channel is enabled / disabled, so try toggling the sserial-enable pin on and off if the nvram setting does not seem to be sticking. 
 

Yep, got it working powerwise now, lifts the Z-axis and when looking halshow it tooks about 11amps to lift with jogging, 5amps to hold still, 30amps spike when accelerating..

But now I have another question, what really annoys me and tends to be REALLY hazard. So i have 8i20 controlling the XYZ, encoders are original fanuc pulse coders with commutation output also, but I am only using the ABZ encoder. But, the problem is, that if i have not stopped the servos in the right position (poles not matching to encoder reading?(I have the linuxcnc set so it will remember last position in startup)) they will jump, usually when trying to jog in the opposite position.. Few more enable and runaways and they usually locate themself. What is the easiest fix to this, just something wrong in hall, or should i use also the grayscale encoder? or what..image missing /media/kunena/emoticons/wassat.png

And for the other important question, i have z axis brake driven now from machine enable+delay in hal, but if the linuxcnc trips and lost the sserial communication, z servo will power off BUT the 7i97 will keep the z-axis relay powered until it gets communication again. So what would be the most safest and reliable way to get the axis brake work in pretty much every situation the servo is off?

Current hal and ini attached. I already have all axis homing working ok (switch+index, atleast i think it works as it should, dont have dial indicator for now to test for sure), just have one home switch per axis, no limits.


  

 

The jump sound like you are not doing a BLDC search for index at startup
(you could avoid this by using the Fanuc commutation data for BLDC commutation)

If you lose sserial communications you will have a high sserial error count
you could use this to cause a LinuxCNC fault or simply enable the brake