Mori MVJR Build Log
- schmidtmotorworks
- Offline
- Elite Member
-
- Posts: 286
- Thank you received: 6
I still suggest that you normalize the PID numbers by setting both PWM scale and MAX OUTPUT to machine units/sec at 1825 RPM.
I must be missing something then because when I set
OUTPUT_SCALE to 11.81 = (pitch*RPM/60) = (0.3937*1800/60)
and
pwmgen = 1
The motor speed can't be adjust by the pot to 1800, it can only get up to about 600 RPM.
Maybe I'm not looking at the pwmgen value properly, I thought it has a range from 0-1.
Should I be setting pwmgen to some number higher than 1?
Please Log in or Create an account to join the conversation.
And yes the range of the PWM generator is +- OUTPUT_SCALE
You should set OUTPUT_SCALE and MAX_OUTPUT to 11.81 (11.81 Inches per second)
now when you set the PWM value to 11.81 you should get 1800 RPM (and 11.81 Inches per second linear motion)
This is what I mean by normalizing the PID, when set up this way, the PID loops outputs values directly in inches per second
Please Log in or Create an account to join the conversation.
- schmidtmotorworks
- Offline
- Elite Member
-
- Posts: 286
- Thank you received: 6
Please Log in or Create an account to join the conversation.
The -1 to 1 is just the default range...
Please Log in or Create an account to join the conversation.
- schmidtmotorworks
- Offline
- Elite Member
-
- Posts: 286
- Thank you received: 6
I made a program as suggested that moves the axis from 0 to 1 and back to zero.
So far I have improved to the point that I have the PID display gain so the scale read 50um and the graphed curve is about 2-3 graduations. Not sure what good enough is yet and if this relates to follow error or something else.
It seems like the machine moves with less acceleration than it did before, as it seems much less violent than it used to be when it was in 100% rapid.
I also notice that the DRO doesn't make it to 1.000, more lie about 0.720, not sure if that is the DRO lagging or if the machine is falling behind.
I have the following setting in INI
MAX_VELOCITY = 9.833 (this equals the max specified in the manual.)
MAX_ACCELERATION = 20
I looked in the manuals for the CNC and didn't find any indication of an acceleration spec.
Is there a best practice for determining the proper MAX_ACCELERATION ?
Please Log in or Create an account to join the conversation.
I have the following setting in INI
MAX_VELOCITY = 9.833 (this equals the max specified in the manual.)
MAX_ACCELERATION = 20
...
Is there a best practice for determining the proper MAX_ACCELERATION ?
There probably is, but I have no idea what it is. However, most of the configs I have seen have acceleration more that 2x max velocity.
You could try increasing the acceleration until things go wrong...
Assuming you are using Halscope, does that agree with the 0.720 figure?
Please Log in or Create an account to join the conversation.
I also notice that the DRO doesn't make it to 1.000, more lie about 0.720, not sure if that is the DRO lagging or if the machine is falling behind.
The DRO is running in userspace land and is only accurate at rest. How closely the DRO follows the path depends on the CPU load among other things that tend to slow up non-realtime things. That is also the reason the backplot does not exactly follow the path the machine takes.
John
Please Log in or Create an account to join the conversation.
HALScope displays what you tell it to, Are you selecting FERROR on the selected axis in HALScope?
FERROR _is_ the following error
Unless you are plotting FERROR I cannot tell whether the machine is anywhere near tuned (the DRO difference could be that you have a long long way to go tuning)
Please Log in or Create an account to join the conversation.
- schmidtmotorworks
- Offline
- Elite Member
-
- Posts: 286
- Thank you received: 6
I don't know if PID error is an indication of following error.
FERROR _is_ the following error
Thanks, I'll start tracking that too then.
You could try increasing the acceleration until things go wrong...
I wonder what the symptom of too much or too little acceleration would be.
Please Log in or Create an account to join the conversation.
Too much would cause too large of a following error during acceleration.
Too little will make the machine slow on small details
There is also an unseen factor here, the actual torque needed for cutting and perhaps more table load, that is if you adjust your max acceleration to the point that the machine is just able to follow with no load, this means all the drives torque is used for accelerating the bare machine at maximum acceleration and none is left for cutting and table loads
It would be nice to know what the original machine spec was. but lacking that, you could bump up the acceleration until you get large following errors and then back of by perhaps 30 to 50%
Note that setting acceleration up to the point that your drives cant follow is liable to test the drives and machine rather severely so be careful!
BTW I would not mess with this until all the other tuning details are worked out
Please Log in or Create an account to join the conversation.