- Posts: 3410
- Thank you received: 0
G93 Inverse Time Mode
12 years 7 months ago #44507
by BigJohnT
G93 Inverse Time Mode was created by BigJohnT
I'm trying to write a Python script to add G93 to a XZA machine. I'm not having a lot of luck figuring it out. I assume I need to use trig to calculate the distance moved and I think I have that correct but I keep getting following errors from what I assume is too fast of a F word. Anyone have any ideas on the proper way to calculate the F word for inverse time mode?
Thanks
JT
Thanks
JT
Please Log in to join the conversation.
12 years 7 months ago #44530
by andypugh
Replied by andypugh on topic G93 Inverse Time Mode
That doesn't make sense, the feed rate should just be throttled if an axis can't keep up (unless this is a complex kins)
I have never used G93, so I am not really qualified to comment, but that has never stopped me in the past.
As I understand it you need to calculate how long the segment should take, in minutes, based on path-length / speed.
Then the F-word is the inverse of that time.
Actually, you can miss a step, the F-word just needs to be (speed units/min) / (length in machine units).
I note that the docs state than an F-word is needed on every motion line.
I have never used G93, so I am not really qualified to comment, but that has never stopped me in the past.
As I understand it you need to calculate how long the segment should take, in minutes, based on path-length / speed.
Then the F-word is the inverse of that time.
Actually, you can miss a step, the F-word just needs to be (speed units/min) / (length in machine units).
I note that the docs state than an F-word is needed on every motion line.
Please Log in to join the conversation.
12 years 7 months ago #44533
by BigJohnT
Replied by BigJohnT on topic G93 Inverse Time Mode
It's just a simple Axis sim with the A axis added and Y removed. It seems like only the really short segments are the problem.
JT
JT
Please Log in to join the conversation.
12 years 7 months ago #44538
by ArcEye
Replied by ArcEye on topic G93 Inverse Time Mode
I am not qualified either :laugh:
The person who wrote this seems to be
www.eng-tips.com/viewthread.cfm?qid=200454
That sort of makes sense to my long disused maths head
regards
PS Looks like you got plenty of answers on the users list
Once they get over the 'F Word' thing, you will probably get the same answer
The person who wrote this seems to be
www.eng-tips.com/viewthread.cfm?qid=200454
F(code) = 1(minute) / (time = 3D distance/velocity)
The 3D distance of the move is calculated in model coordinate space at the NC control point, not in machine coordinate space and not necessarily at the tool tip. For example, a 5-inch move at 50 IPM takes 5 50ths of a minute, yielding an inverse time calculation of 1/.1 and an F-code of F10. The same 5-inch move at 700 IPM would be 1(minute) / (time = 5 / 700) or (1/(5/700)) or (1/.0071428) or F1400.168
That sort of makes sense to my long disused maths head
regards
PS Looks like you got plenty of answers on the users list
Once they get over the 'F Word' thing, you will probably get the same answer
Please Log in to join the conversation.
12 years 7 months ago #44540
by andypugh
Yes, but 1 / (x / y) is the same as y / x
So, there is no point doing 1 / ( distance / speed ), you can just calculate ( speed / distance ).
Replied by andypugh on topic G93 Inverse Time Mode
ArcEye wrote: F(code) = 1(minute) / (time = 3D distance/velocity)
Yes, but 1 / (x / y) is the same as y / x
So, there is no point doing 1 / ( distance / speed ), you can just calculate ( speed / distance ).
Please Log in to join the conversation.
Time to create page: 0.175 seconds
