Re: Jerk Control in machines
Quote:
Originally Posted by
dazp1976
When something says 'request a quote' it really means 'bend over please mate'.
Even having different licence levels, can't give a set price for a basic user.
Usually a red flag to overpaying for a service.
Get the negotiating head on.
This is something I rage at in this industry. I literally go to CMTS and ask company X "how much does this cost" and then can't tell me. If you can't tell me the price of the thing you are selling, I don't want it. What you inevitably find out when you push is the price is whatever they can extract from you.
It is getting a little better though, in part due to companies like autodesk, haas, etc who publish the actual price you pay and force other companies to at the very least have a legitimate base price. I say legitimate, cause if for example you called DMG Mori some years ago and said how "much is a milltap 700" they would tell you an FU base price that was idiotic like 180k cdn and then eventually 5 month later after you have gone to buy something else offer a package with all the trimmings for 107k cdn (an actual viable deal). Reason being they simply didn't care about single customers. These days with all the garage youtubers ready to drop 100k, they are much more responsive.
This goes for part suppliers as well, like THK, delta, siemens, etc when you want to build machines. The will actually respond to you as a customer these days, where NSK and yaskawa (at least as of 2-3 years ago) would give you the FU price and hope you didn't call back. Like NSK will literally quote you 2-3 times the price of the same ball screw listed on misumi usa which is already 2x the price of the same ballscrew on monotaro (japan). They are waiting for foxconn to call, they don't care about selling to you. Now that Yaskawa owns Siemens automation, I worry, although so far nothing seems to have changed (haven't had a quote since before covid though).
/rant
:)
Re: Jerk Control in machines
Quote:
Originally Posted by
mactec54
It is already integrated in my Cam software and there is no extra cost, if you have a cam system check with your manufacture if they have an option to have it integrated, I use Gibbs Cam, one of the higher end Cam software packages
What are the listed prices of the cam software it is integrated in? I can't find a single listed price.
Given the benchmark for pricing right now for fully integrated 3 axis CAD/CAM is "free" for hobby, and $300us a year for pro level with basic 5 axis, any stand alone cam software over that price better be leaps and bounds ahead for any hobby/single user to be interested.
Re: Jerk Control in machines
Quote:
Originally Posted by
ihavenofish
This is something I rage at in this industry. I literally go to CMTS and ask company X "how much does this cost" and then can't tell me. If you can't tell me the price of the thing you are selling, I don't want it. What you inevitably find out when you push is the price is whatever they can extract from you.
It is getting a little better though, in part due to companies like autodesk, haas, etc who publish the actual price you pay and force other companies to at the very least have a legitimate base price. I say legitimate, cause if for example you called DMG Mori some years ago and said how "much is a milltap 700" they would tell you an FU base price that was idiotic like 180k cdn and then eventually 5 month later after you have gone to buy something else offer a package with all the trimmings for 107k cdn (an actual viable deal). Reason being they simply didn't care about single customers. These days with all the garage youtubers ready to drop 100k, they are much more responsive.
This goes for part suppliers as well, like THK, delta, siemens, etc when you want to build machines. The will actually respond to you as a customer these days, where NSK and yaskawa (at least as of 2-3 years ago) would give you the FU price and hope you didn't call back. Like NSK will literally quote you 2-3 times the price of the same ball screw listed on misumi usa which is already 2x the price of the same ballscrew on monotaro (japan). They are waiting for foxconn to call, they don't care about selling to you. Now that Yaskawa owns Siemens automation, I worry, although so far nothing seems to have changed (haven't had a quote since before covid though).
/rant
:)
This is so true.
Years ago I went to Haase to buy a mini mill brand new. They told me flat out we do not sell to a non business entity or a home user and to go buy something used. Period. Next day I ordered a Tormach 1100 mill. And now own 3 Tormach machines
They have changed that policy . Lol not you can get mills, lathes whatever from them no problem.
Same with cad and cam software. There was little to choose from that was under 3k
Re: Jerk Control in machines
Quote:
Originally Posted by
mountaindew
This is so true.
Years ago I went to Haase to buy a mini mill brand new. They told me flat out we do not sell to a non business entity or a home user and to go buy something used. Period. Next day I ordered a Tormach 1100 mill. And now own 3 Tormach machines
They have changed that policy . Lol not you can get mills, lathes whatever from them no problem.
Same with cad and cam software. There was little to choose from that was under 3k
I never saw that with Haas in Canada going back to about 2003. What I do find is you gotta pay cash though. This is a bank think though will all machine brands. They wont least to anyone but an established corporation.
1 Attachment(s)
Re: Jerk Control in machines
Hi Mactec - Just trying to clarify what you mean by jerk. I've said this before lay "jerky motion" is machine compliance, rough motion and rough accelerations. Engineering jerk has no physical expression or symptom in terms of motion control. It is only a way to determine the rate of change of acceleration and value or visualise the smoothness of acceleration. Acceleration however creates inertial forces which can contribute to vibration and rough motion.
IHNF I looked at the sinumeric controller and the jerk setting limits jerk it does not "control" jerk. By control I mean the value is shaped or changed like driving a car and using the accelerator pedal. You control the speed via the pedal. If the car is speed limited once you get to that speed you have no control its capped. That's how the sinumerik works its a limit setting. This is apparent in the graph with the flat top jerk curves. For example the Rosetta controller has rounded top jerk curves, the controller has shaped the acceleration curve to smooth the jerk curve. Peter
Re: Jerk Control in machines
By control the industry means controlling it. There are 2 ways to read it i suppose, both ultimately meaning the same thing. One is that Jerk is controlling acceleration. The other is the control is controlling the amount of jerk (with a constant in this case). Both effectively mean the exact same thing. 3rd order motion profiles.
Your gas pedal analogy is EXACTLY what the siemens control is doing. Look at N110 of the graph. That is a jerk "controlled" ramp. Your gas pedal controls acceleration, your foot motion controls the rate of acceleration (stomping vs easing). Your foot is the "jerk controller" while the pedal is the acceleration controller. Speed is only controlled by matching resistance with force. The speedometer is the encoder and your brain runs the feedback loop. (we're talking mechanical throttle cars here of course, electric ones and fancy new computer controlled ones may behave differently). This is exactly how the siemens (and other) controls are controlling acceleration with a jerk constant. Your foot is of course a crude instrument and may not be applying a perfectly linear acceleration rate like the cnc control is.
I should note moves past N110 show a combined G64 constant velocity MIXED with 3rd order motion, which goes all squirrely. This is why they are 2 separate topics. They interact, and share some of the same results, but they are derived by different things. There are other pages in that article on SOFT/BRISK commands.
Re: Jerk Control in machines
hy pete, i wonder, is all clear with acceleration ?
i try to figure out exactly what your questions are, since you are ok with math, derivate/meaning of acc graph is not a problem
also, if i got it right, you can implement some shapes in your software, and move real things as you wish
so what is it ?
also, i have questions :
... can you analyze feed back data ?
... omg, i forgot the 2nd question / haha :)
Re: Jerk Control in machines
Hi DK - I have no problem with the math. I'm now writing up a control diagram to summarise what has been discussed and how I see a system working.
IHNV I must contest your comments on "control" of jerk via the seimens controller. In my world limiting a value is not control. Control is the ability to make something perform to a set shape or motion path. Machine controllers control position and velocity very well. The motor accelerates the load according to a velocity profile (same as you control speed via a pedal and depending on the load you may need more or less torque to hold constant speed or to accelerate as needed). The controller measures accel and jerk and if any of these get to the set limit the system limits the accel at the set limit or the jerk at the set limit by adjusting the velocity. The seimens controller does not "control" accel or jerk below its set points. The system accelerates at whatever natural accel it can depending on the torque applied until it gets to the set limits. The controller does control the velocity profile which creates the accel profile. Jerk is a result of the path and the accel... The thigh bone is connected to the knee bone, is connected to the foot bone etc
Hi DK - I have no controller or tools to measure the feedback at present. I can make FE models of these things and this will output V-A-J (snaaaaap) if needed but they take a long time (days) to get right so I wouldn't go there for some fun. I do intend to put an encoder on a motor and gather some data. I am investigating what that takes at the moment. I have a mathematics intern interested in this stuff so we are going to work towards a feedback loop on a motor for some fun. He loves doing PDE's matrices and differentiations so seems he'll be a good fit in this game. Peter
edit - the issue is divided between software and hardware. The software side CAD/CAM like volumill and Rosetta and others its easy to do the math and calculate position-velocity-accel and jerk. If the motor can provide the torque to accomplish the set profile then all good. But the software people generally do not know what hardware they are going to get used on. The hardware people, motor builders etc can use physical feedback to accomplish the smooth motion as it deals with the real inertial conditions. Some machine builders that supply SW and HW have this all integrated. Thanks to all that have contributed so far. Peter
Re: Jerk Control in machines
Quote:
Originally Posted by
ihavenofish
What are the listed prices of the cam software it is integrated in? I can't find a single listed price.
Given the benchmark for pricing right now for fully integrated 3 axis CAD/CAM is "free" for hobby, and $300us a year for pro level with basic 5 axis, any stand alone cam software over that price better be leaps and bounds ahead for any hobby/single user to be interested.
Ask Fusion 360 team if they can get VoluMill integrated in their package, it would not be in their Hobby level package package though, there cost per year is good even if a Hobby user wanted full access.
Re: Jerk Control in machines
Quote:
Originally Posted by
mactec54
Ask Fusion 360 team if they can get VoluMill integrated in their package, it would not be in their Hobby level package package though, there cost per year is good even if a Hobby user wanted full access.
So you don't know the price is what you are saying?
Re: Jerk Control in machines
Quote:
Originally Posted by
peteeng
Hi Mactec - Just trying to clarify what you mean by jerk. I've said this before lay "jerky motion" is machine compliance, rough motion and rough accelerations. Engineering jerk has no physical expression or symptom in terms of motion control. It is only a way to determine the rate of change of acceleration and value or visualise the smoothness of acceleration. Acceleration however creates inertial forces which can contribute to vibration and rough motion.
IHNF I looked at the sinumeric controller and the jerk setting limits jerk it does not "control" jerk. By control I mean the value is shaped or changed like driving a car and using the accelerator pedal. You control the speed via the pedal. If the car is speed limited once you get to that speed you have no control its capped. That's how the sinumerik works its a limit setting. This is apparent in the graph with the flat top accel curves. For example the Rosetta controller has rounded top jerk & accel curves, the controller has shaped the acceleration curve to smooth the jerk curve. Peter
Here is SoftServos from 2016 this covers some of what they can do at their control, there is no other control that comes close to what they have done, because most of their system they have is covered by Patents, I won't post the PDF but give you the web link as it has regulations on it's use, open the WMX2 Functions PDF
https://www.bing.com/search?q=SoftSe...ANNTA1&PC=LCTS
Re: Jerk Control in machines
Quote:
Originally Posted by
ihavenofish
So you don't know the price is what you are saying?
I pay $562-50 per year for Maintenance this comes with any upgrades, there may be throughout the year, plus any other support anyone would need with postprocessor changes Etc.
1 Attachment(s)
Re: Jerk Control in machines
Thanks Mactec - looks a good read. I'd like to modify the car accelerator analogy. The accelerator pedal does not control speed it controls the torque of the motor. The driver balances torque, accel and speed as the circumstances dictate. A cnc controller controls position via gcode and speed via the clock rate. ie the controller feeds the gcode out at the required F rate. Depending on the sophistication of the controller the speed and accel are limited by set points or they are ramped using an algorithm.
Now during my driver reading I came across the use of a capacitor across the driver power leads. Has anyone done this? Is it a good thing to do? I haven't done it on drivers and the driver instructions don't call for it??? I see the logic any thoughts? Peter
Re: Jerk Control in machines
Capacitors are used across dc lines to filter voltage ripple and make a more clean dc voltage. This is all done behind the rectifier inside the drive. Some DC drives may ask for external caps, but I've never seen it required on the ones I've had (siemens, indramat, JMC).
Some systems have optional AC line filters and chokes ahead of the rectifier - either inside the drive, or a separate part. The use of these usually depends on how bad your mains power is.
Re: Jerk Control in machines
Keeping it simply, the Jerk in a movement system is the derivate of acceleration. In case of a constant accelerated system the jerk is infinite at first change of speed. The strenght of jerk effects depends by mass of accelerated object. A way to reduce jerk in a speed vector variation is to apply a non constant acceleration quantum es: applying an epicycloyd or a s-curve with a gradient that depends by case. In a CNC system this happens in path-planning/smoothing/look-ahead level so all exes acceleration quantums are calculated to keep interpolated point right (none path deviation). Interpolator work in time domain so gcode single line movements are a flow of infinitesimal positions in time
Re: Jerk Control in machines
Quote:
Originally Posted by
ihavenofish
The okuma seems to be a blanket override. Can't find much on it.
it's mainly closed, even for fellow japanes officials; from early years, some of it's achievements are repacked in a cheapar manner/fanuc, then go worldwide
but in the last decade, this gap started to get smaller, as the worlwide competition cathced up, so the genos line appeared
only later i realized that it is "i have no fish" ... ha ha :)
Quote:
Originally Posted by
mountaindew
I went to Haase to buy a mini mill ... They told me flat out we do not sell to a non business entity or a home user
there are vendors that sell used machines with guarantee similar to a new one; competitive prices, like for the price of a new cnc, you would get an used similar one with higher specs and a new minimill; it's quite of a best deal for a buck :)
Quote:
Originally Posted by
mactec54
Feed rate with this control will leave most speechless
... most of their system they have is covered by Patents
hy mactec, i am pretty sure pete will deliver something to beat it :)
please, if i may ask, what is your background ? and on what machines type/size does that top cnc goes to ? are you involved in it's development ?
Quote:
Originally Posted by
peteeng
what is adaptive control doing?
( trying to ) output within a predifined range, like not only not to overshoot position, speed, inertia or cutting force, but keep it inside an user defined domain
some may argue that position is not caused by adaptive, being too basic to debate, while controlling cutting force is definetly adaptive
adaptive means output within desired tolerance
it may be live, or trained / kindly :)
Re: Jerk Control in machines
Can't control jerk if you have a total jerk controlling it.
1 Attachment(s)
Re: Jerk Control in machines
Hi - Here's a good article on control. Most people into the math understand that accel and jerk are 2nd and 3rd derivatives. That's simple. I'm wanting to know how that is implemented in the CAM. A more nuts and bolts discussion vs the top level maths. Peter
Re: Jerk Control in machines
hy :) from that article
Quote:
It should benoted that “continuous” feed motion cannot be realized on a
“discontinuous” geometry.
actually, it is possible :)
Quote:
Linear and circular segments (Figure 1) are
the upmost heavily used interpolations [6] defined by G00/G01
and G02/G03 commands
movement is not point to point, but interpolation, that is linear, arc or spline, thus machine makes a difference between g01 and g02
that interpolation is broken into segmentes, but at high resolution those segments can be few um or even < 1um, so movement is no longer point to point as it was for the machines that could deal only with g01 and nothing else
Quote:
or also
known as the “cornering tolerance” is introduced at the segment
junction points. This smoothing error will show up in the
machined geometry, and it should be considered in budgeting the
final form accuracy. In most cases, end-user can set the cornering
tolerance by adjusting the NC parameters
please notice, in the image that i shared at post 11, that the control may move on extensions, thus it avoids a filet, taking the corner on a longer way arround, thus this behaviour is way more precise than classical corner tolerance
real part outside corner will come out identicall, even if you use exact stop or continous speed mode; thus the deviations between dinamic modes may occur without altering the real part shape
in other words, beside corner tolerance, there may be also extension tolerance :)
Quote:
An advanced way to generate complex and accurate tool-paths
is to directly move the cutting tool on the originally designed
smooth “spline” CAD geometry itself. Unfortunately, todays CAM
and CNC system architecture limit direct data transfer and use of
the original part geometry.
there is no more the case of such limit
Quote:
As a matter of fact, if the linear segments are densely placed,constructed splines may actually show sharp changes in the
curvature. This generally occurs in precision die and mold
machining, and enforces the NC system to reduce the feed to avoid
actuator saturation on rough spline segments. An efficient way to
overcome this problem is to utilize approximate spline fitting
methods where NURBS curves are fitted with predefined
geometry errors
yes, for complex parts, die/mold, the density of the segments is so high, that it takes similar time to process and execute the toolpath, and desired feed may not be reached
so a larger buffer is needed, toolpath is spline interpolated in advance, and a degree 345 ... n is chosen for that spline; is important that the degree reflects real machine dynamic, for example, there is no use to interpolate at spline degree 5, if the real motion dynamic do not get even close to a spline degree 3
Quote:
As shown in the previous section, although CAM systems
deliver the initial tool-path, NC systems actually perform
significant amount of path planning to attain high performance
cutting.
if i may, this is not ok, because you can not say that the NC performs "signifiant amount" ... it is just/only a machine, it is built to do path planning, not "signifiant amounts" of path planing - see the difference ? :)
Quote:
Due to short segment length and limited path acceleration, planned
path speed may not be reached and eventually modified again.
nowadays it is not modified again, but constant from start to end, smaller than value from g code; there is no trial/trying to see it the feed value fits, but begining with a value that fits, that may be > = < feed value from g_code
Quote:
For instance, if a very slow
speed segment is placed right after a high-speed segment,
maximum velocity of the high-speed segment needs to be lowered
to be able to decelerate and connect to the slower one
here several aproaces are possible :
... it would have to lower it anyway, regardless of what is coming
... it would not care and blow througt it without speed loss, or maybe accelerating
"look ahead" may be there only to handle rad comp, or also dynamics, or also regression :)
Quote:
In complex
geometries computation of axis increments (dx and dy) with
respect to path displacement (ds) becomes computationally
challenging since the geometry is non-linear. If the computation is
not accurate the desired feed cannot be realized and “feed
fluctuations” occur.
again, here are discused systems that may deliver feed fluctuations; actual systems can deliver constant
that article is ok for it's time, meanwhile things evolved a bit :)
Re: Jerk Control in machines
Quote:
Originally Posted by
peteeng
Hi - Here's a good article on control. Most people into the math understand that accel and jerk are 2nd and 3rd derivatives. That's simple. I'm wanting to know how that is implemented in the CAM. A more nuts and bolts discussion vs the top level maths. Peter
It is not implemented in CAM, it is implemented in the machine control. CAM has absolutely zero to do with trajectory planning.
Good CAM can make the control's job easier but all aspects of motion are controlled 100% by the CNC control. The G code merely tells the control where it would like to end up and what speed it would like to go. To fit you your car analogy, CAM is google maps and the speed limit signs are "F".