Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Davek0974 on June 29, 2016, 08:57:41 AM
-
I'm going through the settings on my bench setup of ac servo drives and the CSMIO-IP/A controller with Mach3.
I have steps per set as shown for my leadscrew/drive connection ratio but how do you calculate max velocity - It seems that if it is too high the drive faults on ePid error, too low and the 0-10v output does not reach 10v so velocity is wasted/lost.
How do you calculate the correct velocity to give 0-10v output or does it not work that way??
-
The setting you have in your servo drive for the RPM at 10v is the number you use.
So for example, you have your servo set to 3000rpm per 10v and you have a 2:1 ratio with a 5mm pitch screw.
3000rpm /2 (gearing) x 5mm pitch, Velocity in Mach would be 7500mm/min
Hood
-
Ah, thats interesting then as that is just what i tried, it seems I have a DAC offset so if it set max RPM it works perfect in one direction (negative offset) but in the other direction (positive offset) it faults out with ePid error.
I Have to reduce max el to something like 7550 to allow it to run in both directions without fault.
I doubt whether i'll be running 7500mm/min rapids though on an old mill ;)
-
Never had that issue, both drive and axis/spindle set up that way on the Chiron and it is fine either direction.
What is your tuning like?
Do you have too small of a following error set?
Hood
-
disclaimer - Don't forget this is all on the bench so no load etc on motor. :)
I have a max follow error set of 10,000 in pid tuning, using the pid tune jog buttons, if i set max el at 98% it all works well and the Max I accumulator bar graph shows just two blocks.
If i increase to 99% it faults out when the Max I accumulator tops out.
I have no idea what the Max I accumulator is telling me though???
The value limit is set for 1000,000 in Max I Acc.
This is with max vel set at 7620 in Mach3
-
Just been playing again, its the max following error topping out that causes the ePid fault, at 98% vel it reads about 500 or so, at 99% it just goes up and up until the limit trips.
Any help :)
-
Ok just to be sure, you have the rpm for 10v set in the drive and you have the Vel in Mach set up to be exactly the same or rather you have the Vel worked out from that same rpm?
500 following error with how many steps per unit?
I am still a bit worried about you not being able to tune the servo drive, it is the first drive I have ever come across where there is not an ability to tune manually and yours doesnt even seem to have an autotune function.
You could try altering the accumulator, think I messed about with it a bit and found I got lesser errors when it was actually topping out on a rapid.
Hood
-
The drive is pretty much as delivered - you just set the motor designation in the parameters and thats it it seems.
Steps per was calculated from 10,000 pulses from encoder, 2:1 ratio and 5.08mm pitch screw.
I will try reading the drive manual again but its heavy going ;)
-
That is a high following error then as it equates to 1.27mm
I got a manual for drives that are on eBay Uk, not sure if they are the same as yours or not but that manual has a manual tuning section. If they are the same let me know and I can pass the manual on, heres the ebay link.
http://www.ebay.co.uk/itm/201609613935?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT
Hood
-
Thanks Hood,
different drive, may be similar inside but connectors are different.
I need to do a tune with a load on the motor i think, it behaves differently with a finger on the shaft but that gets a bit painful after while ;)
Hopefully not too long before I can do a lash-up on the machine and see what happens.
-
I was out by 10x :D Error for 500 would be 0.127,mm, still not great though.
Let me know if you want the manual but likely if the connections are different then the drives firmware/internals will be different.
Hood
-
Thanks, thats a bit better ;)
Yes i have a feeling these drives are pretty specific to the manufacturer, so will have to play safe and stick with current manual, the seller was pretty helpful though so if there is a specific question, I can always ask, but as far as tuning goes the answer was "just enter the motor code"
-
I had one of the bigger motors set up today, the following error is better at 256.
I also set the drive to show rpm - they seem to give the ePid fault if it goes above 2950rpm for some reason.
It may be one of the parameters is set too tight in the drive, who knows but in reality I don't think it will matter, i'll just keep away from 100% full speed, 7500mm/min is a bit hairy on a bed thats only 48" long i think.
-
The Chiron has X@550mm Y@320mm and Z @290mm and it runs 20m/min. When I occasionally crank up the Beaver NC5 and see it running at 8m/min I think it is a tortoise :D
There is a big difference though with the stability of the machines, I could also get the Beaver NC5 up to 20m/min but I reckon it might topple :D, no chance of the Chiron doing that.
I would say somewhere between 6 and 10m/min would be the happy range for a knee mill.
Hood
-
Thanks, i would go for whatever it feels happy with but I'm not going to waste time trying to get that little extra a the top end, if it only runs at 7000mm/min then thats my limit.
I still have a mountain of battles to win at present - seems the bearings in the screw end plates are wrong and not the correct pre-loaded pairs they should be, i will need to shim them out to give some pre-load without binding like they are at present. The correct ones it seems are about £100 a pair but these will do if shimmed up for a while.
I had to get the Y axis front bracket machined as the bolt holes did not line up, seems the new dog-bone screw mount is slightly longer than the old BP one so the Y axis screw sits a little lower than it should, found a local guy that did it for a reasonable price. If only i had a mill ;)
Customers keep getting in the way with orders too ;)
-
Just remember for best results the drive needs to be set to same, equivalent value, as motor tuning for motors rpm at 10v.
Hood
-
Er, pardon ;)
Can you expand a little please Hood :)
-
Ok so whatever you have in Motor Tuning as your max Velocity should really be set as your equivalent Max RPM for 10v in the servo drive.
With your 2:1 and 5.08mm pitch screw then
If you had 3000 set in the servo as the rpm for the 10v command then that would equate to 7620mm/min in Motor Tuning for your Velocity
Now say you want to set Max velocity as 7000mm/min then the RPM for 10v in the servo drive would want to be set to
7000/5.08 x 2 = 2755.9, so 2756 would be ok.
If you do not keep these values the same then you can end up with tuning problems. If they are not too far out then it wont likely be too much of an issue but to give an extreme example to show what I mean.
You have rpm for 10v in drive set to 3000rpm
You have Velocity in Mach set to 3810mm/min
You command a rapid move and the IP-A sends out 10v but the drive sees 10v and thinks you want 3000rpm but because of the values in motor tuning it really only wants 1500 rpm at the motor. So the IP-A sees the axis shooting off and sends a corrective voltage but the motor hauls back too fast so IP-A again corrects and so on, you end up with an unstable axis.
This makes it extremely hard for you to tune the IP-A and for the IP-A to keep control of the axis.
As said that is an extreme example but always best to keep things in equilibrium.
Hood
-
Ok so i would need to alter the drive parameters so that 0-10v = 0-2756rpm instead of 0-3000rpm, then set Mach3 to 2756 as well in tuning?
The difficult part is finding that parameter in a Chinglish manual :(
-
For a 7000mm/min Velocity in Mach the rpm in the drive for 10v would be 2756 (with your gearing of 2:1 and pitch of 5.08).
Now as said for a small difference like that then it probably wouldn't be too much of an issue if you still had 3000rpm for 10v in the drive but as you start to get a larger difference from rpm in drive and the equivalent vel in Mach then it can start to cause real problems for tuning.
Hood
-
There are tuning parameters for the BST drives, however the descriptions/translation are pretty poor.
Pn153 up for Speed Control mode, and Pn188 up for Torque Control Mode.
Speed control mode is generally easier to setup, but Torque Control mode will give better performance if it can be set-up correctly (if you don't get it setup correctly, it'll give much worse and far more unstable performance than Speed Control mode).
Without an autotune function, you need to play with the settings manually, and you need to tune them while attached to the machine, as the extra load acts like a damper.
For comparison, my lathe holds within 6 counts on the Z axis, and 17 on the X-axis. The X-axis is a slant axis with a turret on it, so it's got a bit of an imbalance.
I ended up using speed control mode, as in torque control mode I came to the conclusion the internal filtering of the servo drive was causing too much of a delay to change of input voltage, as even with all filtering reduced to an absolute minimum, there was still a delay in the response plots (KFlop lets you plot position commands against encoder feedback, so you can see exactly what is happening).
-
Interesting, thanks. As you said, the translation is pretty poor, I would need an illustrated "How to.." guide before going anywhere near those settings i think ;)