-
Re: Delta servo drives and servos.
Hi,
Quote:
Alinear = 3395rad/s2 * 0,005m / 2 * pi
Alinear = 26,66m/s2
No you have misread the forumla.
dw/dt=3395 rad/s2
Arotational= dw/dt /2pi (revs/s2)....ie divide the angular acceleration in radians per sec2 by 2pi to
get angular acceleration in revolutions per sec2. Now Alinear is trivial:
Alinear= p x Arotational
= dw/dt .p/2pi
= 0.005 x 3395/2pi
= 0.005 x 3395/6.283
=0.005 x 540.33
=2.701 m/s2
Craig
-
2 Attachment(s)
Re: Delta servo drives and servos.
Quote:
Originally Posted by
joeavaerage
Hi,
No you have misread the forumla.
dw/dt=3395 rad/s2
Arotational= dw/dt /2pi (revs/s2)....ie divide the angular acceleration in radians per sec2 by 2pi to
get angular acceleration in revolutions per sec2. Now Alinear is trivial:
Alinear= p x Arotational
= dw/dt .p/2pi
= 0.005 x 3395/2pi
= 0.005 x 3395/6.283
=0.005 x 540.33
=2.701 m/s2
Craig
That makes sense, my bad. I compared the calculations to the Yaskawa software and the results are spot on, if you use ballscrew efficiency of 1 (0,85-0,95 in reality).
Anyway, I quickly made and excel spreadsheet for calculating horizontal axis maximum acceleration, based on user inputs (yellow cells). You can even spec 1:1 direct drive (with coupling OR belt drive) and any other gear ratio with belt drive. Just insert rotation moment inertia for the pulleys OR the coupling. The calculations will include these values automatically.
-
Re: Delta servo drives and servos.
Hi,
Quote:
if you use ballscrew efficiency of 1 (0,85-0,95 in reality).
Yes I forgot to include that factor.
If need be I could use a low lash planetary gearbox of 3:1. That would increase acceleration at the expense of speed.
At this point in time I not going to change anything. My current mini-mill has an acceleration, set in Machs motor tuning of
375mm/s2. The steppers and gearbox combo mean I could have a lot more acceleration but the machine
starts to throw itself around the room unless I tie it to the wall. I've found 375mm/s2 is entirely adequate for
my mini-mill and it follows toolpaths I set it no problems.
My inclination is to try acceleration of 2.7m/s2 and see how it goes.
As you've shown it takes only 90ms to get to full speed or 12mm traveled. It might well be enough.
Craig
-
Re: Delta servo drives and servos.
Quote:
Originally Posted by
joeavaerage
Hi,
Yes I forgot to include that factor.
If need be I could use a low lash planetary gearbox of 3:1. That would increase acceleration at the expense of speed.
At this point in time I not going to change anything. My current mini-mill has an acceleration, set in Machs motor tuning of
375mm/s2. The steppers and gearbox combo mean I could have a lot more acceleration but the machine
starts to throw itself around the room unless I tie it to the wall. I've found 375mm/s2 is entirely adequate for
my mini-mill and it follows toolpaths I set it no problems.
My inclination is to try acceleration of 2.7m/s2 and see how it goes.
As you've shown it takes only 90ms to get to full speed or 12mm traveled. It might well be enough.
Craig
Sounds like a good plan! Anyway the excel should help anyone who is not sure about what size of servo motor they need. Rule of thumb would be to use a safety factor of at least 2, which eliminates the need of friction coefficients, ballscrew efficiency and transmission efficiency. Also remember that what is critical is when the servo motor need to quickly accelerate/decelerate . Then you have a peak torque of 300% at your disposal anyway.
-
Re: Delta servo drives and servos.
Hi,
I thought I'd post my derivation of the forumla that combines linear momentum with rotational momentum. Its been many years since
I opened a physics book let alone read one!
My hypothesis is that there is some effective first moment of inertia, Jeff such that the total kinetic energy, Etot of a linearly
accelerating axis AND rotating ballscrew/servo assembly is described by:
Etot = 1/2 * Jeff * w2...............................equation [1]
Note I use MKS units:
Etot in Joules
Jeff in kg*m2
w in radians per second, rad/s
But we know that the total kinetic energy has two components, first the translational energy of the axis mass and the rotational energy of the rotating components:
Etot = 1/2 m*v2 + 1/2 Jcomb* w2.............. where Jcomb is the first moment of inertia of the rotating parts
But the axis velocity v is related to the angular velocity of the ballscrew, after all thats why we use ballscrews to precisely translate rotoational position
to linear position:
v = w/2pi * l where l is the pitch of the ballscrew in meters. Substituting:
Etot = 1/2 *m*(w*l/2pi)2 + 1/2 Jcomb*w2
=1/2*w2 { m*(l/2pi)2 + Jcomb }...................................equation [2]
Now we have two equations describing Etot, equation [1] and equation [2], using the principle of equating coefficents:
1/2 *Jeff * w2 = 1/2*w2 { m*(l/2pi)2 + Jcomb }
Jeff=m*(l/2pi)2 + Jcomb
So the component of the total effective first moment of inertia that is due to the linear movement of the axis is:
m*(l/2pi)2
For example my axis is 110kg, and the ballscrew pitch is 5mm or 0.005m:
=110 * (0.005/2pi)2
=110 * (0.005/6.283)2
=110 * (0.0007958)2
=0.69 *10-4 kg*m2
Craig
-
Re: Delta servo drives and servos.
Hi,
thought I would do some calculations about what a 2.7m/s2 acceleration would mean for my new build machine.
While the mill beds (75kg grey cast iron) are 700mm long, the linear rails atop them are 650mm and the 32mm ballscrew are 650mm
overall. The ballcrew (BNFN by THK) is double nut so the free travel of each ballscrew is 358mm.
My design calls for travels between limit switches of 350mm x 350mm x 350mm.
I had intended to use 'field weakening' that would allow my Delta servos to run at 5000 rpm despite being rated at 3000rpm.
If I use this strategy then the nominal rated torque of 2.4Nm is available up to 3000 rpm but then it decreases linearly between
3000 and 5000 rpm, to an end point of about 1.44Nm.
This decreasing torque would prove a major complication in the simple claculations I wish to do. I have elected therefore to
not utilise the field weakening operating regime.
The overall parameters are:
Servo: rated torque=2.4Nm
rated speed=3000rpm
Ballscrew: pitch =5mm or 0.005m
G0, or maximum axis speed is 3000 x 0.005=15m/min or 0.25m/s
Time to accelerate to 0.25m/s:
t=v / a
=0.25 /2.7
=92ms
Distance travelled is over that 92ms acceleration interval is:
d=1/2 a t2
=1/2 * 2.7 * 0.0922
=0.0116m or 11.6mm
The same time and distance would occurr at deceleration.
The time take to traverse from one limit switch to the other (350mm) is:
ttot= (0.35 -0.0116 -0.0116)/0.25 +92ms +92ms
=1.49s
So my machine will traverse it heaveist axis from one limit switch to the other in 1.5 seconds or so. I think I can wait that long!!:)
Craig
-
Re: Delta servo drives and servos.
Hi,
in fact if a nominal 110kg axis can lurch 350mm in only 1.5 seconds I'm seriously going to have to tie my machine to the wall
or bolt it to the floor!
Craig
-
Re: Delta servo drives and servos.
Quote:
Originally Posted by
joeavaerage
Hi,
in fact if a nominal 110kg axis can lurch 350mm in only 1.5 seconds I'm seriously going to have to tie my machine to the wall
or bolt it to the floor!
Craig
You don't have enough machine mass if this is a problem, or to much motor for what you are trying to do, dumbing down a motor so it won't move your machine around is as dumb as it gets this makes all your numbers meaningless
-
Re: Delta servo drives and servos.
Gents can I ask your option on running the late 90's Panasonic servos on my Bed Mill for a control retrofit. Will they achieve reasonable finishes given they are in effect around 5000 ppr for 5 mm screws. I found this post which indicates they would be adequate ("reasonable accuracy and smooth motion") but more resolution would be better. https://centroidcncforum.com/viewtop...9bf3846411d6d3
Obviously later servos run much higher resolution and smoothing algorithms. So what's adequate mean and what the difference in finish compared to new servos?
Do I go to the cost and time input and buy 3 new delta or Yaskawa systems? I assume I will see a marked difference??
-
Re: Delta servo drives and servos.
Hi,
Quote:
Obviously later servos run much higher resolution and smoothing algorithms. So what's adequate mean and what the difference in finish compared to new servos?
Do I go to the cost and time input and buy 3 new delta or Yaskawa systems? I assume I will see a marked difference??
It too have 5mm pitch ballscrews and I use electronic gearing to arrive at 5000 pulse/rev or 1um per step. The finish is perfect. Just because the servos
have a 160,000 count per rev does not mean you have to use it, and even if you did then you would require a motion controller that could produce
step/direction signals in the low mega-Hertz range, at huge cost......and to what advantage.......none.
If you went and bought new high resolution servos you would see no, or virtually no difference to surface finish or smoothness. Later model
servos may have better tuning aids and control modes that may help but the extra encoder resolution will make no difference.
Craig
-
Re: Delta servo drives and servos.
Quote:
Originally Posted by
joeavaerage
Hi,
It too have 5mm pitch ballscrews and I use electronic gearing to arrive at 5000 pulse/rev or 1um per step. The finish is perfect. Just because the servos
have a 160,000 count per rev does not mean you have to use it, and even if you did then you would require a motion controller that could produce
step/direction signals in the low mega-Hertz range, at huge cost......and to what advantage.......none.
Electronic Gearing does not change the Encoder you still have 160,000 PPR at the motor your electronic Gearing of 5000 step / per is 32:1 Ratio
Quote:
Originally Posted by
joeavaerage
If you went and bought new high resolution servos you would see no, or virtually no difference to surface finish or smoothness. Later model
servos may have better tuning aids and control modes that may help but the extra encoder resolution will make no difference. Craig
That is total BS :)
-
Re: Delta servo drives and servos.
Thanks for the feedback guys albeit opposing, it adds some perspective, and this post in general has been really helpful.
May I ask for another opinion which is "What Control System?", it's for a clone BT40 bed mill. I did another post here but would be interested to hear your views (opposing or not) on this. There's the mainline suppliers (too expensive??) Chinese GSK and others, maybe with little support, and Acu-ritie and Protract which are 3 axis only, for all the install work it would be preferable to have options for more axis IMHO. . Over Xmas I had some spare time and played with LinuxCNC, all very interesting and potentially powerful, and got an installation running on the computer which can be complied but it was a strain and time soaking, and not sure if the end result would be an easy to use fully sorted system. and that comment is as much about my knowledge as LinuxCNC. I still had to get a card (6 axis) and get it running which didn't have the FPGA setups done, and learn their INI file system and the setup for talking to the outside world etc and thought it would have been still very time consuming to get it up and running. I've also looked hard at Centroid Oak which seems pretty good and sorted system and is good for DIY folks although the 5th full axis is sort of blocked by US regs. Not dirt cheap with options.
I'm aware of motion control cards but you have to get some CNC software to talk to them I guess, not sure how play and play the whole setup is but with limited tech skills I don't think I wanted to get into stuff is too complex. SO if you've got any comment on this or other control ideas let me know. Thanks
-
Re: Delta servo drives and servos.
Hi,
not so many years ago, say 10-15 years, the state of the art servos had 2500 line (10,000 count/rev) encoders.
Mactec54 seems to suggest that a servo with an encoder of that description is incapable of smooth movement
and lacks the resolution required for a good surface finish, and that as we know is BS.
Its enirely probable tha your Panasonic servos have encoders or resolvers in that same category and they were
certainly very capable of great surface finishes, Panasonic was and remains a significant brand in the servo market.
I use Mach4Hobby ($200) and an Ethernet SmoothStepper ($180) as motion controller, I like it and reccomend it.
Others like UCCNC ($60?) and a UCxxx ($150?) motion controller.
Yet others use Centroid Acorn ($300).
Yet others use LinuxCNC (free) plus extra parallel hardware like Mesa (approx $300).
All have their strengths and weaknesses but all are pretty capable and all are within a hobbyists budget.
Craig
-
3 Attachment(s)
Re: Delta servo drives and servos.
Quote:
Originally Posted by
joeavaerage
Hi,
not so many years ago, say 10-15 years, the state of the art servos had 2500 line (10,000 count/rev) encoders.
Mactec54 seems to suggest that a servo with an encoder of that description is incapable of smooth movement
and lacks the resolution required for a good surface finish, and that as we know is BS.
Its enirely probable tha your Panasonic servos have encoders or resolvers in that same category and they were
certainly very capable of great surface finishes, Panasonic was and remains a significant brand in the servo market.
Craig
More of your uneducated BS (wedge)
The Panasonic are a level up on your Delta so would help if you new what you where posting about
20 years ago Yaskawa which was the most used servo system for machining centers where using 17Bit Encoders, and yes the sigma II system has been in use for that long :)
2500 line incremental encoders for wood working would be ok, but everyone had to start somewhere, it sure was not smooth in anyway as you say :)
Bridgeport which was one of the first CNC Machines ever built they used Stepper motors and used no Encoders anything is posable depending on what you want out of a machine as for smooth cutting no that was not posable but was a start of the CNC evolution
To see how far out of touch you are Fanuc have been using 10million CPR for the last 10 years on there standard model machines and offer much higher here is some of there spec's
-
Re: Delta servo drives and servos.
Hi,
well there you have it, according to Mactec54 your Panasonic's are incapable of smooth running.
Makes you wonder how it is that Panasonic became one of the top tier manufacturuers?
Craig
-
Re: Delta servo drives and servos.
Quote:
Originally Posted by
joeavaerage
Hi,
well there you have it, according to Mactec54 your Panasonic's are incapable of smooth running.
Makes you wonder how it is that Panasonic became one of the top tier manufacturuers?
Craig
How did you figure that out by yourself, as I said nothing about what a Panasonic servo system can do or can't do (nuts)
Who said it was a top tier manufacture !! are you sure you can read what is posted, you seem to be good at making stuff up :)
-
Re: Delta servo drives and servos.
Quote:
Originally Posted by
joeavaerage
Hi,
not so many years ago, say 10-15 years, the state of the art servos had 2500 line (10,000 count/rev) encoders.
Mactec54 seems to suggest that a servo with an encoder of that description is incapable of smooth movement
and lacks the resolution required for a good surface finish, and that as we know is BS.
Its enirely probable tha your Panasonic servos have encoders or resolvers in that same category and they were
certainly very capable of great surface finishes, Panasonic was and remains a significant brand in the servo market.
I use Mach4Hobby ($200) and an Ethernet SmoothStepper ($180) as motion controller, I like it and reccomend it.
Others like UCCNC ($60?) and a UCxxx ($150?) motion controller.
Yet others use Centroid Acorn ($300).
Yet others use LinuxCNC (free) plus extra parallel hardware like Mesa (approx $300).
All have their strengths and weaknesses but all are pretty capable and all are within a hobbyists budget.
Craig
My older brother tapping center (1992?) has 1024 line encoders. 5 pitch screws, 700ipm, 400w.
My newer brother tapping centre has 2500 line encoders, 10 pitch screws, 1400ipm, 720w.
Doesn't take long to realise they are "the same".
The encoder resolution does help with high dynamics, but, your machines (sorry), don't qualify in that realm :)
Things to know where though, before this devolves into... well, I cant read them, I've had mactec on ignore for the last 5 years. :)
1: Just like with a stepper, you want your servo to have the resolution that you need in practice. If you want 0.001" precision , then you want an encoder resolution of 0.0001" or better. The end resolution entirely depends on your machine. My brother tc225 has a 5 pitch screw, so that means I would "like" 2048 encoder counts. I have 1024, which is 0.00019". "close enough" for a machine that was meant to do straight line tapping. if you have a rack and pinion, you'll want a much higher resolution encoder, or a gear reduction - although at the same time, that type of machine isn't aiming for 0.001" precision usually.
2: Do NOT use a second linear encoder on a machine that has high backlash and linear guides. you will get "jitter" as the machine tries to find its place. So, if you have rack and pinion, no scales. There are exceptions of course.
3: In addition to being a complete twat who thinks the hobby market is beneath him, the north American Yaskawa rep will NOT support drives not bought from them. If you mention you got them from china (the drive and motor models are tagged with the country btw) they wont even reply. So, be prepared to have no support if you go for the cheap kits for fast2buy
4: fast2buy are great, don't be afraid to buy from them
5: Delta A series drives, are "more or less" the same as the equivalent Yaskawa drive feature wise. the A2 is similar to the sigma 5, and the a3 is almost identical to the sigma 7. The Delta is just 2/3 the cost.
6: Yaskawa own both Omron and Siemens automation.
7: When choosing a servo size, inertia will likely be your main concern, not power. if you have rack and pinion, you may need much larger servos than you think, even though you don't actually need the power. This is one area where the Yaskawa sigma 7 does shine (and delta A3) because it can dynamically tune larger inertia rations. Drives like the delta B2 can only handle small ratios. So you might need for example a 750w delta B2, where a 200w Yaskawa or delta a3 will work.
8: External braking resistors are only really needed when you have very high acceleration and inertia. My brother TC229 has 500w resistors per motor, and 1.0G acceleration. It also has a 1kw resistor on the spindle, because, reasons :). If tour machine is not particularly heavy, and not accelerating very fast, you wont need a resistor other than the built in one.
Ok, you can now go back to your regularly scheduled bickering over nonsense.
-
Re: Delta servo drives and servos.
Quote:
Originally Posted by
ihavenofish
6: Yaskawa own both Omron and Siemens automation.
Fact check Yaskawa have never owned Omron they had a partnership a few years ago that Omron did sell rebadged Yaskawa servo products, but no longer do :)
Siemens they do own 100%
-
Re: Delta servo drives and servos.
"....Finally during the fall of August 2004, Yaskawa brought the last branding partner, Omron IDM, into the fold as well. Yaskawa did so by purchasing the rights to Omron IDM Controls’ HVAC and Industrial channels"
The relationships get really screwy (they do with most big Japanese companies), because Omron safety is a different company. Doesn't really matter, other than knowing they are very closely related and the products are "the same". (there were rebranded siemens Yaskawa drives for a while too, though they weren't machine tool class)
Back to delta... a few more good points:
- You can buy them at digikey with next day shipping. So if one blows out, and your in the middle of a paying job, you can get replacements easy.
- The "B3" is out now. 24 bit encoder, much shorter motor, 350% overload, 6000rpm, smaller drives. Price seems to be slightly more than the B2, but not as much as the A series.
fast2buy has some in stock - I'm gonna try a 400w model for my spindle, should be entertaining :)
-
Re: Delta servo drives and servos.
Quote:
Originally Posted by
ihavenofish
"....Finally during the fall of August 2004, Yaskawa brought the last branding partner, Omron IDM, into the fold as well. Yaskawa did so by purchasing the rights to Omron IDM Controls’ HVAC and Industrial channels"
The relationships get really screwy (they do with most big Japanese companies), because Omron safety is a different company. Doesn't really matter, other than knowing they are very closely related and the products are "the same". (there were rebranded siemens Yaskawa drives for a while too, though they weren't machine tool class)
Back to delta... a few more good points:
- You can buy them at digikey with next day shipping. So if one blows out, and your in the middle of a paying job, you can get replacements easy.
- The "B3" is out now. 24 bit encoder, much shorter motor, 350% overload, 6000rpm, smaller drives. Price seems to be slightly more than the B2, but not as much as the A series.
fast2buy has some in stock - I'm gonna try a 400w model for my spindle, should be entertaining :)
Yaskawa ended the Partnership a little over 10 years ago, and it was only a Partnership, Omron now sell Rebadged Panasonic servo drive and Hitachi VFD Drives, they don't sell any Yaskawa drives Etc of any kind , here is a
Snip from 2009
After months of speculation and rumour, Omron and Yaskawa Electric have announced that they are ending their European drives joint venture, Omron Yaskawa Motion Control, which they established in 2003. In a terse statement issued earlier this month
In future, Omron will sell badged inverters made by Hitachi and servo drives from Panasonic Full details are expected to be released in November at the SPS/IPC/Drives show in Germany. 2007
Yaskawa have only had ownership of Siemens since 2018 have not seen any siemens rebranded drive not in the USA anyway may be something in Europe I can't see them changing any of the Siemens CNC Control system
I like the Delta Servos drives also, the knockoff's from China though is a concern, you can even get Yaskawa and most other brands of servo from China also but not all are knockoff's
-
Re: Delta servo drives and servos.
A few years ago I delved into ballscrew math on the forums. Been away from the forums for a while. Worked alot of overtime during the past two years and now I find myself, like many others, with time on my hands to work on my projects and look for a new job. So in the process of making my own thread about a problem I am now having, I came across this thread (please feel free to help me with my problems in my thread if you know anything about AC servo drives :)).
Did any of the interested parties in this thread find any of my previous posts where I went through the equations I used, to accommodate for inertia, including sample calculations and graphs, in your searching of the forums? They are the same equations that are available in all of the major ballscrew manufacturers literature, although some use different units.
I was hoping that someone else would jump in on the forums and validate my spreadsheets with their own spreadsheets, to compare results, but no one did, at that time.
Quote:
Originally Posted by
joeavaerage
I've done the calclations for the acceleration potential of my machine including rotational inertia. I've had my calculation checked by peteeng, a professional mechanical engineer. Additionally I have Hiwins calculation formulas and can double check all the calculations and they all point to the same conclusion.
Who is this peteeng? Where was he when I wanted to compare notes with someone? LOL. Hey, feel free to ask him to join in the conversation here.
From your full post here's what I got:
Motor rotational inertia = 1.13 x 10^-4 kg*m^2 = 0.000113 kg*m^2
Ballscrew rotational inertia = 5.252 x 10^-4 kg*m^2 = 0.0005252 kg*m^2
0.0005252/650 for a 650mm long ballscrew gives
0.000000808 (kg*m^2)/mm
From the THK rolled ballscrew moments of inertia pdf that I have, a 32mm ballscrew has....
8.08 x 10^-3 (kg*cm^2) / mm = 0.00808 (kg*cm^2) / mm = 0.000000808 (kg*m^2) / mm
And that is the exact value that you have posted. Did you use the same source? Ground ballscrews and other manufacturers will be slightly different, but pretty close.
pitch = 5mm
mass = 110kg = 242 lbs
Quote:
Originally Posted by
joeavaerage
dw/dt= 2.4 /7.07 .10-4
=3395 rad/s2
Ok, I see, you are using 2.4 N*m as the torque. So 2.4 N*m as the continuous and around 7 N*m as the peak? So hard to figure out what you have available for acceleration when you have peak and continuous but the amount of time spent in peak depends on the geometry of what you are cutting. I haven't figured out a simple answer for this yet. But in general, you will be able to do better than the continuous torque.
2.7 / 9.8 = 0.275G
With a ballscrew efficiency of 1, neglecting couplers, neglecting preload drag, neglecting linear bearing friction, etc, I get almost exactly that. Around zero (-0.34 lbs) available cutting force at 0.275G using 2.4 N*m as the available torque. It's almost as if he used the exact same equations as I did :). That is assuming that the axis you are referring to is being run by a single motor. If this axis is being run by two motors (do you have a build log, I clicked on your profile for latest started threads but it said there were no results?), it doesn't change much. So yeah, 32mm ballscrews are a bad idea unless you really need them (which you don't...I can say this with confidence because your axis would in all probability weigh way more than 242 lbs if you did, or if that would make any difference to your design). A 32mm diameter, 5mm lead ballscrew is pretty horrible for most CNC applications.
Of course, I'm just doing this as fast as I can, and I've had several beers. So I might need to double check all of this later. I could have made an error. Standard disclaimers apply.
Keep in mind with these graphs, that they don't include many factors such as preload torque, bearing friction, coupler inertia, etc. Also, this math is just theoretical, and doesn't incorporate many many other factors, like machine stiffness, spindle performance, etc, that can make or break a machine.
If you look at the last graph, you can see that a 25mm diameter ballscrew with 10mm lead would have been a far better choice IMO. I'm guessing (not really knowing anything about your build) that even a 20mm diameter ballscrew would have worked.
The good news for you is that, IMO, from what I've seen, 0.25G (more considering peak power) should still give you OK results. Your ballscrew lengths are fairly short, and your servos are large, and that goes in your favor.
Quote:
Originally Posted by
NordicCnc
Anyway, I quickly made and excel spreadsheet for calculating horizontal axis maximum acceleration, based on user inputs (yellow cells). You can even spec 1:1 direct drive (with coupling OR belt drive) and any other gear ratio with belt drive. Just insert rotation moment inertia for the pulleys OR the coupling. The calculations will include these values automatically.
I tried having a look at your excel spreadsheet to see what you have done but couldn't get it to open. No worries, I was just curious.
To the OP; I really don't know too much about servo drives. I am actually looking for some help on that subject myself. One option to consider as an alternative would be the Clearpath SDSK Stepper servos. I'm guessing that if you are buying your servos new it would be around the same cost or less. Plus you can get them in Nema 34 size factor so you shouldn't have to change your mounts, and with the higher torques at low RPM's and the 20mm lead of your ballscrews, you might not need any gearing. 20m/minute rapids is (20000 / 25.4) IPM = 787.4 IPM or 20000 / 20 = 1000 RPM, you may be able to even buy some Nema 34's that have better characteristics at 1000 rpm than the ones you currently have. What control software are you using? Mach 3 only lets you use one value for acceleration, but there are others, I forget which ones, that allow you to use two, one for cutting and one for rapids, and that could work in your favor.
I need a bit more info if you want a better answer about my opinion on direct drive coupling a 750 W Servo.
2020 ballscrews
What are the approximate lengths?
The 750W servo that you are contemplating, what is the Motor rotational inertia?
70KG X axis...OK.
2.4NM and 7NM (continuous and peak torque, is that correct?)
Off the top of my head, I am thinking that either option (400W, 3:1 Belt, or 750W direct couple should be fine) and that the steppers you currently have, the torque falls off very quickly with RPM....something that is common with many high torque steppers....actually many steppers with a lower holding torque will provide much better torque at increased RPM, which is where you need it to decelerate the machine if you want fast rapids and good acceleration.
-
Re: Delta servo drives and servos.
Yaskawa has an app that you enter in all the data and it tells you what motor combinations work. You can then approximate for other brands that have similar specs. Bit easier and less prone to errors than doing all the math by hand. :)
-
1 Attachment(s)
Re: Delta servo drives and servos.
Attachment 443630
The picture file didn't attach for some reason in my last post. Here it is. You'll need to save it and zoom in on it to see any details.
-
Re: Delta servo drives and servos.
I used the Yaskawa app to size motors for my router which is an ultra heavy version of the xzero predator. 2525 screws (though, not very long), and a 130kg "x" axis moving mass. I wound up with an 86mm sanyo 500w motor (I have sanyo stuff already cause of my brother machines). If I was buying new motors, they would be 750w 80mm delta b3's for that axis. the Y axis, with only about 50kg on it got a 400w 76mm motor. In this case, I would probably still go 750w on the new motors. The older sanyo motors are built for high overload dynamics, and have more low end torque than another motor of equal "watts". This are all direct drive, no belt reduction.
Main difference between my machine and his is length, and therefore inertia in the ball screw.
-
Re: Delta servo drives and servos.
Hi,
I was unaware just how significant rotational inertia would be when I bought those ballscrews.
I reasoned the a 115kg mass axis would dominate the acceleration equation....but it does not, in fact it represents only about 10% of the inertia.
I really wanted C5 ballscrews and my budget did not allow for new ballscrews. Thus when I found three 32mm C5 double-nut ballscrews in near new order
with thrust bearings all for $1000 including shipping I jumped. I would have preferred 25mm but 32mm was what was available.
It was only later that I realised just how sensitive acceleration is to ballscrew diameter.
J=m .r2/2
m=mass in kg
r= ballscrew radius in m
But note the ballscrew mass is:
m=PI. r2.l.(density of steel)
l=length of ballscrew
J= PI/2 .r4.l. (density of steel)
So rotational inertia is proportional to the FOURTH power of radius.
Whereas rotational inertia is proportional to the square of the ballscrew pitch.
The upshot here is that ballscrew diameter (equivalently radius) is vastly more important than ballscrew pitch.
For example, if my ballscrews were the same length but 25mm rather than 32mm:
J= 5.252 x (25/32)4
=1.956 x 10-4 kg.m2
which would reduce the total overall rotational inertia by 46%.
I have just got the cast iron axis beds from the foundry. I have supplied the drawings to the a company to rough machine them prior to
anneal heat treatment, and then they will machine to final size.
I am still working on the pattern for the frame...its looking to be about 300kg of SG cast iron, it won't be cheap, I'm guessing molding and
pouring will be over $3000NZD.
Craig