PDA

View Full Version : To big steppers?



JW1977
14-02-2016, 09:46 AM
Currently reading in about steppers and i wonder, as long as the driver are matched well with the stepper, are there any downsides by "overdoing" it. By chosing a to powerful motor?

Read somewhere that optimal is best and i get that current flow throw the motor when it stands still creating holding torque. More torque = more current, heat and so on but what so bad about that?

JAZZCNC
14-02-2016, 10:36 AM
Provided every thing in the chain matches then no there nothing wrong with having a motor that's more powerful than your needs. . . . However.?

Larger motors commonly have much Higher inductance which means they require high voltage to get speed from them. Even then they struggle to reach the same speeds smaller size motors can with good drives and voltage. The chain doesn't just stop at the drives.?
To get the same or equal performance that of little stepper then Large stepper based system will need to consider linear motion ie: screw pitch. So It's a very common mistake for people to go out and buy large motor thinking BIG is better only to be sadly disapointed with the performance because they didn't match the rest of the system.

Then you have added costs.? . . . The voltage required to get good performance from large nema 34 steppers is much higher than smaller nema 23 steppers require and typicly the drives run at mains voltage. This makes them much more expensive. The motors them selfs are also much more expensive, typicly 2 or 3x the price of nema 23.!!

Regards Heat then yes most heat is generated when stood still and heat reduces the life of the motor due to Iron losses etc. But most good quality drives will provide feature to deal with this by dropping current to 50-60% while at Zero RPM. This is called current recirculation, you'll often hear steppers hissing or ticking when stood still which is the Recirculation circuit doing it's job.

JW1977
14-02-2016, 01:03 PM
Provided every thing in the chain matches then no there nothing wrong with having a motor that's more powerful than your needs. . . . However.?

Larger motors commonly have much Higher inductance which means they require high voltage to get speed from them. Even then they struggle to reach the same speeds smaller size motors can with good drives and voltage. The chain doesn't just stop at the drives.?
To get the same or equal performance that of little stepper then Large stepper based system will need to consider linear motion ie: screw pitch. So It's a very common mistake for people to go out and buy large motor thinking BIG is better only to be sadly disapointed with the performance because they didn't match the rest of the system.

Then you have added costs.? . . . The voltage required to get good performance from large nema 34 steppers is much higher than smaller nema 23 steppers require and typicly the drives run at mains voltage. This makes them much more expensive. The motors them selfs are also much more expensive, typicly 2 or 3x the price of nema 23.!!

Regards Heat then yes most heat is generated when stood still and heat reduces the life of the motor due to Iron losses etc. But most good quality drives will provide feature to deal with this by dropping current to 50-60% while at Zero RPM. This is called current recirculation, you'll often hear steppers hissing or ticking when stood still which is the Recirculation circuit doing it's job.

Ok, but that makes things a bit easier.
Ive allready decided to get 3 of the more decent Leadshine drivers like the 882's or something like that. As i Am thinking they will be one of the parameters limiting what steppers to choose. I will also use the IP-M btw.

I read a thread (saw your name in it to) where a Guy had compared some motors and calculated there (real) torque at 1000 rpm with parametern like inductance and other things in mind. Offcourse thats a more interesting way to go than just read the torque figurers given on the motor Specs.

I dont really have any references yet, what is much torque for my gantry?
Machine size is about 600x900 inside the frame. 2 RM1610 screws at X and 1605 at Y and Z. The x axes will be slaved by belt and a larger motor.
I want to be able to work with alu as good as possible with this gear. The router is a Makita 700 something...

JAZZCNC
14-02-2016, 02:32 PM
Ok well first the Leadshine AM882 are good drives but even those will limit the performance of Large nema 34 motors. I wouldn't go above 8Nm with Nema34 motor using AM882's and would run them no lower than 70vdc. I would also try to find 8 wire motors so can wire Bi-polar parallel. Most nema 34 are 4 wire and wired series which means you'll need more voltage to get speed and torque. Hence why they use drives with Mains voltage.

Also don't take much notice of Speed/torque curves from most motor data sheets you see unless high end brands because they are meaning less in real world.

Regards working with Aluminium then Makita router isn't going to last long or be very good. Wood routers are far too fast for most cutters and don't have enough torque. The higher cutting forces also kill the bearings quickly.

JW1977
14-02-2016, 05:23 PM
Dont u Think that large 34`s are a bit overkill for my setup?

Btw the whole Point in the article i was referring to was to calculate the torque based upon inductance, wiring and so on, not to just read the sheets.

What i am building is a garage/hobby machine, not industrial. I will probably mostly cut MDF for speakerbuildings and so on but i want to be able to cut alu if i want to. There are many examples out there with these kind of routers cutting alu successfully. I guess you just have to do it right with feeds and speeds and all that.

What i was hoping for was som kind of hint like "you need about that size and torque for your demands".
Thinking about not being so lazy and do the math and see what i need but it would be great anyway with a hint...

Robin Hewitt
14-02-2016, 05:59 PM
If you don't have enough metal to dump the heat you will be concerned about the rotor magnets hitting their Curie temperature or the field coils melting their insulation.
If the motor shaft dumps it's heat in to the screw, the screw will expand. If the shaft does not dump it's heat in to the screw, where is the heat to go?
You will probably never have a problem, but if your motors run hot you will always worry, especially when considering leaving it unattended for that long job :beer:

JAZZCNC
14-02-2016, 06:13 PM
Dont u Think that large 34`s are a bit overkill for my setup?

If you are running 2 screws connected together with belts which you said you was then no it's not at all.

You like others think that because Mdf is soft it's easy to cut, which it is. But what you don't realise is that while easy to cut it requires feed rates which are fast so you require certain amount of speed from steppers. Now because steppers lose torque when speed rises and you have one motor turning 2 screws along with drag etc from belts then you'll need the extra torque 34's provide.

I'm not guessing on this I build these size and type of routers all the time and I can tell you without doubt that 3 or 4Nm nema 23 cannot handle 2 screws connected with belts on well built machine and give you feed rates enough to cut MDF properly and reliably without chance of lossing steps under load.

Cutting Ali is easy enough provided the machine is strong enough. It's achieving the correct speeds for woods, mdf which catches people out. If you don't have strong enough motors then you'll have to lower feeds so don't stall or miss steps. If you do this too much and can't reach correct feed rates required for cutting then tool wear and finish suffer.
Really has nothing to do with Hobby or industrial usage. It's just case of selecting the correct motors and components so they work together and give balanced machine that will give feed rates required for range of materials you intend to cut.

If you want a hint to which motors then here's one.?. . . . Square ones.!! . . . . . Seriously how can we give you a clue without you give us clue to how machine is built.!. . . Is it built from Tissue or Cast iron.? Does it's use Skate bearings or levitation for linear motion.? So many factors come into play that without a clue it would be wrong to recommend sizes etc.

JW1977
16-02-2016, 08:04 PM
I havent assembled the gantry yet because there are some decisions left before i can do that. However, its based on 2pcs of 80x80 alu extrusion profiles with TBR 20 supported rails bolted directly on them.
They will be held by two (dont know the English term) "short sides" in 10 mm alu. The ballscrews are 2 pcs of RM1610 for X and RM1605 for Y and Z.
The moving piece is also based upon a 80x80 with two TBR 20 supported rails bolted on it. That should give a good idea about friction and the masses that should be moved.

Right now i have to decide one of the following Before i can go on and calculate the steppers,
1. IP-M with 4 steppers. (- for the gantry geometry correction)
2. IP-M with 3 steppers and a belt for the "slave drive". (- for the belt drive but + for no problem with geo. corr. and only 3 motors/drives)
3. Smoothstepper, Pokey or simular with a good breakoutboard, 4 steppers and no problem with the gantry geometry correction. ( - Quality compared to IP-M? Do need a good BOB)

JAZZCNC
16-02-2016, 08:29 PM
Right now i have to decide one of the following Before i can go on and calculate the steppers,
1. IP-M with 4 steppers. (- for the gantry geometry correction)
2. IP-M with 3 steppers and a belt for the "slave drive". (- for the belt drive but + for no problem with geo. corr. and only 3 motors/drives)
3. Smoothstepper, Pokey or simular with a good breakoutboard, 4 steppers and no problem with the gantry geometry correction. ( - Quality compared to IP-M? Do need a good BOB)

Why do you think #2 is (-) using belts.? It's not a bad thing it's a good thing. Belts are not a problem regards accuracy and they keep the screws perfectly sync together. This shouldn't be under estimated and I'd much rather have belts linking screws than using two motors slaved together.
Where it becomes cloudy is if the machine is wide. Then belts get messy and awkward so not always ideal.

JW1977
16-02-2016, 10:35 PM
When im building i always comparing options. Which is best but what compromises must i make? At first glance, no belt seems like the better option. Its one "link" less in the drive. There may be pros also like with the sync you mentioned. I would like to use the IP-M because you have managed to convince me. If it werent for the gantry squaring it would have been clear.
You and others say that the "hard correction procedure" work well so im leaning towards that or the belt drive...
The other options arent so cheap after all since they demand a decent BOB to work well...
Btw the belt would be roughly about 1400 mm total...

njhussey
16-02-2016, 11:22 PM
Machine size is about 600x900 inside the frame. 2 RM1610 screws at X and 1605 at Y and Z. The x axes will be slaved by belt and a larger motor.
I want to be able to work with alu as good as possible with this gear. The router is a Makita 700 something...

My router has an inside area of 970 by 500 and I use an 8Nm Nema 34 driving twin 1610 ballscrews via a belt (I'm using a 230V driver not an AM882 to get the speed) in my X, Nema 23s on Y and Z with Y having a 1610 and Z having a 1605 ballscrew driven by AM882 drivers...it munches MDF and I have it at 7500 m/min plus for cutting plastic such as Acetal....I also use the IP-M and its a great piece of kit.

Your belt will be more like 2m to 2.5m long with a cutting area you're looking for.

Clive S
16-02-2016, 11:42 PM
You and others say that the "hard correction procedure" work well so im leaning towards that or the belt drive... I would go with the single belt rather than the solid stops It is a tried and tested route with many people using it. If you can afford it go with the mains 230V drive and the nema 34 for the twin screws. The AM882 drives are good they also have stall detection. Make sure you get the 8 wire low inductance motors for the Y and Z axis

Good luck with the build

JW1977
17-02-2016, 09:06 AM
My router has an inside area of 970 by 500 and I use an 8Nm Nema 34 driving twin 1610 ballscrews via a belt (I'm using a 230V driver not an AM882 to get the speed) in my X, Nema 23s on Y and Z with Y having a 1610 and Z having a 1605 ballscrew driven by AM882 drivers...it munches MDF and I have it at 7500 m/min plus for cutting plastic such as Acetal....I also use the IP-M and its a great piece of kit.

Your belt will be more like 2m to 2.5m long with a cutting area you're looking for.

Thanks! I went through your build log, nice work! :applause:
I’m also into RC and are really looking forward to get the machine ready.
After all the input from you and others here I will probably go for the IP-M and I will slave the X by belt.
About the belt and pulleys . Can you recommend any products? If I remember correctly the distance between the centers of my ball screws is 684 mm at X.
I was figuring twice that + a little for the pulleys and some kind of tension adjustment device at center.

JW1977
17-02-2016, 09:09 AM
I would go with the single belt rather than the solid stops It is a tried and tested route with many people using it. If you can afford it go with the mains 230V drive and the nema 34 for the twin screws. The AM882 drives are good they also have stall detection. Make sure you get the 8 wire low inductance motors for the Y and Z axis

Good luck with the build

Im gonna do do that!
I havent studied the drivers that much yet but i mentioned the AM882 as an example of a "better" driver.
I will definetly go for Mains 230V.

Thanks! :smile:

JAZZCNC
17-02-2016, 10:58 AM
Can you recommend any products? If I remember correctly the distance between the centers of my ball screws is 684 mm at X.
I was figuring twice that + a little for the pulleys and some kind of tension adjustment device at center.

It will depend on how you setup the tensioning system and idlers. But with 20 teeth pulleys which is good size because gives nice balance of teeth engaged and pulley size which affects inertia of the screw and can have big affect on power requirements. Arranged something like whats in the picture then your belt length will be in the region of 1600mm.

17613

JW1977
17-02-2016, 11:16 AM
It will depend on how you setup the tensioning system and idlers. But with 20 teeth pulleys which is good size because gives nice balance of teeth engaged and pulley size which affects inertia of the screw and can have big affect on power requirements. Arranged something like whats in the picture then your belt length will be in the region of 1600mm.

17613

Yes, something like that.
I liked the principle of this one...
17614

JAZZCNC
17-02-2016, 11:32 AM
Yes that way works ok and I've used similar setups my self in past but ideally you want to reduce the amount of idlers to minimum. So if you have no restrictions to where motor goes etc then other way is slightly better. Less friction, less to go wrong.

JW1977
17-02-2016, 11:48 AM
Yes that way works ok and I've used similar setups my self in past but ideally you want to reduce the amount of idlers to minimum. So if you have no restrictions to where motor goes etc then other way is slightly better. Less friction, less to go wrong.

Yes, you’re right. One pulley is a better solution… But the symmetrical one looks nicer… :shame:

JAZZCNC
17-02-2016, 11:59 AM
Yes, you’re right. One pulley is a better solution… But the symmetrical one looks nicer… :shame:

Both look the same after you put cover over them to protect . . :hopelessness:

Sven
25-09-2016, 03:12 PM
My router has an inside area of 970 by 500 and I use an 8Nm Nema 34 driving twin 1610 ballscrews via a belt (I'm using a 230V driver not an AM882 to get the speed) in my X, Nema 23s on Y and Z with Y having a 1610 and Z having a 1605 ballscrew driven by AM882 drivers...it munches MDF and I have it at 7500 m/min plus for cutting plastic such as Acetal....I also use the IP-M and its a great piece of kit.

Your belt will be more like 2m to 2.5m long with a cutting area you're looking for.

I'm building a 1250mm sq router planned to run at 10m a minute, using 2 nm steppers....
Calculations suggested 8 nm would get me 12m a minute or so.

Direct drive, 1610 screws, 2 motor/screws for the gantry.

JAZZCNC
25-09-2016, 03:29 PM
I'm building a 1250mm sq router planned to run at 10m a minute, using 2 nm steppers....
Calculations suggested 8 nm would get me 12m a minute or so.

Direct drive, 1610 screws, 2 motor/screws for the gantry.

Don't need 8Nm for machine that size if slaved motors 3 or 4nm nema 23 will be more than enough provided the gantry is less than 80-90Kg and your using ballscrews. If used with good motion control card you'll get 12nm or more.

I wouldn't use 2Nm because while they spin faster there will be very little torque left at the Mid/higher end of speed curve which is were you want it while cutting.

8nm would mean Nema 34 which would actually run slower with the same amount of volts. So you'd need more voltage which means larger drives and PSU. Really Large steppers need high power drives that run at mains voltage for best performance.

Sven
25-09-2016, 04:18 PM
Don't need 8Nm for machine that size if slaved motors 3 or 4nm nema 23 will be more than enough provided the gantry is less than 80-90Kg and your using ballscrews. If used with good motion control card you'll get 12nm or more.

I wouldn't use 2Nm because while they spin faster there will be very little torque left at the Mid/higher end of speed curve which is were you want it while cutting.

8nm would mean Nema 34 which would actually run slower with the same amount of volts. So you'd need more voltage which means larger drives and PSU. Really Large steppers need high power drives that run at mains voltage for best performance.

At 10800 mm/minute (is 10.8 m/minute with 10 mm pitch) the 2 Nm had 0,8 nm of torque left, which is enough for my machine.
The cutting force was set at 100N in these calculations.

I cant run higher speeds on a ball screw of the length I am using.
Steppers with higher torque had a torque curve that plummeted as soon as they start rotating, the 2 and 3 nm motors were comparatively flat.

JAZZCNC
25-09-2016, 07:49 PM
At 10800 mm/minute (is 10.8 m/minute with 10 mm pitch) the 2 Nm had 0,8 nm of torque left, which is enough for my machine.
The cutting force was set at 100N in these calculations.

I cant run higher speeds on a ball screw of the length I am using.
Steppers with higher torque had a torque curve that plummeted as soon as they start rotating, the 2 and 3 nm motors were comparatively flat.

Ok if you don't want to listen to experience and would rather go with calculations then who am I to argue.:cower:

Sven
05-02-2017, 02:14 PM
I thought I'd post a video of the machine during it's acceleration tests today.

Now I'm unsure of what Jazz means but here is the machine running at 0,4 G with two 2 Nm steppers and 1610 ballscrews:

https://www.youtube.com/watch?v=1RzP1VLtK9I

Topspeed is 10200 m/min

JAZZCNC
05-02-2017, 03:02 PM
I thought I'd post a video of the machine during it's acceleration tests today.

Now I'm unsure of what Jazz means but here is the machine running at 0,4 G with two 2 Nm steppers and 1610 ballscrews:

Topspeed is 10200 m/min

What I said was Don't require 8Nm motors which is still true. Also that 2Nm will struggle while cutting at higher feeds. Cutting air is very different to cutting the hard stuff so suggest you stick cutter in and see how well it handles cutting with decent DOC at 8000mm/min.! (Which is MDF like feeds)

Sven
05-02-2017, 03:13 PM
Ah, I thought it might be in that direction.
I'll test and let you know!

Robin Hewitt
05-02-2017, 03:42 PM
Topspeed is 10200 m/min

Blimey, that is 380mph :onthego:

Sven
05-02-2017, 10:00 PM
Whoops, well spotted!
10200mm/minute.