[jossa]

you are right, my thought is bigger is better...
i am going to move the big and heavy head of the machine with that stepper and my logic was that i "need what i can get"

maby this could work?
Brand New NEMA34 Hybid Stepper Motor 86BYGH450B-06D (KL34H2110-50-4A) Shaft Size: 14mm Max torque: 920 oz-in | Automation Technology Inc
it has:

Current: 5A
Resistance: 2.9Ω
Inductance: 12mH


so what is reducing the current? just the way you wire the motor or something in the drive itself? ie. dip switches..




PSU: i don't quite follow you here.. isn't power a function of voltage times current?
for my Kstepper 48V is max so should get a 48V SM-PSU with around 1000W? giving me almost 21A

i have read the thread you linked to, but i am not sure what you want me to see. maby that i should get a switch mode regulated PSU.. (thinking of buying it locally in my town so i can get a brand name one and not a generic china one)

[TomKerekes]

Hi jossa,

i am going to move the big and heavy head of the machine with that stepper and my logic was that i "need what i can get"

I think you can lift a heavy load at about the same speed with a smaller size 23 motor than a larger size 34 motor with the proper gearing/mechanical advantage because although the size 23 motor has less torque it can usually run faster.

maby this could work?
Brand New NEMA34 Hybid Stepper Motor 86BYGH450B-06D (KL34H2110-50-4A) Shaft Size: 14mm Max torque: 920 oz-in | Automation Technology Inc
it has:

Current: 5A
Resistance: 2.9Ω
Inductance: 12mH

I think that must be a typo on the resistance. Probably should be 0.29 ohms. 2.9 ohms at 5A would dissipate over 70Watts in heat! The inductance is quite high.

I'm not saying these are bad choices but just want to point out the tradeoffs.

so what is reducing the current? just the way you wire the motor or something in the drive itself? ie. dip switches..

The physics of the motor. Inductance and back emf. All the drive can do is apply the full power supply to try to get the current to flow. A mechanical analogy to voltage, current, and inductance is force, velocity, and mass. Given a fixed force the velocity will ramp slower with a larger mass. Back emf is an effect of the motor similar to a generator. When running at some speed it generates a voltage. The applied voltage must be at this voltage before even begining to get current to flow into the motor. Unfortunately the things that increase torque in a motor also tend to increase inductance and back-emf.

PSU: i don't quite follow you here.. isn't power a function of voltage times current?
for my Kstepper 48V is max so should get a 48V SM-PSU with around 1000W? giving me almost 21A

No it is more of a power/conservation of energy thing. Any chopping stepper drive, especially KSTEP is extremely efficient. Virtually all of the power supply's power goes into heating the motor or doing work. When stopped the motor does no work and the motor just dissipates a few watts in heat. The power supply current may only be a fraction of an amp. It only takes several volts to make full current flow in a stepper when stopped or moving slow. At high speed (under load) the power will increase. But because the torque drops off rapidly with speed it becomes impossible to put more than a certain power into a stepper.

Along this same line when decelerating at high speed negative work is being performed by the motor. This energy is pushed back into the power supply. This can cause the power supply voltage to rise (regenerative braking). In some cases this may damage the supply (or the stepper drive) or cause the supply to fault. I think this is usually why most stepper drive manufacturers recommend a linear supply with huge capacitors to be able to absorb this energy. KSTEP has a voltage clamping feature so this isn't an issue.

HTH
Regards

[jossa]

Hi jossa,


I think you can lift a heavy load at about the same speed with a smaller size 23 motor than a larger size 34 motor with the proper gearing/mechanical advantage because although the size 23 motor has less torque it can usually run faster.

What you said got me thinking..

If i use a 570oz stepper on my Z axiz, with the gearing the CNC fusion kit uses 1,23:1 that gives me 701oz with a nema 23 motor @ 5A

That should be a match made in china! 700oz stepper on a quality ballscrew with 5mm pitch lifting a head of aprox. 5-60KG on dovetails

For the X and Y axis, NEMA23 425oz/in 2.8A Stepper Motor ¼” Dual shaft @ 1:1 direct drive.
it is a bit bigger then what i have seen others use, but i have the driver for it, and have read that a bigger motor is not that prone to missing steps, but it might be slower to move..
crevice reamer uses 381oz in his optimal design, but he is limited to 3,5A with the G540 drive he uses and he might be thinking economics


Powersuply: SWITCHIG POWER SUPPLY 48V /12.5 AMP
its it big enough for 4 axis? (adding 1 extra for the future)
if i add up max current for the 4 motors i get 13,4A and the PS is 12.5.. just a bit to smal?


Homing: PNP NC inductive switch...


Motor enclosure: nema 23 kit


MPG: MPG4 - 6 Axis Pendant hardwire it to the Kflop, it looks like all the outputs are there




what do you think about this setup?

i know you are not my personal advisor, but you are the first person here that has cared enough to answer me in a way that makes this possible and i appreciate that enormously
it made me so confident that i went ahead and bought the Kflop and the Kmotion and it is in the post now.. =) i will even try to document the setup in my build thread for others to see at a later time.

[TomKerekes]

Hi jossa,

I think the 570oz-in stepper could be a good choice. Although the Torque graph they show is pretty poor. Torque drops dramatically at 400RPM. I'm not sure why. I would have expected it to have usable torque out to 800 or 1000RPM

http://www.automationtechnologiesinc...100-50-4BT.pdf

Here is a very similar motor that a User has reported good results with to 1000RPM

Nema 23 570 oz/in Stepper Motor Model #RS23-570

If I did the math right the theoretical linear force would be 1400lbs or a safety factor of 10 to lift your 60Kg head in Z. I attached an Excel Spreadsheet if you want to check or adjust the numbers. You might want to have a brake or counter balance to stop Z from falling with power off. 400RPM would be 1600mm/min.

Attachment 241242

I would probably use the same motors for X and Y also.

I think the power supply would be adequate (If I did the math right lifting 60Kg at 1600mm/min theoretically only takes 16Watts)

The rest of the setup looks reasonable to me.

HTH
Regards

[jossa]

thanks!

yea there is a big dip in performance around 400, but over that would be for rapids only?
but there seems to be other benefits with the RS 23 steppers making them worth the extra cost..


so you don't see any problems with using the same size steppers on all the axis?
except price maby

[TomKerekes]

Hi jossa,

I don't see any problems with using the same size.

Regards

[jossa]

is there any way of getting the same performance from that setup we have arrived at now.. as you would get in this easy-servo-motors system?

could i add encoders to double shafted steppers maby?


basically what i want to do i maximize the accuracy in my setup..
first thought was to wire inn the DRO scales, but if that feedback don't offer any major benefit is there any other ways?

[TomKerekes]

Hi jossa,

I think those are nearly the performance and cost of servos. I do recommend you get dual shaft motors so you can add encoders. From an accuracy stand point the glass scale feedback would be the best approach. But the accuracy of a system is determined by the combination of all the errors. Eliminating any one may not make a substantial difference.

Regards

[jossa]

is the Kflop able to compensate for lost steps etc using the glass scales?
i remember you saying something about that it would detect it but not compensate for it.. is that possible with encoders on the steppers?

yea i agree with that, and to be honest i don't think my mill is that accurate in the first place..
but i want to learn about the possibilities in this amazing piece of hardware, my plan is to become familiar with it and later on use it to convert an "older" vertical machining center using modern electronics .
and my learning is going to be done on a cheaper SX3 mill =)

[TomKerekes]

Hi jossa,

Miss-stepping is sort of like a gear slipping a tooth. Once the gears re-engage it is possible to drive further to make up for the miss-step (and KFLOP can do this), but before this happens the position will be incorrect. A stepper motor miss-steps by missing a pole which results in 1/50th of a revolution error. Furthermore once a stepper miss-steps it is possible to completely stall which results in very little torque until the motion slows down and allows the rotor to re-synchronize to the phases. For these reasons you can't expect your system to work well even though your steppers are miss-stepping or stalling even with linear scale feedback or rotary encoder feedback.

But we do highly recommend adding encoder feedback. Otherwise the system is basically running blind. The encoder feedback can let you see immediately if you miss-step, stall, crash, or resonate. Linear scales provide more information regarding backlash, friction, and compliance (and can often correct those).

Normal stepper drives like Geckos and KSTEP drive a stepper with constant current and only advance the phase to generate motion. More advanced drives can both control the current and phase. This allows the drive to control a stepper in torque mode just like a brushless motor. Our SnapAmp is capable of doing this. This type of control can prevent miss-steps and stalls in the real sense. But I believe for about the same cost and complexity a servo system will provide better performance.

HTH
Regards

[jossa]

thanks again for clearing things up
i will add encoders to my steppers..

btw. i have powered up my kflop and got the drivers installed on a win 8.1 pc..
it is connected and the program is talking to the board, console confirmed..
but i am kind of lost on what to do next..

do you have a list of things to do?
i assume defining steppers ore something like that? and setting inputs&outputs? ... but to be honest i tried reading the manual but it just made me more confused on where to start..

[TomKerekes]

Hi jossa,

The first step would be to wire up a stepper motor to KSTEP. One coil would connect to the A0 terminals and the other coil would connect to the B0 terminals. You might verify which wires on the stepper are a coil with an ohms meter.

The next step would be to set the current jumper setting for the current level for your motors.

Then connect your stepper motor supply to VBB+ and GND.

See:

KStep Connectors

Let us know how much of this you understand.

Regards
TK

[jossa]

thanks again.. your patience with me is amazing! (and part of the reason i went with the kflop =) )

that takes care of the kstep.. (i don't have the steppers yet, the store that you linked to had a bug in the checkout process so i have to wait to Monday when they see my email.

but what about the Kflop.. do i need to do anything there?

[jossa]

And i have gone trough KFLOP Connectors

i am more thinking of the program running on the pc..

[TomKerekes]

Hi jossa,

Its hard to do much without motors. But since steppers are open loop you could just pretend they are moving

InitKStep3Axis.c should setup 3 KFLOP axis to reasonable settings for 3 motors connected to KSTEP Outputs 0, 1, and 2..

#1 Run KMotion.exe, go to the C Programs Screen, load InitKStep3Axis.c, and Compile/Download/Run it Attachment 241838

#2 on the Config/Flash Screen Import the settings of the 3 C Program Axis into KMotion.exe from KFLOP
Attachment 241840

#3 on the Step Response Screen have Axis#0 selected, enter a move size of 10000, time 3.5secs, and push "Move". Axis #0 motor should move (if it was connected), the generated trajectory will be uploaded, and Plotted. Something like this:
Attachment 241842

Calculate your resolution for all your axes in terms of uSteps per Inch (or mm). KSTEP has 16X microstepping so there are 3200 uSteps/rev. Then factor in your lead screw pitch and any gearing you might have. Then set the Vel, Acceleration, and Jerk to what you think might be reasonable values for your system. The units are all in uSteps so relate to normal units by dividing by the uSteps/unit. So for example if your resolution is 1000 uSteps/mm then moving a velocity of 10000 would be 10000/1000 = 10mm/sec or 600mm/min

Push "Move" again and observe the new trajectory. Also change the Plot type to Velocity, Output vs time to observe the velocity profile. Note because you don't have encoders or servo feedback ignore the red and green plots (or turn them off). All that is useful at this stage is the blue Commanded Destination Plot. Try and understand the various parts of the trajectory: where limited jerk is ramping the Acceleration, where acceleration is increasing the velocity, and constant velocity regions. There are actually 7 "Trip states" when moving from point A to point B ("Move" does a back and forth motion).

#1 ramping (increasing) acceleration
#2 max acceleration
#3 ramping down acceleration
#4 no acceleration (cruise speed)
#5 ramping deceleration
#6 max deceleration
#7 reducing deceleration
#8 Stopped

If you aren't familiar with motion parameters - using a car analogy - Max Velocity would be the speed limit, Max Acceleration would be how far you are allowed to push the accelerator, Max Jerk would be how fast you are allowed to move the accelerator pedal. So for a car anaolgy they would be:

#1 gradually pushing the accelerator down
#2 holding the accelerator down
#3 gradually releasing the accelerator
#4 coasting
#5 gradually applying the brake
#6 holding the brake down
#7 gradually releasing the brake
#8 fully stopped

Experiment changing Velocity, Acceleration, Jerk, and Move sizes to understand the relationships.

Another concept to grasp is that KMotionCNC coordinated motion uses only second order motion (acceleration is applied instantly - infinite Jerk). You can simulate this type of motion by temporarily setting Jerk to a huge number (1000X the Acceleration value will apply the acceleration in 1millisec). Later the Velocity and Acceleration values for infinite Jerk will be entered int the KMotionCNC Trajectory Planner (after conversion to inches).

After determining reasonable Velocity, Acceleration, and Jerk for all your motors Export the Setting back to the C Program. This Flash video may help although it was created before the Import/Export function was added to KMotion.exe. It shows how to use the C code->Clip Board to convert settings for a single axis at a time.
http://www.dynomotion.com/Help/Flash...ers/index.html

Let us know how much of this you are able to do and understand.
HTH
Regards

[jossa]

ok.. i think i understand.

the 1-8 steps is PID controlling the steps and thus shaping the curve.. right?
i know some PID from programing VFD's that run pumps, fans and compressors but i am not an expert.


i am going to buy theDeluxe kit
so my leadscrew has a pitch of 5 mm .. and i am planing on using the steppers you recommended
so how do i calculate the resolution?


this is what my screen looks like.. i have enterd the values tat you said.. why is the a negative dip after the positve one?
Attachment 241878

i have also included numbers on the different things i wonder about..

in number 2 what is the values measured in? steps?
and in number 3: is that volt and amp? if so what is the J? joule?
number 4: ?

yea.. you get the picture.. :P i am unsure about many things here.. also steppers being new to me is not helping.




ooh! i almost forgot.. have i forgotten to by a breakout board for my Kflop?
is there a recommended way of connecting to the I/O's?

[TomKerekes]

Hi jossa,

the 1-8 steps is PID controlling the steps and thus shaping the curve.. right?

Actually no. PID has to do with servo feedback to help keep on the desired trajectory. In your case you have no feedback. The parameters you have marked #1, #2, #4, #5 all have to do with feedback and can be ignored until you add feedback (encoders). The thing we are currently concerned with is the trajectory. Sometime called motion profile. This is sort of like planning a route on a map. It consists of your desired position, speeds, and times along the route on a theoretical basis. If you were to actually do the route then the actual path may differ from the original planned route. And if you somehow knew you were going off the planned route you could make corrections to get back on route.

i am going to buy theDeluxe kit
so my leadscrew has a pitch of 5 mm .. and i am planing on using the steppers you recommended
so how do i calculate the resolution?

I can't find the details in that link. But assuming the lead screw turns one rev for each motor rev then:
5mm = 3200uSteps
or
1mm = 640uSteps

Resolution = 640 uSteps/mm or 16256 uSteps/inch

this is what my screen looks like.. i have enterd the values tat you said.. why is the a negative dip after the positve one?
Attachment 241878

The "Move" test makes a move of +Size and then a move back of -Size. You are plotting velocity. So the plot shows a positive velocity, a stop (zero velocity), and then a negative velocity back to the starting point.

i have also included numbers on the different things i wonder about..

in number 2 what is the values measured in? steps?

Again these will only come into play after you make your system closed loop. But yes they will all be in uSteps for closed loop step/dir mode.

and in number 3: is that volt and amp? if so what is the J? joule?

No. V=Velocity A=Acceleration J=Jerk
See:Step Response Screen

BTW you can push the Help Button then click on the parameter for a description. For example if you see a parameter on a screen you don't know what it is. Do this:

#1 click the "Help" Button on the Screen
#2 click on the parameter you don't know

number 4: ?

Feed forward is an advanced parameter that you shouldn't concern yourself with at this point. See: Step Response Screen

have i forgotten to by a breakout board for my Kflop?
is there a recommended way of connecting to the I/O's?

Winford.com sells screw terminal breakout boards for IDC connectors. As well as ribbon cable and connectors. See:
For KSTEP JP33
2x13 0.1" Header (26 position IDC ribbon connector) Breakout Board with Screw Terminals - Winford Engineering
for KFLOP J4 and J6
2x8 0.1" Header (16 position IDC ribbon connector) Breakout Board with Screw Terminals - Winford Engineering

HTH
Regards

[jossa]

Actually no. PID has to do with servo feedback to help keep on the desired trajectory. In your case you have no feedback. The parameters you have marked #1, #2, #4, #5 all have to do with feedback and can be ignored until you add feedback (encoders). The thing we are currently concerned with is the trajectory. Sometime called motion profile. This is sort of like planning a route on a map. It consists of your desired position, speeds, and times along the route on a theoretical basis. If you were to actually do the route then the actual path may differ from the original planned route. And if you somehow knew you were going off the planned route you could make corrections to get back on route.

i see.. that is when feedback comes into play.

I can't find the details in that link. But assuming the lead screw turns one rev for each motor rev then:
5mm = 3200uSteps
or
1mm = 640uSteps

Resolution = 640 uSteps/mm or 16256 uSteps/inch

so if the pitch is 5mm, then one rev. of the screw is 5mm forward? and one rev of the stepper is 200 steps, how do you get the Usteps?

The "Move" test makes a move of +Size and then a move back of -Size. You are plotting velocity. So the plot shows a positive velocity, a stop (zero velocity), and then a negative velocity back to the starting point.

DOH! :P i see.. my mistake

Again these will only come into play after you make your system closed loop. But yes they will all be in uSteps for closed loop step/dir mode.

No. V=Velocity A=Acceleration J=Jerk
See:Step Response Screen

thanks, it all makes more sense now..

BTW you can push the Help Button then click on the parameter for a description. For example if you see a parameter on a screen you don't know what it is. Do this:


#1 click the "Help" Button on the Screen
#2 click on the parameter you don't know

THANKS! sorry for bugging you all the time, i did not know i could click the picture in help.. nice feature.

Winford.com sells screw ........................ctor) Breakout Board with Screw Terminals - Winford Engineering

a bit pricy but looks like high quality, orderd a full sett of breakout's. thanks alot!

[TomKerekes]

Hi jossa,

so if the pitch is 5mm, then one rev. of the screw is 5mm forward? and one rev of the stepper is 200 steps, how do you get the Usteps?

KSTEP (like most stepper drives) breaks each motor full step into smaller steps called micro-steps to get smoother and finer resolution. KSTEP does 16X microstepping so there are 16x200=3200 microsteps per revolution. uStep is an abbreviation for micro-step.

HTH
Regards

[jossa]

ok, i see..

and according to this page, 16 micro steps is a good place to stop since the torque is dropping drastically after that point. indicating good design and not just chasing numbers


Thanks again so much Tom!
I would not hesitate to recommend your products and your business to anyone searching for a good CNC setup.
The service is A+ and i wish the company's where i live could be 1/5 of what you offer to a stranger on the internet.