[louieatienza]

boom.

those parkers make all the difference in speed. mmmm. what settings are you running? mach/emc unfortunately cant push them to their full potential of 256 microstep at 3000rpm.

Not pushing them right now, making sure everything's moving first before I tweak the speed. Right now, 192IPM rapids, accel set to 50in/s/s, 5x microstep... I'm not a big advocate of high microstep resolutions for a few reasons. I have a mechanical resolution of less than .0008", no microstepping. So 5x is probably overkill. I'm thinking half stepping would be sufficient, and give me better overall torque.

I should add the EdingCNC software defines (rapids) speed in in/s not in/min, so that would be 3.2in/s as set in the EdingCNC software...

[Tkamsker]

i really like the sound reminds me on some sifi movie - nice lets see how it is milling question are the x and z rails 15m 20 or 25 mm ?

[Tkamsker]

mach/emc if yu run them at the lpt you mean ? i think i remember some 33khz limit but if you use PCI card it should have no problem with it
I am right or mislead ?

[louieatienza]

the max throughput of the EPP parallel port is about 2MB/s, so the port itself is not the bottleneck. The bottleneck is likely the software and software control of the CPU clock Moving the pulse generation to an external controller should give you better speeds.

Yes, the sound of a ground ballscrew whirling is a bit more pleasant than hearing a rolled ballscrew.

[louieatienza]

i really like the sound reminds me on some sifi movie - nice lets see how it is milling question are the x and z rails 15m 20 or 25 mm ?

Good question, don't have a clue.... Have to measure them.

[ihavenofish]

high microstepping is for smoothness, not precision. less resonance helps surface finishes, and it helps lessen the likelihood of stalling the drive in an aggressive move. when you have some time, run through the jumper settings. 8 id say is the minumum youd want. 16 is likely your compromise for speed with your controller. but as you go higher, youll notice something. by the time you are at 256, the motor will be nearly silent.

those motors have flat torque right up to about 400ipm at 2000rpm. at 16 microsteps thats 3200 steps per turn. 33 turns per second. so 105khz. thats the upper limit of mach3. a little higher than emc. i dont knoow what the control you are using is capable of.

[ihavenofish]

mach/emc if yu run them at the lpt you mean ? i think i remember some 33khz limit but if you use PCI card it should have no problem with it
I am right or mislead ?

mach3 artificial limit is 100khz. this limit is based on the reliability of the system. a brand new 4ghz box with ample ram would likely manage much higher than that, but the limit is hardcoded into mach3.



- - - Updated - - -

i really like the sound reminds me on some sifi movie - nice lets see how it is milling question are the x and z rails 15m 20 or 25 mm ?

15mm

[louieatienza]

high microstepping is for smoothness, not precision. less resonance helps surface finishes, and it helps lessen the likelihood of stalling the drive in an aggressive move. when you have some time, run through the jumper settings. 8 id say is the minumum youd want. 16 is likely your compromise for speed with your controller. but as you go higher, youll notice something. by the time you are at 256, the motor will be nearly silent.

those motors have flat torque right up to about 400ipm at 2000rpm. at 16 microsteps thats 3200 steps per turn. 33 turns per second. so 105khz. thats the upper limit of mach3. a little higher than emc. i dont knoow what the control you are using is capable of.

The max step frequency of the EdingCNC card is 125kHz.

Eding CNC - PC based CNC control

I will run tests with different microstep resolutions. I understand that it leads to smoother operation and not necessarily accuracy (I've advocated these points in the past.) The reason I hesitate getting too carried away with that is it leads to false readings in the computer DRO. Since the stepper has to overcome not just the weight of the axis, friction of the bearings and screw, and the force needed to cut through whatever material, it also has to overcome its own detent torque. You will reach a microstep where this combined torque needed exceeds the torque at a particular microstep. Therefore, the controller will send pulses to the drive, but the motor never moves, until it reaches a step where the torque of the motor does overcome that of the total load.

[ihavenofish]

because the torque of these motors are flat to 2krpm and have generally speaking, way more torque than the little mill needs, you shouldnt have such an issue. if you were running to the edge of usability, then youd have some trouble im sure.

393ipm, 16 microsteps, 100i/s2 acceleration. thats where id aim with this little guy. i have these drives too of course, and have done the math on them several times. they are possible the most impressive nema23 setup you can find (too bad they cost a fortune new and parker service and distribution sucks hard)


PS: i hope that desk is sturdy, haha

[louieatienza]

because the torque of these motors are flat to 2krpm and have generally speaking, way more torque than the little mill needs, you shouldnt have such an issue. if you were running to the edge of usability, then youd have some trouble im sure.

393ipm, 16 microsteps, 100i/s2 acceleration. thats where id aim with this little guy. i have these drives too of course, and have done the math on them several times. they are possible the most impressive nema23 setup you can find (too bad they cost a fortune new and parker service and distribution sucks hard)


PS: i hope that desk is sturdy, haha

It's moving incrementally that problems occur. At half step you have about 70% of the torque available at that point. At 1/4 step it drops to less than 40%. At 1/8 step less than 20%. And at 1/16 step less than 10%. This is just a fact of the way the stepper motor works. It's like a spring, for example every 1/8 and 7/8 step wants to pull to the nearest full step, for example.

On my chart it shows about 200in-oz available up to 30rps, or 1800rpm which gives me about 283ipm (and about 1200in-oz available torque)... Drops past there but I probably wouldn't operate past that corner speed anyway (remember, these are custom 4mm lead GROUND ballscrews, not the standard 5mm rolled screws that comes stock.)

Yes the desk is temporary... I have a nice cabinet I'm using for a base that I need to fit a pan to. I have one more Parker drive for my temp rotary axis. When I'm ready I'm switching all these motors to servos and using these motors on my router table. I have a nice Parker HV-342 as well.... Beast of a stepper...

[louieatienza]

I should note I also have some Kollmorgen CTP-22 steppers, one which drives the Z axis. I have all three axes tuned the same, a testament to the engineering of the Z assembly. I don't understand how Parker's drives outperforms Kollmorgen's. They have a drive that outputs 320VDC and the torque curves are nowhere near Parker''s... and Parker's design is about 10 years young.

There are some new steppers I'm looking at from Lin Engineering, which are the same size as these Parkers but rated at 600in-oz stall torque. I think these Parkers are made by Lin looking at the way they're made, so it would be just a matter of funding the right match that will work with these drives... They make them with different current and inductance ratings as well.

[ihavenofish]

the parkers are indeed lin as far as i know. usa made. same as gecko (except those are low voltage). lin will wind you a motor any way you like. basically go in and tell them the drive spec and say "i want more low end, but still 150 oz at 2000rpm and they will find the best match in a winding.

i and going to get something in nema34 from lin if its not crazy expensive to match my parker e-ac's.

i didnt know it was 4mm screws. that makes sense then yeah. 283ipm is still a decent target for a cutting speed. you could do rapids at full throttle (360ipm, 2340krpm or 125khz) where it doesnt use near the max torque. i wouldnt worry about the microstepping. all this is happening in the sub .001" range.

you may have also noted parkers torque curves are for microstepping (not sure which level, probably 1/4 step). they full step rating is also provided. the motors are also designed with low detent torque to minimise the loss of microstepping torque. you also have dip switches for waveform control, which all give different performance trades. in other words, they designed these as a paired drive and motor set and its not likely youll see any drifting more than a few tenths in practice at 16 microsteps.

[louieatienza]

the parkers are indeed lin as far as i know. usa made. same as gecko (except those are low voltage). lin will wind you a motor any way you like. basically go in and tell them the drive spec and say "i want more low end, but still 150 oz at 2000rpm and they will find the best match in a winding.

i and going to get something in nema34 from lin if its not crazy expensive to match my parker e-ac's.

i didnt know it was 4mm screws. that makes sense then yeah. 283ipm is still a decent target for a cutting speed. you could do rapids at full throttle (360ipm, 2340krpm or 125khz) where it doesnt use near the max torque. i wouldnt worry about the microstepping. all this is happening in the sub .001" range.

you may have also noted parkers torque curves are for microstepping (not sure which level, probably 1/4 step). they full step rating is also provided. the motors are also designed with low detent torque to minimise the loss of microstepping torque. you also have dip switches for waveform control, which all give different performance trades. in other words, they designed these as a paired drive and motor set and its not likely youll see any drifting more than a few tenths in practice at 16 microsteps.

Yes very high quality from Lin.... They do have them standard with different specs, so I think it's just a matter of finding the right current and inductance ratings that fit the drive specs. I'd probably have them make it as 8-wire so I can get more top end torque.

While the HV steppers were designed with the E-AC in mind, they also pair with their VS and OS lines (I believe made by Pacific Scientific.) Not that you need to necessarily pair Parker's steppers with this drive. The reason is that this way Parker has tested and can state the performances of each pairing, which makes it easier from an engineering standpoint. You can pretty much use any motor you wish but they won't "guarantee" its performance or have any graphical data to show.

Check out Sanyo Denki, Oriental Motor and Leadshine, they also have offerings that will pair with high voltage drives.

[ihavenofish]

yeah ive looked through everything. lin is surprisingly (or not) actually the cheapest and top notch as well. i have my non specified novakon nema 34's on the drive already. the drive runs them, but not in an ideal manner (series, with alot of the low end torque clipped due to the 3.5 amp rating). ill use those to set everything up, and when its all working well, ill get lin to make me up some ideal motors.

theres a funny thing with these though. the hv233 actually has 200w at 3krpm. the hv342 is a mere 127w. unfortunately i cant do a belt reduction, otherwise the 233's would be perhaps a better choice.

anyhow, it will be another month before thats all in place. in the meantime i can drool over your machine instead!

[LeeWay]

Do you have Novakon's AC steppers?
If so, those are some fine steppers when coupled with some AC Leadshine drives.

[ihavenofish]

i have parker e-ac stepper drives, but only the stock nm145 motors.

[louieatienza]

i have parker e-ac stepper drives, but only the stock nm145 motors.

You should be able to run them with the Parker drives though would be nice to test. If you're interested Ihave a HV-342....

Sort missed that last post.... All the Parkwr steppers I do have are indeed 8-wire.

[ihavenofish]

the novakons are already running on the drive

they are just not optimal is all - though i have NOT attached them to the machine yet. it may turn out they work rather fine. another reason to replace though is seals... they got a bit of the heat treatment and are a tad grindy :x



oh, the hv342/1 is about $100 more than an equivelant lin motor. weather thats just markup, or those motors are 'better", i dont know.

[louieatienza]

the novakons are already running on the drive

they are just not optimal is all - though i have NOT attached them to the machine yet. it may turn out they work rather fine. another reason to replace though is seals... they got a bit of the heat treatment and are a tad grindy :x



oh, the hv342/1 is about $100 more than an equivelant lin motor. weather thats just markup, or those motors are 'better", i dont know.

Honestly I think you just pay for the fact that Parker ran their performance data with the stepper and they certify that performance for that particular pairing. I have the same settings on all 3 axes despite running the Kollmorgen on the Z axis.

[Tkamsker]

15 mm ? in x and y. is z axis bigger ? i expected minimum x and y 20 mm in z 25 ,.. but lets see the results when you mill steel ,..


Gesendet von iPad mit Tapatalk