Servo Motor Drive

[SignTorch]

I used ordinary gecko servo drives - https://www.geckodrive.com/products/g320x-digital-servo-drive - two required



The servo drive controls high power voltage going to the motor, based on step and direction signals from the BOB, much like a stepper motor.

Unlike a stepper, servos rotate continuously and have an optical encoder to track rotation, which the servo drive monitors to move a certain distance per step, and to hold position when idle.

One problem with a servo is, when it gets a step signal, it basically takes off at full speed, and depends on the encoder signal to tell it when to stop when the step is complete. If the encoder isn't working, the motor doesn't stop, it runs away at full speed. So we have to be careful during hookup and testing, be ready to kill the power if that happens, and make sure the encoder power source and wiring is reliable.

The encoders are high resolution with 500 lines which produces 2,000 pulses per revolution. Which makes for a high frequency signal at high speed, so the encoder wiring must be shielded and no longer than necessary.

Servos outperform steppers because they are free spinning DC motors and can turn at several thousand RPM with no loss of power at low or high speed.

Servo drives have tuning parameters to control how abruptly they attempt to correct differences between the actual and commanded position.

Getting a servo drive hooked up, tested, tuned, and working is a careful, methodical, step by step process which must be worked through. Unlike a stepper you can't just hook up an arbitrary motor to an arbitrary drive and expect it to work right off the bat.

 

[James Taylor]

I am in the process of embarking on a similar adventure (converting my PlasmaCam Z-98 to Mach3) and was wondering if you had any thoughts in regards to swapping out the servos for stepper motors? I've got some gecko stepper motor drivers and it is cheaper for me to purchase the equivalent stepper motor than it is to purchase a gecko servo driver.

 

[SignTorch]

I am in the process of embarking on a similar adventure (converting my PlasmaCam Z-98 to Mach3) and was wondering if you had any thoughts in regards to swapping out the servos for stepper motors? I've got some gecko stepper motor drivers and it is cheaper for me to purchase the equivalent stepper motor than it is to purchase a gecko servo driver.

that would be entirely feasible

but you'd be sacrificing a lot of performance for what can't be enough savings to justify not using servos - even if the whole stepper drive system was free (even if you sold the servo motors and came out ahead) - it wouldn't make up for what you'd lose - which is high acceleration - which there is nothing more important for accurate plasma cutting

trust me - the servos are worth it - I soon will be posting a video about mach3 constant velocity in which the servos can run 200+ ipm and 100+ acceleration - and it shows what happens as lower and lower acceleration values are used - things go downhill fast -

my Hypertherm 65 cut chart for 14 ga mild steel with 45 amp consumables recommends 270 ipm cut speed

per my video - even at 100 acceleration - if the cut is not long and straight lines - then the machine will never get near 270 ipm - and it will round off corners a lot - because that acceleration is just too low for that speed - stepper motors simply can't accelerate fast enough for plasma cutting thin material

that video should get posted soon - meanwhile take a look at this video and this little tiny servo I used to replace the stepper motor on my Z axis - it runs the z axis at 400 ipm and 200 acceleration with a torch and router - that's easily worth the $120 cost of a gecko servo drive - it's a night and day difference compared to the stepper motor that it replaced - I wouldn't want to see what size of stepper motor it would take to do that -

not to mention how servos don't lose steps - I forget how low acceleration was on the original Z stepper motor but it was paltry and it'd still lose steps at times

and when it comes time to add automatic height control - the servos will excel at that to.....

the only way I'd agree that it's a good idea is if you were trading me them servos for some equivalent stepper motors - I'd be the one getting a good deal out of it - you'd be taking a huge step backwards - if that's what you want then let's do it ...

 

[James Taylor]

Excellent advice/points. I'm glad I asked. Thank you for your thoughts.

 

[SignTorch]

any time, let me know if you need any help

what keeps giving me the most trouble is the plasmacam wiring harness - it was a lot of trouble to connect everything to one plug and the connector to the carriage keeps loosing connection ever once in a while -

and your connector would (probably) have a different pin out

but once you get past bench testing the servos you'll be elated

 

[James Taylor]

Awesome, thank you! That would be extremely valuable to pick you brain on occasion and get your input during this consversion "adventure". One thing that I would be interested in getting your input on is the Z-Axis. The Z-Axis on the 98-Z is complete garbage. I am debating whether to do a ground up redesign or whether to look at he Z-Axis carriages on the newer plasmacam models (DHC, DHC2, etc) and base the design off of those. What are your thoughts? I would love to be able to get some close up detailed photos the Z-axis and X-Axis carriage of these various machines.

Thanks again for your help and for taking the time to document your process of migrating to Mach3.

- James

Oh, on a side note, I've got 2 Gecko 320x drivers that will arrive tomorrow. I've got to sell some machinery before I order the rest of the hardware. I recently retrofitted my Vertical Mill with Mach3/smoothstepper/gecko drivers/etc so my budget for machinery improvement is lacking to say the least .

 

[SignTorch]

sorry I didn't respond sooner there was glitch preventing me from seeing new messages

sure - I'm glad to help -

I would replace the Z - mine didn't quite have 2" travel - for cutting objects like frying pans and such that was a little too short for me

here's a quick vid of my latest Z axis - it doesn't take much - this little servo can go 400 ipm with 250 acceleration

stay away from anything heavy - and I am not too fond of direct drive or lead screws - should you ever add auto height control then you want it to be as fast and nimble as possible - lead screws are very inefficient and even with this large servo it had more limited speed and acceleration than the above - this design totally overloaded the gantry - and the screw noise was disturbing

 

[Thick8]

Hi, This tutorial is great. I am using a PlasmaCAM DHC1 table that I got for free. The controller box is non-functional and I don't have the software. My intension is to convert it to a CNC router so speed and acceleration shouldn't be a factor in the retrofit. Would you still recommend using the servos or would I be better served (cost wise) to swap over to steppers? My intension is to use a Windows10 laptop to run it via USB.

 

[SignTorch]

the ideal speed (feed rate) depends on the tool and material - going too slow is not good for the bit

highest possible acceleration is always best - it speeds up the job time and maintains more constant speed for more uniform cut quality

I'd never recommend steppers over servos - you'd need a heck of a stepper to compete with a scrawny cheap servo like on my Z axis in the video above - I got those on Ebay for a few bucks

the plasmacam gantry is not well suited for routing - everything is supported by small springs - check my youtube channel - I made a video about that

 

[Thick8]

I was thinking I could use a screw and jam nut to replace the springs. I haven't taken a real close look yet to determine how to do it yet. I will also be replacing those rollers with proper bearings. I may just clean the table up real good and sell it to fund a router table. I think I found the issue that made them believe the controller was bad. Now the pause light blinks as it should when plugged in. Is there any way to verify operation without a computer hooked up? Maybe a series of button presses that will put it in a function test state.

 

[SignTorch]

the plasmacam X and Y axis is supported and aligned with some small roller bearings rolling on painted steel tube AND the drive gear rolling on the gear rack - neither of those surfaces are smooth and both wear down here and there unevenly becoming less smooth as usage and paint peeling and rust and dust accumulate - the spring is necessary to maintain even tension while rolling on those uneven surfaces - you can't replace the spring with a screw because it will bind or be too loose here and there - you can't use a much stronger spring because it will cut through the paint and wear into the steel and wear out the gear and rack faster - then adding a router on there adds weight which also rides on those springs which effectively reduces the spring tension on the bearings - so, the heavier the tool then the less rigid the gantry - the only way to improve it for routing is to replace it - I'd just use it for plasma cutting

standard repair procedure is if it don't work that's too bad you'll have to pay - my correction procedure would be to press the red button as hard as you can 5 times with a 20 pound sledge hammer