Servos instead of steppers for TAIG?

[Smertrios]

I am searching eBay for "TAIG" and IC that one of the sellers of TAIG is elling 225oz servo motors that can be used for replacement of steppers. I REALLY want a closed loop system that won't miss steps either because of too little force, too much force or some combination of the 2 just because the machine has warmed up.

Okay so now to the point... On eBay deepgroove1 is offering 225 oz servos motors for use with a TAIG saying they can be mounted just like steppers for direct drive. I want... at least I think I do but I need U

3 servos http://cgi.ebay.com/dc-servo-motors-...QQcmdZViewItem
3 geckos www.geckodrive.com
6 limit switches ???
1 or 3 power supply http://www.cncresource.com/store/ind...&products_id=7
and of course alot of wiring for leads and connecting the geckos

225oz enough when its a servo?

[Runner4404spd]

those run at 3000RPm you will want some kind of speed reduction it would think. i'm running a 3000 RPM motor on my mill and i have a 3:1 gear reduction. this also increase my available torque by 3:1

[Smertrios]

what is the torque of your servos?

[phantomcow2]

Since servo motors have such high RPM's (compared to that of steppers at least), typically people gear them down at least 2:1. For the diameter of your leadscrews, it will probably not be practical to have them spinning at 1500RPM, so you will want to consider a good 3:1. Since you gear it down so much, you can get away with a lower torque servo. Which also means cheaper.
So in my opinion, those are way overkill for your Taig. I've seen people convert Taigs with 200oz steppers directly coupled to the leadscrew, you iwll have close to 3x that torque after the gear down.

[warpedmephisto]

What about if you're putting a beefier spindle and motor on the Taig that can take advantage of the higher torque/traverse rates? I've been thinking of upgrading my Taig with a more powerful motor and more rigid spindle assembly, among other things. Then I could use those servos and perhaps put ball screws in it instead of the V leadscrews in it now getting a max traverse rate somewhere in the 200 IPM area with a 3:1 reduction.

[Smertrios]

The more I read about how the control systems work I have to wonder why more of the driver board makers don't make a closed loop stepper system. Adding some kind of counter for pulses coming in that only get sent to the stepper chip after the encoder shows its ready for another pulse. Sounds simple enough I guess but its probably difficult or expensive to add.

[Mariss Freimanis]

Yes, it's very difficult to achieve a true PID closed-loop stepmotor servo. The complexity comes with coordinating motion between 2 or more axis while one or all axis motors are speeding up or slowing down due to their individual loads.

It's something that's being worked on with all effort right now though.:-)

Mariss

[nicad]

I have a MaxNC 15(used to be CL) that I put ~220oz/in DC servo motors on wired to some Gecko 320's. I also changed the 20tpi acme screw to a 8tpi 3/8"ballscrew. Servo motors are still direct drive, and I have plenty of torque (2.5hp treadmill motor making the chips) to push through aluminum all day. I run Mach3. My encoders are 500cpt, for a resolution of 2000"steps" per rotation.

Here is the catch- With most any PC-based parallel-port driven software out there (ie- Mach2, Mach3), you are limited to 45000steps per second. Actually, good luck getting a PC to put out a clean 45khz, and most say not to run more than the slowest 25khz pulse rate for a clean pulse stream. At 25khz I can only move ~96ipm at my current resolution of 0.0000625" per step (max speed of 750RPM with the 25khz and 500cpt encoder). Lower resolution will result in faster moves, and vice-versa.


The MaxNC CL system is steppers with encoders, and will drop speed to prevent losing step. that part was nice, but it was somewhat quirky and it had no micro-stepping. I could get up to 250ipm on it, but my resolution was only 0.000625" (may sound small, but its not really).

My point- just because you have DC servo motors that can spin 3000rpm does not mean you can control them up to that fast. I can only run 750rpm with my 25khz pulse stream. If I were to run at the max 45khz, I would still ony be getting 1350RPM, or 168IPM.

Solution to resolution AND speed?? Enter the Gecko G100.

[Mariss Freimanis]

Solution: Get a G902, plug it into your G320 to make it a G340. It will turn your encoder into (your choice) a 50-line, 100-line, 250-line or the original 500-line encoder.

1) 500-line = 2,000 steps/rev = 750RPM@25kHz = 0.000025"/step@20TPI
2)250-line = 1,000 steps/rev = 1,500RPM@25kHz = 0.00005"/step@20TPI
3)100-line = 400step/rev = 3,750RPM@25kHz = 0.000125"/step@20TPI
4)50-line = 200 steps/rev = 7,500RPM@25kHz = 0.00025"/step@20TPI

Mariss

[imserv]

Alternate Solution: run DeskCNC at 125,000 sps

25,000 sps = ~96 ipm becomes 125,000 sps = ~480 ipm ( more than enough for a Taig, but maybe not other machines)


Fred Smith - IMService
http://www.cadcamcadcam.com/hobby

[nicad]

Fred-
wow.. I did not know there was a faster 125k versoin out!!
very nice will have to look into that.

[Smertrios]

I wonder what it would take to get something like this running on a TAIG It looks like POWER that might sell cheap! What else is needed? It has 3 servos, 3 drivers, 1 power supply but no "controller". Can it run from a PC easily?

http://cgi.ebay.com/3-Cleveland-brus...QQcmdZViewItem

REST OF POST IS COMPLETELY OFF TOPIC: It is risky purchasing on eBay IMO. I bought an air compressor a few days ago for $140.00 (that includes the S/h) and the seller claimed it to be in great shape... I get it and one of the sheetmetal valve "flappers" is bent up 1/8" completely destroying the ability of the pump to create a vacuum or compressed air. The valve also has alot of tiny dents on it.

The pump...
http://cgi.ebay.com/ws/eBayISAPI.dll...%3AUS%3A1&rd=1

This is what I sent the seller

Originally Posted by Smertrios This pump does not pull a vacuum at all and the pressure which is rated at 100psi on a new machine can't even be more than maybe 2 or 3. I took the head off to inspect the flapper valves and discovered the problem. This is the prob... there are 2 flapper valves one that lets air in the vacuum line and one that lets air out the compressed air line as a piston moves up and down. The compressed air valve was bent way up and has little round dents all over it. Since all the damage with the little dents is isolated to the one valve its definitely (IMO) a real hacks fixit job that didn't work. Don't get mad or anything I'm not saying it was your handy work. Point being that now I have to either return the pump or I need to buy a rebuild kit if one even exists for a pump descibed as 'This pump in in very good cosmetic and working condition'. https://www1.fishersci.com/ still sells a pump model similar to the one I bought and I have sent an email asking them about a rebuild kit.

then after no reply for 1-1/2 days I sent this

Originally Posted by Smertrios I contacted you about this pump to let you know there was a faulty valve 2 days ago. I also contacted fisher about a possible rebuild kit but they have not replied to my email. Anyway knowing the valve was already a mess and couldn't be made worse short of tearing it off the metal sheet it is formed in I tried my best to rebend the valve to flatness with just my fingers. It was literally 1/8 inch bent up before I started and now it is flat enough to let the pump work although still not 100%. Pump now supplies over 100PSI according to the gauge on it but I suspect that gauge is busted and 20 inches of vacuum according to my own vacuum gauge (new the pump is rated for 24 inches). The pump is working to my satisfaction now so I won't be requesting a return.