[Hirudin]

OK, made a little progress. I got my first "real" part finished today! It's a mount for the servo motor. I've attached a couple pictures and the .3DM (rhino) file, I also uploaded the high resolution originals and a couple additional pictures to my web server.

I BARELY have my machine working with steppers and I've never used servos before. I plan to get some more work done on this project soon, but I might want to play with my machine a little more before trying to figure out how to make this servo work. I'll keep updating this thread as I progress though.

[dfurlano]

How is the servo working out? I would be interested in trying this also.

Dan

[Hirudin]

Well, I can say there is technically some good news: it hasn't failed. The bad news: I haven't really been working on it.

I let the fact that I don't really have any good heatsinks for the servo driver bog me down. But, I corrected that earlier today: I bought a whole bunch on eBay. I hope they come by Friday. If they do, I hope to at least have something up and working by... I don't know... the following Tuesday.

Short, short version of my lame excuse: Until I decided to use individual heatsinks I was trying to come up with a way to use a single, giant heatsink for all the drivers (3 axes and 1 spindle). I couldn't get a design that I both liked and I thought I was capable of completing. Now with these individual heatsinks I'm just going to get it done instead of trying to make it perfect/professional.

For kicks I attached a render of the Taig with the giant heatsink on the bed...

[Hirudin]

The heatsinks came in (got 'em Wednesday more or less) and they are pretty much the perfect size. I made some g-code to drill some holes in one and got the G340 mounted.

If I can get on the stick I hope to have something a little more substantial to show pretty soon here...

[Hirudin]

Hmm... I wired everything up but it's not working. I'll keep messing with it.

Attached is a pic of where I'm at.

[edit]
I don't know what I did (I think I jiggled some wires) but now it's working much better... Well, that is to say it's working at all.

I've got it so it'll turn at 3000 RPM back and forth without much trouble. I haven't figured out how to tell Mach3 it's speed yet though. I'll continue to keep messing with it.

Also, I dropped my "laser" tachometer in my bucket of coolant...
[/edit]

['nuther edit]
Sweet, got it working on the bench. Forward (M3), reverse (M4), stop (M5), and speed (S<RPM>) all work!

This process would have went a lot faster if Mach3 would just IDENTIFY the settings that will not be implemented until Mach3 is restarted. *grumble*

[Hirudin]

This ain't workin'.

I figure the 500 count encoder I'm using is too sensitive, it reaches the ~120 count fault limit too easily. That's only 20&#37; of one rotation, at my desired top-speed of at least 3000 RPM I'm pretty sure that means if it messes up for ~0.004 seconds it'll fault out.

If the encoder isn't plugged in, the motor will spin up to 3200 RPM almost instantly. With the encoder plugged in I can only get the motor to spin up to 2000 RPM using Mach3. Once it's spinning at 2000 I can usually bump it up to ~2800 by manually typing "S2200"; "S2400"; etc. but that's way too much hassle.

I'm thinking about buying a much lower count encoder, probably a 32.

Should I throw in the towel? Even if it works am I risking blowing up the servo because it's not built for constant duty? Is there really anything the servo control offers over a standard DC motor and controller? I'm too much of a newbie, please comment if you have any suggestions.

[jalessi]

Hirudin,

For rigid tapping or making threads a servo is awesome.

Jeff...

[Jeff-Birt]

I recently started a thread over at the 'Taig Owners Club' about a new variable speed DC spindle motor for the Taig. Hirudin had asked about the servo idea there and I thought maybe I would recap my reply here to as it might be of benefit to others.

Background Info:
I've had several requests for a variable speed spindle motor for the Taig. After doing a lot of research of the last few months I found that I could get some top quality, custom made in the USA, DC spindle motors that are small enough to mount directly behind the Taig spindle. They are rated at 1/4 HP with a 100% duty cycle. This would be a big improvement over the factory cantilevered arrangement.

For a mill the size of the Taig there is almost no benefit to having the spindle motor under servo control. A good DC speed control with a tach mounted on the motor will give you speed regulation to +-1%. CNC control programs like Mach can look at a index pulse on the spindle motor and synchronize the other axis to it.

For a CNC lathe there may be some benefits to threading or if you had live tooling (doubtful on the Taig).

In general when you buy a servo it will not be rated for 100% duty cycle. So if you try to run it like a spindle motor you will burn it out. I had a customer with a CNC plasma cutting table that insisted on cutting at the maximum traverse speed (even though the cut quality was awful). They burnt out a servo motor (that cost more than a whole Taig machine) every three months or so. They could not understand that just because the machine could do 350 IPM that did not mean it could maintain that speed 24 hours a day. It would be like trying to drive your car with the engine red-lined all the time. How long do you think it would last?

If there is enough interest in a variable speed spindle motor kit for the Taig Mill and Lathe I'll proceed with getting some motors made up. Just let me know if you are interested.

[Dynamech]

I've thought about this from time to time and thought I'd post a link to a 1/2 hp variable speed motor that is intended for a turncrafter lathe upgrade. Is there any reason this motor along with a tachometer (either DRO or output to Mach3) wouldn't serve the same purpose as an expensive servo and drive? I'm sure the manual control could be modded to accept an input from Mach3 - a PWM driven digital potientiometer or something similar instead of the twist knob.


1/2 hp variable speed (manual) motor:
http://www.pennstateind.com/store/TCLVSKIT.html

[Jeff-Birt]

A CNC machine can run for hours on end and can quickly ruin a motor not rated for the duty cycle. That is why Taig uses a different, more expensive, motor on CNC mill than a manual mill.

I suspect that the 1/2 HP rating on the Pennstate may be a peak rating, not a continious duty rating. Given that the intended purpose is a small manual wood lathe that seems very likely. Some speed controls can be used in a voltage following mode and some can't. I have heard some folks have tried using this very motor though.

After the research I've done the last few months I'm convinced that a custom built motor is the way to go.

[5artist5]

What would a custom variable speed kit cost ballpark?

[Jeff-Birt]

I'm targeting the $250 price range for the whole kit. The motor is small enough it will mount behind the spindle instead of hanging off the side.

[5artist5]

What would the speed range be like?

I would be interested in something like that in that price range if the speeds were the same as the current pulley system.

[MechanoMan]

I'm still wondering about these beasts:
http://cgi.ebay.com/1-5KW-AIR-COOLE-MOTOR-SPINDLE-ENGRAVING-MILLING-GRIND_W0QQitemZ370197941616QQcmdZViewItemQQptZLH_D efaultDomain_0?hash=item5631844d70&_trksid=p3286.c 0.m14&_trkparms=72%3A1205|66%3A2|65%3A12|39%3A1|24 0%3A1318|301%3A0|293%3A1|294%3A50

1.5KW (2hp) spindle motors, with ER11 collet, which would probably just need to bolt onto the headstock dovetail directly.

See I've noted that no matter how tight I try to make this, with the highest or lowest belt ratios the Taig's little belt is prone to slipping on the smaller pulley before it ever reaches the motor's max HP output. Things then go bad, the tool breaks and the work is damaged.

They have a water-cooled spindle too. That sounds pretty awesome because air-cooled spindles with shaft-driven fans are very vulnerable to overheating at low rpm since the fan slows down too, but the user may be asking for high torque out of it which means high current and high I2R heating. Water cooling with a large CPU cooler should actually be fairly easy, only about 15% of the motor's output will be heat whereas in a CPU almost the entire electrical power consumed becomes heat (CPU does no actual "work" output). And CPU coolers are typically over-the-top to create very low temperature increases at the processor, a spindle would do fine with higher temp rises.

The mere fact that they offer water cooling implies to me they may indeed be into quality mfg. If they only sold cheap crap to n00b end users, well, n00bs don't typically know why they need water cooling nor bother to set one up, since it's more trouble than simply buying the motor. So they wouldn't be profitable to carry that line if that was you business model.

I'm really thinking higher RPM will help a LOT. You just can't belt drive at extremely high RPM- the belt will overheat from the flexing and the bearings get stressed by the side force. Also without an idler pulley, the belt flies outward a lot too.

$370 for the 0.8KW 24,000 RPM, water cooled, with inverter Or you can go $475 for 2.2KW.

[Egoblitz]

How is the servo spindle going? I plan on doing the same thing for tapping on a mini mill, any ideas on getting mach3 to run the motor continuously for drilling and using the encoder to tap? As far as running continuous, I think the 60V MS31 Series from Dynetic will be ok with 426 watts of power or .57 HP.
Any thoughts?

[Hirudin]

I thought about updating this topic to say that I have given up on this project for the foreseeable future, but I decided not to give this topic a bump. But, now that it's back, let me make it official:
Sorry, at the moment I do not have plans to move on with this project.

I don't know if there is something wrong with my servos, encoders, wiring (the wiring job I did), power (the wiring in my apartment), computer, drivers, or (now) my smoothstepper - all I know is that my servo sounds like a cement mixer whenever the speed is being controlled by Mach 3. If I unplug the encoder wire the motor spins up to ~3200 RPM nearly immediately and is nice and smooth and quiet.

I now have one of those "Chinese spindles" on order. I also ordered a VFD that is made for 110VAC input power (apartments suck). On paper the VFD looks like it will work to me, but that's what I thought about the servo. I plan to start a new topic about my experience with it only after I've successfully made it spin at least one time.

Of course, the spindle/VFD combo will not help with tapping.