Help me decide. To Retrofit or not to Retrofit.......

Machine -
1993 Fadal VM15XT
Geneva style tool changer
7500rpm spindle (that's done with a 10volt signal) also easy FWD and Reverse connections
I already have Mesa 7i96s and a 7i85s. I'd want to run stepper online 1000watt servos possibly even the ethercat versions, but the 7i96s I have would run the step/pulse version just fine. Might have to snag another mesa car for some inputs, but the Fadal is a simple machine. Air operated spindle orientation, simple operation of the tool changer. I like the idea of absolute encoders as I could treat it kinda like the original control that used resolvers and simply call the old cold start my "home".

The reason I'm thinking about this is I need to replace the Y axis servo motor for a cool ~$1700. Instead op throwing $1700 at older tech, I can spend about that getting everything else that I need. Only down side is I would have some down time while setting up LinuxCNC.

Once I get the bugs out of the linux install, I'd imagine if I could control myself from tinkering with it, it would just run like its supposed to. So do you think its worth it or not? I have about $20k of work for the machine in the next 8-10months. Being down right now is not a HUGE issue, but I have a solid month / two months or so before it needs to be able to make parts again.

Also I could sell stock axis amps along with amp cards and such and cover more than it would cost me to retrofit the machine.
 

I know its not going to be as easy as replacing the servo motor, but it SHOULD be more rewarding in the fact that I will not be depending on 30 year old hardware that may or may not get me for 300/600/1800 at a time lol

To be honest the only thing that has me concerned is writing the logic for the tool change.

I know what needs to be done and the steps, but I do not know anything about classic ladder or programming so that will be some learning on its own.

That's a price for new servo. DC Brushed machine. I like the old fadals cause you can get parts for almost everything. I just don't wanna have to buy this one lol.

Glentek GM4050 (MTR-002) runs about $300 to $800 on ebay.

I have some thoughts/suggestions on a possible retrofit if you're interested.  Little late tonight, but if you're interested let me know.

A lot of those GM4050s are not the -38 that I need. 4050-38. I did find a few for ~$800.

What are your ideas on retrofit...Im all ears.

ConSonar wrote:

A lot of those GM4050s are not the -38 that I need. 4050-38. I did find a few for ~$800.

What are your ideas on retrofit...Im all ears.
 

Sorry if I gave you bunk ebay info - the usual Fadal spare parts manual didn't specify a "-38" sub-model.

Ok, so retrofit thoughts...

Do you have any experience operating a machine under LCNC?  If so, no issues on that front.

If not, then along with all the retrofit stuff you'll have to learn how to write programs and operate LCNC.  I love LCNC, but it's NOT Fanuc (or Fadal 88), and there is a learning curve. Little things - like single-block - are different, as are some big things (Macro B, GOTO, etc).

There are a number of people on the forum who have demonstrated a successful Fadal retrofit (including a recent one using Probe Basic).  But I suspect it took them more than the  "month or two" you indicated you have.

Those stepperonline servos ( I have some T6's) are decent, but the programming/tuning software is a miserable hack of Leadshine's OEM version.  And Leadshine's version doesn't work any better on the Stepperonline drives.

Torque:
T6-1000W - 3.19nm (9.56nm peak)
GM4050-38 - 4.67nm (23.35nm peak)

So they're way down on torque from the Glentek's.

Along with fighting the software, those servos will require an adapter plate and a different coupling at minimum.  Not a big deal for someone with a working mill, but not easy to accomplish without one.

For a hobby situation, I'd say go for it.  But for a working shop, I'd suggest being a bit more... conservative.

I'd buy a used servo and fix the Fadal, and also the parts needed to get LCNC working on the Fadal.  Then, while the mill is making parts, put together a working benchtop configuration of LCNC with all the I/O mocked up.  That includes choosing a GUI, building a new operator panel, as well as a mocked-up ATC.

You can 3D print a geneva mechanism and get a cheapo DC motor, along with some $10 mini air cylinders, and have a working model of the ATC to drive around while you sort out all the LCNC programming & config.

The operator panel you have could be hacked up, but it's worth more intact to somone else.  Use the 'pre-rebuild' time to fabricate a decent panel with physical inputs & MPG.

You can take your time, run some programs, even run the 'toolchanger' - about the only thing left will be tune the servos against the actual loads.

Get it all sorted out and only then take the mill out of service and do the retrofit.

As far as the ATC goes, there are a number of ways to skin that cat.  Probe Basic has built-in ATC subroutines and it's pretty easy to connect to your existing geneva drive.  Andy Pugh's recent edits of carousel.comp are also great, and there are plenty of examples on the forum of working ATC's.

Just my two cents.