Poorly N sick Fadal 15 VMC Will live again!

[JAZZCNC]

I was going to mention the CRT but though that might be a little facetious! I won't be so reticent in future

The overall machine looks to be in good condition bar some reasonable wear and tear. Jazz, do you plan to get the original control system working or upgrade it to something more contemporary?

Kit

Sorry Kit I missed this.! . . . Watch this space for more info.!

[JAZZCNC]

Ok so stuck some power to it and sure enough, it was still poorly sick n dieing. . . . Kicking it didn't work.!

So after further investigation and checking out the price of Fadal replacement cards it soon became apparent it wasn't worth sticking with the original control.
Apart from the fact parts are expensive and difficult to find in the UK it's also limited in Memory for running G-code generated by modern Cam. It could be made to do so by drip-feeding or adding USB etc but this all adds up and still got old control.

Now the question is what to replace it with.? and what to replace.?

The machine uses DC brushed Motors with Resolvers and AMC Amps, which all work fine and good quality units. However, I've decided to replace these with AC brushless servos and Modern Servo drives. The reason for this is simply to bring all up to modern standards that are readily available. The drives and resolvers for these old motors are getting harder to find and very expensive when found. Also, I can sell the old ones which will go a long way to AC servos or even cover the costs completely depending on what I use.!

Doing this also simplifies things in lots of ways.??? . . . Why the question marks, well there's a twist to the decision.!
Modern drives mean I can use Step n Dir controller rather than Analog if I like which will make things a little easier come set up time. The twist is that I want to use the 4th Axis which uses Brushed DC motor with resolver and the drives require Analog.!

So my options get a little more limited unless I change the 4th axis motor as well which I'm not so sure is easy to do. Need to check it out.!
The 4th Axis is high-quality Japanese Nikken unit which is also worth lot of money so possibility I might sell it.? . . . Not sure how much I need a 4th axis.!
I also have to consider the Resale value it will add to the machine compared to how much it's worth if sell it on its own.? Could shoot my self in the foot because there's a high probability I'll sell this machine when it's finished and could add a lot more value and salability to machine than the individual resale value of the 4th axis.?

To keep the 4th Axis without changing motors leaves me very few choices because the controller will need to mix Step & dir with Analog and be able to handle resolvers rather than encoders.!

Basically it boils down to just two, LinuxCNC with Mesa cards or Dynamotions Klop, even then I'm not sure the Kflop can use resolvers and requires encoders.? It is a possibility I can just change the Resolver for encoder on the 4th axis which I don't think will be too difficult. Need to check both out more.!

If I drop the 4th Axis then I can pretty much use whatever I like. However, this machine deserves a little better than just being closed loop back to the drives, so I want the loop closed back to the controller as well.
This rules out some of the weaker and cheaper options like Mach3/4 or UCcnc unless I go Analog, even then it doesn't come cheap and is less than ideal.

So I'm back to LinuxCnc, Kflop or some of the more expensive options like Centroid Oak, etc.

At the moment I'm strongly leaning towards Linux CNC and mesa cards which I know is more than capable.! . . . The weak link here is ME.!. . I've never used Linux CNC or Mesa cards but I know a man who knows is onions, hey Clive S..Lol . . . I'm sure together we'll figure it out.!

Watch this space.!

[JAZZCNC]

Here's a couple of vids of me getting little excited thinking I'd found something good.!

[Jonathan]

Nice machine. Bit odd how they leave the rails sticking outside the castings - does the carriage move onto that region or stop before?

At the moment I'm strongly leaning towards Linux CNC and mesa cards which I know is more than capable.! . . . The weak link here is ME.!. . I've never used Linux CNC or Mesa cards but I know a man who knows is onions, hey Clive S..Lol . . . I'm sure together we'll figure it out.!

Has everyone changed since I've been gone . I'm sure between the people here you'll manage it. I've not tuned a servo motor with closed loop control through LinuxCNC either, but in other systems I've worked with it's generally not been so hard so I wouldn't shy away from it if that's a solution you can try without spending extra cash.

[JAZZCNC]

Nice machine. Bit odd how they leave the rails sticking outside the castings - does the carriage move onto that region or stop before?

Honestly I don't know I've not moved axis to end stops but I've seen other industrial machines use similiar setups. Think some older Haas machines use this setup, which isn't surprising because the First haas VF machines where based on Fadals.

Has everyone changed since I've been gone . I'm sure between the people here you'll manage it. I've not tuned a servo motor with closed loop control through LinuxCNC either, but in other systems I've worked with it's generally not been so hard so I wouldn't shy away from it if that's a solution you can try without spending extra cash.

Only thing that changed is a Few Old mates are missing and missed.!

Regards the Closed loop and Analog it doesn't really bother me if honest but why make things difficult if can use Step & Dir.
If wanted to keep the costs down I could just throw encoders on motors and use the Cslabs IP-A controller I already have and use all the existing motor/Amps etc But don't want to take that route, I'm thinking more long term for when things fail.

[JAZZCNC]

Nice machine. Bit odd how they leave the rails sticking outside the castings - does the carriage move onto that region or stop before?

You got me curious about these rails Jonathan so been to work today and wound Y-axis to the limits. The carriages stay within the casting limits and don't ride onto the overhanging rails at all so not quite sure why they would waste so much rail.? The only thing I can think to explain is that they do an XT version with longer travels and can not imagine the XT uses a different frame casting only longer rails and ballscrews so maybe it's got the longer rails on it.? That said the X rails only hangover 50mm each end so not sure about X.?
Put a longer ballscrew on Y-axis and could easily get another 200mm travel.

[m_c]

You got me curious about these rails Jonathan so been to work today and wound Y-axis to the limits. The carriages stay within the casting limits and don't ride onto the overhanging rails at all so not quite sure why they would waste so much rail.? The only thing I can think to explain is that they do an XT version with longer travels and can not imagine the XT uses a different frame casting only longer rails and ballscrews so maybe it's got the longer rails on it.? That said the X rails only hangover 50mm each end so not sure about X.?
Put a longer ballscrew on Y-axis and could easily get another 200mm travel.

I've not watched the video, but could it be for supporting/mounting the way covers on?

[JAZZCNC]

I've not watched the video, but could it be for supporting/mounting the way covers on?

No Moray the Fadal doesn't have conventional way cover on the Y-axis ie: telescopic covers. It's a one-piece cover that goes through the saddle Quite neat setup actually.

[JAZZCNC]

Ok quick update.
I'm still in the process of trying to decide which controller I'm going with. I'm narrowing it down slowly and I think I'm nearly there.?
These are the candidates that are still in the running.?

Linux CNC with Mesa Cards 7i76e + 7i85 which I've already purchased and are currently being played with by Clive (Clive S) testing with AC incremental servo's and closing the loop back to Linux CNC.

Next up are two Chinese industrial controllers both using 2Kw Bus type Absolute Servos. Newker 1000-18i-Mi and Gunt - 350iMa both with C type Subpanel with MPG.
These are interesting to me because I like the idea of an industrial controller without a PC. I also like the idea of Bus type Absolute servos because of the ease of use wiring/setup and servo's always knowing their exact location even after power down so not having to install limit or home switches.
My reservations are both of them are Chinese and I'll be dealing with Chinglish manuals which won't help when it comes to writing ATC macro's and using PLC.
However, I'm used to Chinglish and dealing with Chinese company's and the price is very very good.!
Only downside is I won't be able to use my 4th Axis so I'll probably sell that if take this route.

I was looking at Adtech controllers/servos but I've seen some negative comments from users about lag on 3D toolpaths or on toolpaths like adaptive with high data rates due to slow DNC feeding the code from external cards etc. Also after carefully checking out specs and manuals etc, they are quite limiting in speed at 10mtr/min when really this machine would run closer to 20mtr/min rapids with 2Kw AC motors.

http://newker-cnc.com/index.php?m=co...=165&l=3&id=90

Lastly and new to the running and one that I'm seriously considering using is Puruvesi ET10 controller using UmyCNC software.

This is a company I've only discovered in recent months through helping another member of the forum and after doing a lot of research and reading the manual cover to cover I've become very impressed with it.
The Ethernet controller is very well made and provides both Analog and Step N Dir closed-loop system with more than enough I/O for my needs. It also allows me to mix Step n Dir with Analog so I could keep the 4th axis with an encoder change.

The UmyCNC software is very nice indeed, not so much because of how it looks but for what it allows you to do and the S curve planner it uses. It pretty much allows you to build a system exactly as you'd like it using a mixture of G-code and C++ Macros and PLC. S-curve planners give very nice smooth toolpaths.
The GUI is fully customizable so you can build custom front end if required and it's very Touch screen friendly. (I'm not keen on the profiles provided so I probably will be doing this.!) It's also very flexible regards what computer and OS it will run on, Linux, Windows even Raspberry py's. I'll probably run it on Linux as it's a very stable OS.

The only downside with this setup is that I will have to use Incremental encoders as it doesn't support absolute encoders.

Another BIG plus is the backup and fact hardware and software are both in house. I've already been in touch with Ivan at Puruvesi who has been super helpful with quick replies answering technical questions and offering experienced advice. The support and backup is a big Plus.

I'm also seriously considering switching to there lower ET6 and ET7 models controllers on the routers I build. They have everything I need from a controller, like slaved motors and gantry squaring/homing, probing, built-in camera vision system and tangental knife friendly which is great for some of the plotter type machines I build. Again the S curve planner is also interesting and exciting because they really shine on routers and improve performance massively on 3D toolpaths.
They will also be more than good enough for the Mills I convert and the future plans I have for super-strong Mini Mill I intend building. They are also very plasma orientated which is a bonus and an area I will be looking into more also.

However, as always I'll be thrashing them to death first on a test machine to within an inch of its life before I commit to switching. But to be honest I haven't been this impressed by the Spec of a controller/software for a long long time and after communications, with Ivan, I'm very optimistic it will be a good experience. You can rest assured if it's not you'll hear about that as well.!

I'll keep you posted.