How bad an idea is this ancient Matsuura MC-400H-30-PC11?

sansbury · Jan 13, 2017

The suspect: MATSUURA 4H PC11 CNC Horizontal Machining Center -HMC # | eBay

I have something of a side project going where it would be really nice to have a 40-tool HMC with a pallet pool to play with. I say "play," because this is really more of a research project than anything else, and I wouldn't have customer orders or other bill-paying activities lined up that would suffer when it goes down.

I'm willing to take the risk of the control dying because I will likely want to retrofit it with LinuxCNC in order to have complete control/hackability of the machine for my own purposes. I know that's a few hundred hours easy plus some money for PC and interface hardware (I have experience with the system and am very comfortable with software configuration and build) and I'm OK with that.

What would make this a pointless boondoggle for me would be if I bought the machine, spent easily as much money rigging it in and wiring it up in my space, then spent a hundred or two hours retrofitting the control, and then find I need to replace a bunch of mechanical or electromechanical parts that cost 4-5 figures each. At that point, I'd be better off buying something newer and less beat up.

Anyway, I guess what I'm trying to get a sense of is, if a qualified inspection found the machine to be basically sound and in good enough condition for my immediate needs (think ~+/- .001" tolerances with no concern for speed), other than the control dying, how much risk do I have of ending up with a pile of scrap iron due to failures that I can't solve with sweat and elbow grease?

Thanks in advance, including for the people who will tell me this is the worst idea since New Coke (around same time this machine was made, I think) :nutter:

lowCountryCamo · Jan 13, 2017

Might be a good buy but I would never, ever, retrofit with any control! Unless you are one in 10,000 it will never make chips with linux. If you buy just use it like it is. Cant make it better.

dkmc · Jan 13, 2017

lowCountryCamo said:
Might be a good buy but I would never, ever, retrofit with any control! Unless you are one in 10,000 it will never make chips with linux. If you buy just use it like it is. Cant make it better.

Well that didn't take long....for the first clueless comment to appear...
It will never make chips with Linux.....what a riot.

sansbury.....you're playing to the wrong crowd on this forum.
All they know is OEM. Anything else is crap. NOT doable.
Knowing Linuxcnc's potential, I say go for it!

Mud · Jan 13, 2017

What year is it, does it have DC brushed servos? What brand, and what drives, are they available used?

Does it come with hundreds of BT40 toolholders? BT40 seems to be getting scarce on Ebay.

Vancbiker · Jan 13, 2017

dkmc said:
Well that didn't take long....for the first clueless comment to appear...
It will never make chips with Linux.....what a riot.

Probably would make chips, but be a solid reliable machine tool, not very likely. If the OP has stellar skills at creating great logic he might get there.

Getting the axes to move and position properly is the easiest part of a retro-fit. The drives and motors on that machine are pretty ideal for retrofit use. They are DC motors with encoders for position and possibly tach feedback for velocity though they could be using F/V conversion. Depends on vintage and OEM specs. 0-10V analog drive signal so compatible with common PC based retrofit hardware.

I've been doing maintenance and repairs on CNCs since the late 70s. I've seen many retrofit attempts by non-professionals and pros alike. A couple of the pro jobs worked out OK. In both instances that come to mind, it was because the original machine ladder logic was pretty much duplicated in the new control. I've never yet seen a retro-fit where all new logic was created be robust and reliable.

In the OPs case, he has a tool magazine, tool changer, a pallet changer, and a pallet magazine to generate safe, solid, robust code for. That's a huge task if one is looking for reliable and safe operability.

sansbury · Jan 13, 2017

Vancbiker said:
In the OPs case, he has a tool magazine, tool changer, a pallet changer, and a pallet magazine to generate safe, solid, robust code for. That's a huge task if one is looking for reliable and safe operability.

Yeah, I definitely see a lot of sense in that. I've converted a few manual machines to LinuxCNC, but just benchtop stepper-driven toys, definitely a much smaller challenge than this. This machine would definitely be the Ph.D. course. I wouldn't really be putting this machine into production so I'm OK with the reliability being less, but I wouldn't want to make compromises on safety.

On the plus side, building and configuring software systems is thing I'm good enough at to be able to afford a side project like this. Hard I'm good with, stupid or pointless, not so much.

ewlsey · Jan 13, 2017

It's just the amount of time it would take. I'll wager a retrofit from a legitimate integrator using a new Fanuc control would come in between $100K and $150K. Even cheaper hardware like Fagor or Dynapath would probably be in the $50k to $75K range.

And you guys are telling me a DIY guy can do it for $100 with some napkin sketches and shareware?

BGL · Jan 13, 2017

Matsuura retrofit

Garwood · Jan 13, 2017

That machine has a negative value. It's costs to move it or get it to the scrapyard far exceed it's value in scrap or usefulness.

They would have to pay atleast several thousand to remove that machine.

Big old HMC's are scrapped. When scrap is low they sit there. Because rigging outweighs what 40K lbs of iron gets in scrap.

If you need an HMC spend a bit more and get something ready to make parts.

JP Machining · Jan 13, 2017

As someone who has built and retrofitted several machines and worked with PLC's (but nothing yet with linuxcnc), I'd say go for it. Once broken down its all really just simple if this than that type basic stuff, just that theres alot of it for each step. If you can somehow find the ladder logic of the original control it will make the retrofit alot easier, just little things like, oh that switch should be open when the pallet is in this position type of stuff. The biggest question would be condition of spindle and if its rebuildable (thinking its too old for cartridge style spindle, so should be easy bearing swap), servos and drives are are replaceable these days just may require adapters etc. If it has box ways id want to know how they are, if linear, well they can be replaced too.

If any of that is questionable and by the time you figure in rigging costs etc, it may be better looking at newer, and in my shoes Id be looking for faster better control anyway, but you said you didnt care about that.

ewlsey · Jan 13, 2017

Used HMCs don't have very good resale value. I see nice looking 2000ish vintage Moris and Makinos on ebay for less than $20k all the time. Good VMCs from that era are worth more than that...

dkmc · Jan 13, 2017

Vancbiker;2898981 I've been doing maintenance and repairs on CNCs since the late 70s. I've seen many retrofit attempts by non-professionals and pros alike. A couple of the pro jobs worked out OK. In both instances that come to mind said:
As a repair tech of course it's in your best interest to poo-poo retrofits.
Older controls that keep having problems for you to re-visit are your bread and butter. Retrofit the machine with a reliable modern control (heaven forbid an approach like LCNC that includes a FREE software solution) and your out of a job. I get it. Never met a machine tool repair tech yet that has anything good to say about machine tool retrofits.

Wonder why.....

ewlsey said:

And you guys are telling me a DIY guy can do it for $100 with some napkin sketches and shareware?

Click to expand...

Nobody said that.
You're just delusional as usual.

Vancbiker · Jan 13, 2017

dkmc said:
As a repair tech of course it's in your best interest to poo-poo retrofits.
Older controls that keep having problems for you to re-visit are your bread and butter. Retrofit the machine with a reliable modern control (heaven forbid an approach like LCNC that includes a FREE software solution) and your out of a job. I get it. Never met a machine tool repair tech yet that has anything good to say about machine tool retrofits.

Wonder why.....

At least in my case, you're reading the wrong story. For starters, I'm no longer in the repair business, just an old retired guy.

I've had the fun of trying to keep folks retro-fits going during the time that I did do field service. They actually are excellent for a service tech's income. Many problems take twice as long to diagnose due to shoddy or more often, zero documentation. Wiring diagram? What's that? You mean this stack of scratched up, unorganized sketches? Where's the maintenance manuals for these servos? Oh, the retro-fitter didn't leave you any?

My experience with retro-fit machines may be unique. Maybe all the ones I did not work on were properly done and well documented. Somehow I doubt it.

ewlsey · Jan 14, 2017

Vancbiker said:
My experience with retro-fit machines may be unique. Maybe all the ones I did not work on were properly done and well documented. Somehow I doubt it.

Not unique at all. I've worked on tons of retrofits. Figure any CNC machine tool in active use that was built before 1980 has been retrofit at least once. Some of the OEM retrofits I have worked on were awesome. Documentation was excellent and everything worked with OEM support.

The problem is the backyard retrofitters. They don't have the resources to get in bed with a real control manufacturer like Fanuc, Mitsubishi, or Siemens. So they push the cheap controls like Centroid.

I'm working on 3 machines now that were retrofitted about 8 years ago with PC based MDSI controls. The problem with all PC based retrofits is the PC hardware. At just 8 years old, repair parts for those PCs are already obsolete and impossible to get. The software was built for the hardware and you can't just drop in a newer PC and expect it to work. These machines are in a large government facility. They can't just get on ebay and buy a used motherboard. They need OEM support.

So now you are stuck doing a complete retrofit again after just 8 years. The iron on these machines was built 40 years ago and can easily outlive me.

Steveo · Jan 14, 2017

That machine is late 80`s. I have an old Mats and parts, the mechanical kind, are hard to get and extremely expensive. I had the spindle drive go on mine a while back. Yasakawa wanted 6k to rebuild it and no warranty. I found some one local who did a good job for 3k. If I wanted to upgrade to a new one, it was 10k installed. The weird ass coolant pump is a grand from the dealer. Old Japanese iron is great until it breaks.

gregormarwick · Jan 14, 2017

There are two discrete points of argument in this thread and everyone is getting them mixed up and bickering unnecessarily.

The first concerns the capability, suitability and quality of the TOOLS that are proposed to be used for a retrofit.

The second concerns the quality of the WORK that is done using said TOOLS to complete the project.

To address the first point, LinuxCNC is mature and extremely capable in the right hands. It provides everything required to complete a feature complete, modern retrofit of basically any machine tool ever made with comparable performance to any of the mainstream CNC controls. Anyone who claims otherwise is willfully ignorant or for whatever reason has an agenda to discredit it.

Per the second point - There are plenty of bad or mediocre retrofits out there and plenty of well-made points about that in this thread.

• LinuxCNC is a software product. The quality, suitability and performance of the hardware used is entirely up to the integrator.
• It runs on Linux, the most hardware agnostic OS by a huge margin. Future hardware compatibility is as close to guaranteed as you can hope to get.

Anyway...

You WILL sink a lot of time into this if you intend to do it properly. The fear of unforeseen expenditure in the form of replacement parts is absolutely valid unless you plan in advance to do a complete retrofit (replace all the drives, servos, spindle motor, wiring etc.) - in which case the project will never make sense from a cost perspective anyway. The value of the machine when completed will be low, but depending on your work it might perform very well.

ewlsey · Jan 14, 2017

gregormarwick said:
To address the first point, LinuxCNC is mature and extremely capable in the right hands. It provides everything required to complete a feature complete, modern retrofit of basically any machine tool ever made with comparable performance to any of the mainstream CNC controls. Anyone who claims otherwise is willfully ignorant or for whatever reason has an agenda to discredit it.

That's just not true.

The Linux CNC software can move multiple axes and run a tool changer. However, it's obvious that it was written by software people and hobbyists, not by anyone who knows anything about machine tools.

It lacks the following things that have been available on "real" CNC controls since the late 1970s:

-A real editor
-Real rigid tapping (it can kind of do it with limited control on depth)
-Canned cycles for lathes
-U,V,W incremental moves for lathes
-A real industrial interface that doesn't require a mouse

Linux CNC is so concerned about the esoteric details of CNC like realtime control. However, it skips the details that make it really work in an industrial environment.

Ox · Jan 14, 2017

ewlsey said:
Not unique at all. I've worked on tons of retrofits. Figure any CNC machine tool in active use that was built before 1980 has been retrofit at least once. Some of the OEM retrofits I have worked on were awesome. Documentation was excellent and everything worked with OEM support.

The problem is the backyard retrofitters. They don't have the resources to get in bed with a real control manufacturer like Fanuc, Mitsubishi, or Siemens. So they push the cheap controls like Centroid.

I'm working on 3 machines now that were retrofitted about 8 years ago with PC based MDSI controls. The problem with all PC based retrofits is the PC hardware. At just 8 years old, repair parts for those PCs are already obsolete and impossible to get. The software was built for the hardware and you can't just drop in a newer PC and expect it to work. These machines are in a large government facility. They can't just get on ebay and buy a used motherboard. They need OEM support.

So now you are stuck doing a complete retrofit again after just 8 years. The iron on these machines was built 40 years ago and can easily outlive me.

Yep - been saying this for years.

-------------------------

Think Snow Eh!
Ox

dkmc · Jan 14, 2017

MDSI was a software ONLY solution as far as I remember.....NEVER seemed like a good idea.

The people that make the hardware used with LCNC
are keeping it current with modern computer platforms.
All ya gotta do is a bit of research.

ewlsey · Jan 14, 2017

dkmc said:
MDSI was a software ONLY solution as far as I remember.....NEVER seemed like a good idea.

The people that make the hardware used with LCNC
are keeping it current with modern computer platforms.
All ya gotta do is a bit of research.

That's not fair either. The Mesa cards or whatever others are using are not forward compatible. If the PC dies, you load the software onto a new PC, but now your card won't plug in or won't play nice if it does.

How bad an idea is this ancient Matsuura MC-400H-30-PC11?

Aluminum

Stainless

Diamond

Diamond

Diamond

Aluminum

Diamond

Hot Rolled

Diamond

Stainless

Diamond

Diamond

Diamond

Diamond

Cast Iron

Diamond

Diamond

Diamond

Diamond

Diamond

Similar threads