PDA

View Full Version : Turning centre retrofit help/project on a Mazak/Yamazaki



spluppit
05-09-2010, 04:24 AM
Right then! Cant sleep again and been meaning to do this post but dreading it at the same time. I will have so many questions and half I will forget to mention in this first post, so you are prolly gonna hate me after a while, so sorry in advance. Before I start I will admit to being an electronic idiot. So this is also about if I will be capable for doing this myself or not.

This is the first machine I want to try to tackle and be rid of commercial controls once and for all! The plan is to retrofit the control with something else. Obviously this is a commercial machine so it really has to have the ability to function as per its design IE all functions available.


First of all some specs to give you a feel for what the score is

2 controlled axis Z and Y
10 Mtrs a min rapids.
12 station turret
15 Hp spindle motor with speed controller and encoded.
Fully automatic hydraulic tail stock this includes the body and the quill.
Automatic Hydraulic chuck.
Dc brushed servos (big buggers)
Automatic lubrication.
Bar feed (at a later date)

This is where the fun starts, as I really don't know where to start! I am lost, totally! I have spent months reading about this supplier that supplier this board and that control board. I'm still none the wiser!

Hardware

This is the bit i find totally confusing with all the boards out there. Im looking for a realistic way financially to do this. I have noticed there are things like control panels out there pre made like this.

http://www.slidesandballscrews.com/g60tt-lathe-controller-p-324.html?cPath=64

What would you experts out there advise, go for something as above or or go the pc route. I want to avoid being in the situation where in 5 years spares are not available for things like that controller. Its a pain in the neck very often with industrial machines where controls become out of date, boards are no longer made, so the manufacturer uses that as an excuse to attempt to charge 5 million quid for one board. This situation is exactly what I’m trying to avoid. I would think the pc route is the best way to go if you guys think iI can get all I need out of the machine via a pc.

The next point is compatibility of hardware, are all these boards out there like servo cards compatible with other makes of boards? Its an issue as I don’t have a clue about compatibility. For instance if a board goes down and I need to replace, then find they are no longer available, and have to replace with another manufacturer, how much of a compatibility problem is there?

Who to use as a supplier is next. I have scoured many sites over many months a lot do not cater for the servo market as they are literally home cnc jobbies for steppers. I would welcome advice on who is a valid candidate. The criteria is cost, reliability of product, quality of product and back up. Also a company that has been around for a while and is likely to be around in the future.

Next is safety systems. A machine like this has many safety systems in place so things cant be activated manually when the machine is in automatic mode. Like for instance the chuck cant be opened when in auto mode, or the tail stock cant be activated. Are these safety systems handled by things like mach 3 and is the hardware capable of handling such safety features?


Software

I am open to all of your ideas on this subject. The only one I have tried to study is mach 3 as there is a a host of information out there. I am still left with many questions if it can realistically handle a machine such as this.

Manual control of the machine is also important for a machine of this type, for setting purposes and dry running/ testing of programs, as such the machine (ideally) needs to have a control panel (as in the pic) to cover certain features such as:

spindle speed override
Feed rate override
spindle speed
Hold feed
Axis jogging
Tool selection
Coolant
Dry run (means the spindle/feed overrides are active overriding the programmed speeds/feeds. It also reduces G00 rapids to a much lower rate like 2 mtrs per min)

There are many other types of features, I would expect a lot will not be needed as they are really for that control type.

I guess I would like to hear from a pro out there that can tell me if this really is viable with the home build style of parts available.


Obviously I am prepared to spend money on this machine, but it has to be viable against the value of the machine overall. I would also gladly entertain paying someone that has the knowledge and wants to play with a industrial machine. You could be someone who is a electronics wizz building your home build cnc machine, but is lacking in facilities/ ability/knowledge in Engineering proper, and would would like to exchange work done for you/ machining facilities, or proper payment or a combination of the two. Either way I'm very open to suggestions.

I'm sorry for the long bleat and numerous questions. Pics below of the beast in question, I know you all like pics.


Kind regards,


Scott

29362940293729382939
http://www.mycncuk.com/forums/images/misc/pencil.png

Wobblybootie
05-09-2010, 06:58 AM
If costs are a big issue then you may be able to consider EMC2 instead of Mach3 it costs £0 BUT it is a Linux program and will require a Linux PC. Like Mach it does have a really good support forum and there are folks here running it as well, it is what I will be using when I reach that stage.

Tom
05-09-2010, 09:23 AM
Hi Scott,

Lovely machine! Will be very powerful when it's finished....
The spec (large servos, turret, auto chuck, bar feed, etc) mean that it's beyond the typical scope of a hobby machine. But that's not necessarily a good reason to stop!
Converting a machine like this is not trivial, but it is entirely possible, while still satisfying your wish for modular systems which are "future proof" and replaceable.

To start though, you'll need to shake your "electronic idiot" tag! :) (I'm one too when it comes to servo control). To do this on your own you'll need to do enough research and learning to genuinely shake that tag.... For some basic DIY builds it's possible to talk people through what steps to take - yours though is a level of complexity beyond.... You'll need to do enough research to become an expert before really starting. It won't be quick, but it will be hugely satisfying...

If it were me, I'd be reading up about EMC2, and digesting some of the servo projects that people have done with it. Check out EMC2's "ladder logic" for the turret control and tool changing. If it's interesting reading, keep going... If it reads like double dutch, this lathe conversion is probably for someone else!

Robin Hewitt
05-09-2010, 04:12 PM
Anything can be hacked but most hackers start with small cheap gadgets and work their way up. This is probably not a good machine to start a hacking career with.

OTOH if you don't hack it then you get screwed everytime a plug comes loose.

You could hack it by degrees, adding functionality as you need it.

First question is, does it program exclusively through the front panel or does it have a data port or does it have a floppy drive/ magnetic tape reader?

Great looking machine though :smile:

Robin

spluppit
05-09-2010, 05:54 PM
Gentlemen, many thanks for the replies.

Firstly Tom,

This is the great shame about machines of this nature. These machines being old were really built to last, unlike the modern cnc's which are now considered disposable. The carcase of this thing is really very good indeed, plus the maker of this particular machine is renowned for being the best in the world. Its actually a Mazak, back in those days sold under the name of Yamazaki, Yamazaki being the actual producing co of the machines.

Things like the the bar feed, tail stock, and chuck are run from M codes, which i understand are all included in the PC type controls mach 3 etc. My understanding is they are there to be used, but its down to you if you want to actually use them, often for the home build they are not required. I'm not sure on the tool turret though as when programmed you use an T number. T01= tool 1. T02 = tool 2 and so forth. My understanding is all PC home build software are only using all the generic G and M codes invented donkeys years ago. The thing is a lot of the systems are there in the machine and wired up, spindle controller, relay circuit for the tail stock, chuck, servos, limit switches etc. Its a matter of knowing if those existing systems can be integrated into modern software/ hardware.and the correct signals sent. These are the things i do not know, due to being a total numpty.


This is the hard part as you correctly say, 'shaking this electronic idiot tag.' Don't get me wrong, I understand the basic principles of how it all works. For instance: Servo driver card controls the current inputs to the servo, which then moves, the encoder reads the position, sends signals back to the control so it knows where it is and what its doing, but this is where it ends! I understand a signal has to be applied somewhere to make an action happen, After that i pass out!
I have tried researching as mentioned, but its very difficult with this sort of thing there really is no information out there, set out in the simpletons terms i need. Its all related to people that have x knowledge to start with, so trying to read jargon and terminology that means nothing to me, It's tricky. As always having someone to show you the ropes makes life much more tolerable. It's exactly why i have made an offer for someone who has a lot more understanding than myself.

I have taken an hour look at that EMC a lot i understand with the actual control side as its familiar being standard G code, but the technical side of the actual software itself, as you say double dutch!


Secondly Robin,

The machine was known as a NC machine, NC being paper tape reading machines. How it works is the machine reads punched paper tape, different configurations of holes on the tape represent x command. So you stuffed this tape into the reader, it reads the tape and then holds the program in memory ready for use. You can see the tape reader panel cover in the 1st and 4th pics on the left hand side, its the darkened perspex panel.

This is where it gets confusing! You can still manually program the machine via the key pad. So if you loaded a program via punched tape and it had errors upon dry running the program, you can go into the program at the control panel and edit the program for use. You can totally program the machine at the machine control panel also. I have never used the punch tape, I have always programed the machine at the control panel.

It does have a coms port, I believe its a rs 232 or something, so it can be linked up to a PC, configuring might be another matter. I did look into this, speaking to the machine maker. The info i had was it can be used to dnc/ upload/ download etc.


Thirdly Wobblybootie,

Thanks for your input. I have gone and had a look at that software, I didn't even know it existed. I'm still none the wiser of the route i should take.


Kind regards,


Scott

irving2008
05-09-2010, 06:28 PM
Scott,

A question first: Is this machine earning its keep today, and if so, how long a downtime can you afford? While I think there are enough people on this forum to advise you on this project, its not going to be a 1-2-3 and you're done job. I am not aware of anyone who does this sort of thing as a full time job, so it might be a little harder to find someone who can provide a one-stop service. (is there a demand for such? A question for all to discuss... but not on this thread!)

There is a level of research to be done, and possibly some machine specific coding/configuration. Like Robin I agree having your own in-depth understanding is the only way to stem the 'manufacturers drain on the bank', and having it fully and accurately documented means you can always hire someone to change it in the future, especially if you stick to opensource and common spec parts, e.g. making sure, for example, that the servo drivers you eventually buy conform to certain standards such as signal levels. Fortunately many of those standards are in place these days whereas in the past everyone had their own (again the manufacturers revenue generating 'lock in').

Oh, another question - where exactly are you in 'south east UK'?

spluppit
05-09-2010, 07:05 PM
Many thanks for your reply irving,

Currently it's not earning it's keep, It has been pretty much a pile of cast iron in my way for some time. I will elaborate on that if required.

As far as I'm concerned the machine can be down for as long as it takes, if it's 2 years so be it. The only other option is the scrap heap, which would be a crying shame. I have been down the route of having a commercial control fitted to my 3 axis mill as a full retrofit, servos the lot. It's a long story that i don't really wish to go into on an open forum. There is something that may interest you irving, regarding another thread i read about commercial conversational software.


As regards to demand, this is a subject i was going to touch on. I do feel there is a demand, I can not be the only small business that hates commercial controls and the idiotic prices they charge. Plus the problem of no longer available parts and what is available is a fortune + service engineer. I do think anyone that could cobble up a control and all the ancillaries could well find a nice niche. As you say, for another thread.

I know its not as easy as 1-2-3 as you say. I think the thing to remember here is all commercial controls are g code based, every single one uses g code in the back round, that includes all conversational controls. The only difference is they have given a graphical way to input the data so its easy to understand, so rather than having to know what g code you are writing, you get a graphical prompt. Its a little bit like the macro facility in mach 3 (that i have seen) where you can say i want to machine pocket and then configure the macros for all the variables related to a pocket of any given type. Thats all the conversationals are. The g code you guys have on your software has all come from industry standard and it is totally the same.


I know that does not account for integrating hardware and making everything work, but its a start.


I'm in in the Maidstone area of kent.



Kind regards,


Scott

Gary
06-09-2010, 03:14 PM
This is a nice lathe, and it would be a shame to do nothing with it.
I have a Hardinge lathe that i am updating, but i will be also changing the servos from brushed to AC servos.
the initial problem you will have is the interface to the drivers.
A lot of the old servos used +/-10V and this will be a problem with most of the PC based systems.
Then you have the PLC functionality that you will need for the tool changing turret.
This will require I/O, and even without the turret the lathe will need quite a bit of I/O, so will need additional hardware for the I/O and a dedicated CNC controller has PLC and a lot of I/O built in.
you also have all the Estop, collet opener coolant and other switches and pendant control already built into the panel.
With the PC based system, i would need a faster system than the parallel port, so looked at the smooth stepper, but was put off from the feedback it has had, i would also need some sort of modbus I/O expansion to handle the additional I/O.
Also with Mach 3 it would be open loop, so the controller would not know if the axis was out of position.

When you take all of this into account the costs between a PC based system and this dedicated CNC controller are not that far from each other.

I know someone who converted a hardinge lathe the same as mine to use Mach3, but it was not that straight forward, but can be done.
I considered Mach3 for the update, but then decided to go the with a CNC controller.
The controller that you pointed out would work, but you would most likely need to update the drivers because it only gives out clock and direction.
There are others that can use +/-10V but they are a lot more expensive, so it would most likely be better to change the motors and drivers.

ptjw7uk
06-09-2010, 07:34 PM
I think you are worrying about something called obsolescence which we can do very little about.
Even if you go the route of mach3 etc you could still be left with motors and drivers that will eventually change hopefully into something better.
My own take on the subject is that in a working environment I would go for a dedicated controller system have alook at the http://www.eaziform.co.uk/ systems they can use drivers from a variety of sources or they make their own drivers.
No doubt you caould use a mach3 system as many have done but I have my doubts about using a pc in a factory environment.
I know from personal experience that a PC in a factory environment will suffer from some unusual mains variations which cause alsorts of problems.
I wish you well but I suspect you reaching for the moon!

Peter

Swarfing
08-09-2010, 11:25 PM
The factory environment should not be an issue for the PC after all an office block with a thousand PC's running and all sorts dust clogging them up. Roger on here is doing something with Modbus, is absolutely sold on it for his mill control (converted full size job). As for obsolescence i have 1938 southbend which i have just got running again, i keep looking at some spare steppers i have?????????????

battwell
13-11-2010, 01:43 AM
mach3 is an extremely powerfull program (not seen at first)
it can nicely run old dc servo drives etc etc with the right hardware. took me a lot of research but well worth it.
i use the galil motion controller with mach (80 input/outputs, analogue servo control,analogue vfd, analogue input feedback etc)
The machines i convert are big! 5 tonners but now ive got the speed plus accuracy i need and for peanuts
see my utube to see what this control combination can do with a bit of vb programming (which i had never done before)
i dont know if everything works properly with lathe but the mach plugin guys will iron out any bugs for you as found. they have been amazing with my projects which i would say are 99.8% right now
http://www.youtube.com/user/automateanything?feature=mhum