Older controllers

[flatbottom]

A friend of mine, and myself, are looking to buy an older cnc mill. After wading through a wealth of information in the forums, I have one question. What would be the pros and cons of a controller on an older machine?

We are looking to purchase a machine with small scale production in mind. At first, it will be used more for one offs, since my friend is going to make some custom parts for a motorcycle project he's been dying to do. Then later on, it would be used more for batching out small runs.

We have been looking at machines from the early to late 90s, as a start point. Most likely, a Mori or Okuma. My friend worked for an aircraft manufacturing company for several years, and likes the Haas, Yasnac, and Mitsubishi controllers. He hates Fanuc. As far as I'm concerned, the type of controller doesn't matter, since I have no experience at all, and will be learning it fresh. I'll be the one doing the CAD part, since I know a couple of CAD/CAM (mastercam, qccad) inside and out. I will also have to run the machine at times when my friend can't.

My biggest concern is will an older controller be able to handle the large files generated by the newer CAD/CAM applications. I know memory in the older controllers will be limited. Drip feed? Will accuracy be as good as the newer machines? Also, I was wondering what types of interface problems might arise from an older controller and newer desktops. RS 232 shouldn't be a problem, but as I have found through the years, its when you think there's not a problem, you've got one.

I also worry about parts associated with the control system on an older machine. In some cases, parts may be non existent or hard to get. What controllers still have a good parts inventory available? Software and firmware ware upgradable, and so forth.

Many thanks in advance.

[Al_The_Man]

'Early to late nineties' are not really that old machines, so you should not have that much of a problem as far as memory, or parts come to that, if you are looking at 90's machines, especially if it is a big name that is still around, and even for 70's~80's machines, brands like Mitsubishi will offer rebuild or repair parts, if the original is not available.
You may want to stay clear of control makes that are not around anymore, if one of these becomes a 'basket' case then your only option would be to retro-fit if you cannot locate parts.
You should be OK with the likes of Mitsubishi, Fanuc, Fagor, Siemens, Heidenhain, Mazak's are nice machines used mostly Mitsubishi control , some were Fanuc.
Some of the older Cinncinatti's parts are getting scarce, but the machines themselves were built to last.
Al.

[flatbottom]

Many thanks Al. We are currently looking a a 92 Mori. Its got the Yasnac MX 3 control. I went to the Yaskawa web site, and they offer remanufactured control units. They also list options available for that control, so I've emailed them to see if the options can still be added to this used control in the feild, or will it have to be sent in for the upgrades.

I got to thinking about G codes also. I would assume that G codes have evolved over the years, and would a newer CAD/CAM program generate codes in the NC file that an older machine might not recognize. I would also assume that M codes have evolved also, but to a lesser extent.

Hmmm. Just more to think about.

[Al_The_Man]

G codes have remained pretty standard, except for the canned cycles, but Macro programming makes G code more powerfull, M codes are machine specific and are written by the OEM not by the control manufacturer.
Al.

[ratbike]

Hi,
I have recently purchased a Victor TNS 2 not in bad condition for its age, being a 1985 machine, however the Fanuc 3T control is set so that the X values are always programed as minus values, this is very confusing as I have spent my working life using positive values for X. Please could anyone provide me with the correct parameters to alter this, or tell me how to enable the G68 mirror image option ?
Many Thanks
Ratbike

[Fireflowers]

Hi,
Could you set the work zero at the chuck (G54/59) then all your figures will be positive working back from the chuck along the workpiece if you are programming in absolute (G90) during the cutting moves.

John.

[ratbike]

Hi john,
Turning controls usually do not support the work coordinate codes G54 through G59 they are normally for machining centres. However on this control you do have to use a G50 which is the equivalent and loads the X & Z distance from reference point to datum point on the workpiece X0. Z0. which is effectively chuck centre line, so no that doesn't work but thanks for the suggestion, it is the way the machine has been setup parameter wise initially.

Very Best regards,
Dave

[Fireflowers]

Ok Dave,
Should have checked this! I've just bought an Hitachi- Seiki 4NE11 with 6T .
It says in the manual that for reference point return and back lash comp the parameters are 012. Reading from the left if the first two digits are set to 1 (ZMX ,ZMZ) then the reference return value is a minus ,ie turret going to home position, so i presume the values travelling towards the chuck should then be positive. If those two digits are set to a 0 then the reference return is a positive.
Not sure if 3T is the same but hope this helps anyway,
Cheers, John.

[ratbike]

Hi John,
You lucky person the 6T & 6M is reputed to have been Fanuc's finest control and on a Hitachi-Seiki to boot. Anyhow back to my issue Fanuc parameters change from control to control so the 6T listing would be different from the 3T as it happens I do have the 3T listing but it doesn't describe all the listings as in 995 through 999 these I suspect are the options parameters as with cautious exploring I have been able to switch on the geometry offset file. Parameter 996 bit 5 count from 0 right 00100000 replicate all others as is. Which allows me to do away with loading G50 values in program and makes the control more usable but still X-. The parameters that you describe on the 6T. Do control which way during reference point return the machine approaches the reference point return dog or switch and not anything else. but thanks again for looking, & do you use the geometry offset file when setting tools or the dreaded G50 in prog
Best Regards
Dave

[Dan Fritz]

ratbike: The X axis can be reversed permanently, but it takes a little surgery on the X axis pulse coder cable. If you reverse the direction of the pulse coder feedback and also reverse the armature wires to the X axis motor, the servo will run "backwards". After that, there are a few little details to worry about, like the direction of zero-return and the layout of the jog buttons.

We used to have this problem years ago when my friend Terry sold Mori Seikis here in N.E. Ohio. All the Moris would be "backwards" from the other lathes on the market, so the X dimensions were all negative if the spindle centerline was X0. Actually, the coordinate systems were the same on Moris, but the tool turret was "upside down", so your perspective of the cutting tool was the same as if you were lying in the chip pan of a "brand X' lathe looking up at the spindle. A few of Terry's customers wanted the X axis reversed, so we did this for them:

The servo pulse coders have 6 signal wires, plus a couple of power and ground wires. The signals are called the "A pulse" and the "Not A pulse", then the "B pulse" then the "Not B pulse", the "Zero pulse" and the "Not Zero pulse". On the schematics, they are labeled PCA, *PCA, PCB, *PCB, PCZ, and *PCZ. To reverse the direction of the pulse coder feedback, you must swap the PCA and the PCB wire, and also swap the *PCA and the *PCB wire. This can be done at either end of the cable, whichever is easier to get to. You must solder these connections.

On the 20-pin Honda plug end of the cable (plugged into the CNCs main CPU board), the pin assignments for these pulse coder signals are:

pin 16 -- PCA <--- swap pins 16 and 18
pin 17 -- *PCA <-- swap pins 17 and 19
pin 18 -- PCB
pin 19 -- *PCB
pin 14 -- PCZ
pin 15 -- *PCZ

After you do this, DO NOT TURN THE CONTROL ON until you've also reversed the armature (DC) wires to the X axis motor, otherwise the servo will "run away" really fast! The armature wires are attached to the terminal strip T1 at the bottom of the servo amplifier (on the back, below the servo's fuses). There will be 2 wires for the + side of the motor amature, and 2 wires for the - side. You must swap both + wires with both - wires so the motor's polarity is reversed.

Once you do this mod, you can turn on the control (keep your hand on the E-stop button, just in case!) and you will notice that the X axis will jog backwards. The X position display should go positive as the tool moves away from spindle centerline. Now, you must reverse the direction of the zero return parameters that you mentioned so that you can zero out the X axis in the new + direction. If your +X and -X jog buttons are confusing, you may want to swap them as well. Just look at the back of the switches and find the wire that changes from 0v to +24v when you push the button. Swap this wire for the similar wire on the other button. You can then swap the buttons on the switches so that +X and -X are correct.

Be careful, and Good luck !

[Fireflowers]

Thanks Dave,
I also have two Bridgeport BPC 320's with 6MB, there really nice to use as the 6 control can be made to do just about anything with a bit of patience!
The Lathe isn't wired up yet so I'll be looking into using the geometry offset file as an alternative as you have suggested.
It looks like you have your problem sorted anyway, good luck!
John.

[ratbike]

Hi Dan,
Thank you very very very much for your wonderfully detailed solution, I like to think I am quite good with the old soldering iron, have a good one Dan, in fact have a better one. How good is CNC zone with all the brilliant people like this feller taking part, I have been running CNC's and automatics for the past 25 yrs, and owning them personally for the past 12 yrs never had so much help Thanks again Dan, and every one else who takes part.
Very Best regards
(Ratbike) or Dave