Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: piv on February 16, 2011, 03:34:43 AM
-
I am about to do an upgrade on a big old Mori Seiki TL5 lathe. Its about 1100mm between centres and swings about 500mm. Its got an 8 place auto tool changer. Physically its a good machine. The controller is an ancient old Fanuc 2000. Thats the kind with LED lights that show which G code is current. We have actually had it going a few times but getting the steam pressure (just kidding) just right is a real trick. Its time for a new controller. I have just put DSPMC on a pair of five axis routers and it works pretty good. I am going to use DSPMC to control this beast. Right now I am pondering whether to use the existing motors that are big old Fanuc Gettys DC units and try to figure out the million wires on the control board to use the existing amps, or pull them off and put on some AC servos and new drives, which of course will cost lots more $ but hopefully be more reliable. Also I have to write some brains or macros to run the solenoids to control the hydraulic tool changer and figure out what goes on for the gear change. I also have to decide on keeping the 20hp DC spindle drive and the mitsubishi spindle control unit. That might be worthwhile. If I pull that out it will get about a 5kw AC servo instead. 20hp is probably a lot more than I will ever need on this machine. Or has anyone found good (cheap and reliable) DC servo drives that can run these old motors, id say they are 15 to 30 amp, but no idea of voltage.
So the question is, has anyone else ever done this to a TL5 or Fanuc 2000 or similar vintage lathe and does anyone know much about these things. Frankly looking inside the control box is scary, its like looking at a museum piece. As tempting as the money saving option is, I really think I will save myself a lot of pain by just putting in new electronics, cables and maybe motors. All comments, info and referals to the loony bin are welcome. I guess the other question is how far should I go in replacing things like limit switches? Of course that would end with a new machine, but if anyone else has gone down the conversion path, Id like to know. Also does anyone have a detailed idea of what else the PLC might do like run the lube pump etc. If not I am going to have to figure out all the things that need to do something.
-
Sounds like its a similar sized and vintage lathe as mine, mine was hardly used and had a GE 1050 control.
At the time I did mine DSPMC etc were not available and the only StepDir to analogue converter available was the pixie and it was seemingly a hit or miss and availability was patchy at best.
In the end I went with AC servos, have a 10Nm on the Z and 7Nm on the X both connected 1:1
The spindle originally had a 17Hp induction motor which went through a gearbox and the gear changes were done via electromagnetic clutches, I have since picked up a 12Kw AC servo and use that on the spindle with good results, however I dont think I would go much smaller than that.
I thnk if I was in your position I would try and keep at least the original spindle motor and drive and possibly even the original axis motors and drives. There are Dugong Drives from CNCDrives,com which may handle your motors but they are step/dir so no use with the DSPMC as far as I know.
Originally my lathe had a 6 position turret on the rear and 4 position toolpost on the front, both were totally hydraulic. I have now swapped out the toolpost with another 6 pos turret which again was hydraulic but I have converted that to servo and it can now rotate both ways and doesnt have to clamp at each tool like the rear turret does.
Limit switches etc should be good quality and I would keep, also see if the DSPMC can do index homing as it will be much more accurate than homing to even a good quality switch.
I use a PLC on mine for controlling the turrets and coolant etc, it also handles the drives I/O and all the switches I have on my panel.
Try and keep as much of the I/O as you can at 24V as it will make noise a non issue.
Would like to see a pic or two of your lathe. ;)
Hood
-
I have been working on Mori's since 1980, Your machine is about a 1977/78. The main spindle motor should a Mitsubishi or Fuji and will be a 17.5 Kw/25HP. Just had one rebuilt for a customer at a cost of $3500. If it runs leave it alone, and the drives until they break, It's not worth spending tons of money on a 1978 machine unless you are going to regrind the slides because I know they will require grinding after 30 odd years of hard use and cast iron on cast iron. Turret is electric driven, Hyd, clamp, well from 1979 on, through a Geneva gear and about 4 proximity switches for the count, I would put an encoder. The gearbox in the head is no problem, 2 electromagnetic clutches 24vdc. Not sure about the Gettys Fanuc motors, they could have tach and resolvers.
I have lots of customers with Mori's that I want to put Mach3 with DSPMC and I am in the testing period as we speak, but I am having some issues and Hood has been kindly helping me along with some of them My PC integration sucks, but I do know machine tools. DSPMC is giving me the most problems.
I should have a bunch of old data for the TL's, let me know if I can help.
Ray.
-
Hi Hood and Dresda, thanks for the reply. Next step for me is to get out the books and figure out just what is there and pull off a few covers to check out the tool change mechanism. Id like to keep as much as practical, but will do it in a way that can let things get replaced with new when the old stuff stops working. Ive built quite a few new machine tools, mainly big five axis routers. The advantage with new designs is you get what you want and can find the information. This is my first go at an old one.
Dresda, I am happy to share info and PLC code with you (not that Ive got any code for the lathe yet). Ive got a young PLC engineer who does some work for me, and I help him learn about CNC machine tools, so if you need help, let me know. The best suggestion I can make here is start simple, put the minimum functionality in the PLC to get you going, then you can add in the bells and whistles.
What problems are you having with DSPMC, maybe I can help a bit? Ive used both versions, 7761 and 7762, the new 7762 has encoder debounce which is nice as the five axis routers have a real noisy VFD on the spindle that was corrupting the encoders with the old DSPMC (7761).
I will be using index homing, that does work good on the DSPMC.
If and when I get the controls sorted out, we will then do whatever mechanical work is needed to get this beast back in shape. The ways arent too bad, but a machine this big and strong is worth fixing as an equivalent new machine is well over $100k, and we can get the work for it. Of course a good second hand machine can be had for $50k, but then whenever something goes wrong its a $1k to 10k exercise to fix it anyway.
-
Hi Piv, The first problem I had was over a year ago using the 7761 hooking it up to Fanuc motors and drives.
The Fanuc drive I was using was the 3 in 1 type, kept getting over current for some reason, the motors were not hooked up, then it was all over, drive pooched.
I thought is was fault until I spoke with the guys at Machmotion, it was the 7761 unit, the new firmware upgrade supposed to fix that,$2000.00 to repair the drive.
Now I am testing some SEM dc servo motors with pulse coders and tach and at 40"/min when I hit feedhold the Acc/dec is so slow it's pathetic, Rufi said he will try to repair it soon....
I even ordered the new 7762 unit because I have a customer with a 60" lathe that he wants retrofitting and I had to put it on hold.
I posted a few days ago about the Acc/dec problem title " .300" overrun at 120"/min".
When I rapid and stop it's fine, but not in program.
Other problem is when I run the famous roadrunner program and press program stop (I know I shouldn't use program stop) and restart it moves to the block look ahead postion and carries on without loosing postion, so it skips like 100 blocks or so. I tested it on a desk top with just Mach3 installed no DSPMC and the program just restarts right up from when it stopped.
Ray.
-
Hi Dresda,
Sorry I cant help out with those problems but I do agree they are real. They will be a problem on the lathe, fortunately they are not a big issue on the routers because wood is a lot softer than steel. The feed hold is a bit of a problem but our work around is to just decrease feed rate to zero using the FRO.
The program stop thing, our customer backs up a hundred lines of code and starts from there. Not ideal but liveable. The last controller we couldnt start mid program so its luxury now for them relative to before!
Im still pretty happy with the DSPMC, I think it will get better too. There are a few issues with it but at least they do get fixed. Ive had bad experience with a few other controllers. Smooth stepper and DSPMC are both pretty good other than the few issues they have.
Ive had some of my drives (Baldor Flexdrive) show overcurrent too. But I thought it was in the drive, because the drive is supposed to manage its own current limits, so theoretically it shouldnt ever let itself overcurrent and if the controller asks for too much current for too long, the drive current should foldback. How did DSPMC cause your problem, and more importantly, is it actually fixed?
I guess this highlights the issue that just because something works in one application doesn't mean its good for all. Some things are important to some poeple and not others and there are lots of things that have to be right to suit all applications. I guess thats why CNC costs so much to get bullet proof.
-
our work around is to just decrease feed rate to zero using the FRO.
That is the way I do things, I have a FRO pot on the panel and if using new code I wind that down and up as needed. One thing to be aware of is Mach will never actually stop even when the DRO reads zero, reaon is it never actually reads zero ;) It will go down to something like 0.1% if I recall. Normal day to day this is not a problem but if you wish to set and hold on a FRO of 0% it will creep along so could cause problems if left for extended periods at zero. I have asked for this to be changed in Rev4 if possible but not sure if it will be.
The way I get around this is I have a Brain looking and reading my analogue inputs from the pot, this is how the FRO is set, I however also have it looking at the % value and if below 1% it will issue a feedhold and thus will actually stop, when the brain sees the FRO% again increase then a cycle start is issued. As the FRO has been turned down the feedhold is almost instantaneous but that may not be the same with the DSPMC. I also have the Brain set 100% if the FRO is between 98 and 102% this just helps setting it to 100% and allows for any slight variances in the analogue signal.
Hood
-
Thanks Hood,
Dresda could make a work around (fix) for his problem by making a brain that reads the feed hold switch and starts to ramp down (using a timer loop?) the feed rate over ride and then sets the feed hold when FRO is < 1 % say. Its clunky but could work. Probably easier for vital to fix the real problem.
-
I would think the best way was to set up a pot for FRO and use that for testing out new code, great thing for me is I can wind it down as it approaches the part, once I see its fine can wind it back up again. Also because I have the FRO and RRO linked then the rapids are also overridden and for me it is really the rapids that are of most concern when testing out new code, feeds are usually no more than 0.35/rev so even at full spindle speed of 2000 on my lathe that is 700mm/min where rapids are 8000mm/min.
I am not sure but I think the DSPMC has analogue?
Hood
-
Standard DSPMC has no analogue in but you can get it as an option. Or there are lots of digital inputs so a grey code or even use an encoder and feed it into one of the encoder inputs.
-
Ah Ok thought the DSPMC had them. I use current type modules on my PLC for it, got a few very cheap on eBay :)
Hood
-
Hi Hood,
What PLC are you using and is practical to send me the code as a starting point.
-
I am use a DL06, specifically the D0-06DR from Automation Direct.
What code are you wanting? The PLC ladder?
Hood
-
Hi Hood,
Yep the PLC ladder and any brains that you have done. I will probably do it as macros and brains and just use the DSPMC, but the ladder will be a big help in doing the brains.
-
I think this is my latest ones, not at the workshop so cant grab them off the lathe and the netbook which will definitely have the latest is aldo at the workshop.
Hood
-
Thanks Hood.
-
No probs, if you cant make sense of anything that will make two of us ;D Seriously though, just ask.
Hood
-
Guys, As I said before, I can live with feed hold and I do drop the FRO down to Zero before the next cycle, but not everbody does.
After my mill has been tested I will be doing a 60" chuck lathe with Fanuc motors and drives so you can imagine the size of the job that's going on it..I need to stop the slide in a reasonable time and this is unreasonable. Rufi did say the acc/dec is set in the DSPMC and he hopes to make it the same as Mach, I thought It would be just a parameter change. I can show you cheesy cheapo 2 axis PC machines that can stop on a dime. It can't do your cutters any good the way it slows down then ramps up again.
Ray.
-
Yes I agree, it should be made right, its so close to being good and it is a very important thing to have control over the machine, especially big metal cutting ones.
Ive had a quick look through the machine books. Fortunately I have them. The DC motors have pulse coders and tacho feedback. Ive found the pin outs for the velocity control. I havent found any pin outs for the spindle drive. The servo drives are SCR. They give a lot more info than I thought they might.
It looks like main axes can be controlled by hooking up the DSPMC analogue out to VCMD on the velocity control units and then finding the old pulse coder input to the "position control board" and using that as encoder input.
The biggest drama will be figuring out the IO, maybe I just have to trace it back.
Hood, you took out the motors and drives and put in new ones, was that because the old ones were dead or did you just want new ones, or was it getting too hard?
-
Piv, Is this what you in have in your spindle?
VRDY1 & 2 (velocity ready)
TLM5 TLM6 (torque limit)
OVL 1 & 2 ( overload)
FA FB (fuse alarm)
SDRY 1 SDRY 2 (servo ready)
TH1 & TH2 (thermal)
TLM1 & TLM2 ( it says pin in orient)
TSA TSB (tach)
I will look for the motor pin outs, as I said I did have my motors running, but not for long.
Ray.
-
Hi Dresda,
Thanks but I couldnt find those terminals (yet?)
The spindle drive is a Mitsubishi Diarol/DL-SCZ-22K-D Serial W20133005-7803 Is that March 78?
Under the front cover the signal wires (maybe?) go to TB6 and TB5
TB6 has
Terminal Wire
N12R 158
N12
SET SET
SRN SRN
SRI SRI
DEF 159
MAN MAN
SRO
FC R11 LOOPS TO ZSI
FB 119
FA 120
ZSI R11 LOOPS TO FC
Z32 128
VR1 N1
VR2 160
USC 250
USB
USA 175
CDA
CDC
TB 5 HAS
SD5
OR3
OR2
OR1
Just as a joke it would be nice if I could plug my PC into USB javascript:void(0);
There is another connector con53 that has a few wires going to behind the top pcb.
None of the above means anything to me. Ill have a google search for a manual or schematic.
-
Has anyone got a manual or schematic that can help me figure the pinouts for this spindle drive
Mitsubishi Diarol/DL-SCZ-22K-D Serial W20133005-7803
On a Mori Seiki TL5 lathe with a Fanuc 2000 controller. Either a schematic for the spindle drive, or schematic for the Fanuc control connected to it would be real helpful.
-
Hood, you took out the motors and drives and put in new ones, was that because the old ones were dead or did you just want new ones, or was it getting too hard?
The drives and motors were like new, I reckon the lathe had done under 50Hrs total. It came out of a University in Hull, UK and had been used for testing the differences between CBN and Ceramic insert life. I found a paper on the net that detailed the tests and the results, was quite interesting :)
There were two reasons I didnt use them. First was that my friend bought the drives and I bought the rest of the lathe, he has older CNC's that use HiAk drives so he wanted them as spares and the second reason was that when I did this there werent any devices like the DSPMC about so no way to convert the Step/Dir of Mach to analogue for the drives. There were the pixie boards but they were seemingly a bit hit or miss whether they would work and also supply was patchy.
Pic below of inside of motor showing no signs of carbon from brushes, also a pic of the old drives and finally one of the old control panel without the slightest mark on it :)
Hood
-
Wow hood, that was like brand new. I will get some pictures of my beast.
-
Look forward to them ;)
Hood
-
The guy that got my drives just came up with a manual for the spindle drive, its too big to attach so if you pm me your email address I will forward it.
Hood
-
Your old memory must be failing,Hood.ISTR I sold you the lathe less the drives and control. :D
For anyone who wants Mits manuals,go to www.meau.com and click on downloads then manuals.
If you are located outside of North America set the time zone on your computer to an American one first.
Mark.
-
Its a wonder you can reember anything at your age ;)
Hood
-
I can remember how to spell remember anyway.
This is like taking sweeties from a wain. ;D ;D ;D
Mark.
-
Wain, is that Weegie/fffb for wean? And yes I have heard that is a favourite past time for Fifers.
Hood
-
What does fffb stand for,o little fat one?
-
Sorry Piv, That was the old Fanuc spindle drive, I will ask my buddy see what he has.
Hood, No wonder you changed the control, Mark century, I know guys that have not only run a mile from those, they are still running...
Ray.
-
The manual can be downloaded from Mits site that Mark (Gridley51) posted a link to, seems to be quite a lot of good info in it.
Yes it looks kind of antiquated, did fire it up and got a couple of things running quite quickly but that was due to Mark being familiar with GE controls, more modern ones but still I would imagine the basics are similar.
Hood
-
I would be impressed with Mitsubishi if you find your manual on there site.
-
Are you meaning Piv's manual?
Hood
-
See if this works, you have to have the time set for North America to get to the downloads page but a direct link should work fine without setting the time.
http://www.meau.com/functions/dms/getfile.asp?ID=010000000000000001000000458800000
Hood
-
Oh, Wow, that's great..
-
Post some pics.... Should move to the machine builds section?
JH
-
Piv, Do you a electrical drawings for your machine? You know what the 22K is for, 30HP. I have drawings for DL-SCZ 18K-D (18.5 KW) 25hp.
Ray.
-
Hood, Thanks for the link to the drive manual, it worked! I did have a look before on the meau.com site, but didnt know about setting time zone to North America. Thats going to make this conversion possible. Page 68 (connection diagram that actually matches what I have!) is gold. It looks like I have the option of in putting speed as a 12 bit digital input or as an analogue voltage. Thats nice.
Dresda, I guessed the 22K was for 22kW oe 30hp. I dont have an electrical schematic for the machine. Do you have one? That would be a time saver too. Otherwise its lots of wire tracing.
So now I feel like I can use the existing axis and spindle drives. My next step is to figure out the existing "PLC output" to the relays on the machine so I can work out what my new PLC output from the DSPMC will be.
Thanks everyone for your help.
-
Piv, its Mark (Gridley51) you need to thank, I only asked him if he had the manual as I know he has a lot for all the machines he has.
Hood
-
Hi Hood and Gridley 51, thanks, and thanks again.
For anyone else looking for a Mitsubishi Spindle Drive Manual or Diarol Manual or SCZ manual or SCZ schematic then go to www.meau.com and set your time zone to North American then you can download it.
-
I have an electrica drawing from Nov, 1977, don't know if it's mori TL5, I bet it is. Control Ser # 393.
Ray.
-
Hi Dresda,
That would be fantastic if it can be emailed. I just PM you with my email.
-
Update on the TL5. Simon has been working on it and we now have X and Z axes moving, spindle turning and the tool turret indexing and most of the IO working. To give an idea of the work involved, Simon has been at it for 7 weeks full time. He is a Mechatronics engineer, new to Mach3 but I have been helping him out with lots of stuff. The big problem is the lack of complete documentation and working on a machine that has a had a few undocumented changes and repairs over the years, plus the whole Fanuc 2000 and spindle controller and drives are so old they are like they from a different planet the way stuff was done. Any way we are using the original Fanuc servo drives and Mitsubishi spindle drive. Currently we have issues with not seeing mpg input in mach3 (shows up as bit changes on the input lines in DSPMC status window) and figuring out a screen set to work with the existing hard switches on the old operator panel. Any help appreciated.