[lerman]

...but probably not today.

I did a milling machine conversion using picosystems controllers (see: http://www.pico-systems.com). That's about $300 for the controller plus $100 per axis if I recall correctly. The controller can support up to four servo amps. One of the four channels could be used for speed control. This would be a very cost effective solution for you.

Unfortunately, lathe support is still in it's early stages with EMC. But I believe that threading has just been made to work. I suggest that you subscribe to the EMC user list and state your requirements there. The list is populated by a group of dedicated volunteers. I believe that some of the members also do retrofits.

You might wind up being the catlyst that gets lathe support to be useful.

I'd understand if you didn't want to get involved with that. But at least the cash risk is relatively small.

Ken

[Al_The_Man]

Shizzlemah:

In a Motion Science servo amp like that in an Ezpath or Eztrak, torque is hard limited by a current sense circuit in the servo amp.
There is NOT a velocity feedback in these systems, only a position feed back from the motor directly to the PC. Current/torque is limited by the amp itself.

This refers to current limit, I believe arr5ow was refering to torque control, if you do a search for transconductance amplifier it should give a clearer explanation, this is current out directly related to a voltage in,
A transconductance amplifier outputs a current (torque) proportional to its input voltage.
Al.

[Shizzlemah]

Okay, I guess, it seems like a silly description. There is of course a current limit, and IV does result in torque.

But how is that different than any of the common platforms? The centroid uses a plain jane encode feedback.


NC CAMS - a centroid hack literally was screaming at me once that of course I must be able to reuse my encoders. They were selling a package to refit a machine ONLY available with resolvers. And became belligerent when I said I didn't have encoders. "Well how is it gonna work without encoders ?!" Yeah, well, um I guess you are a well skill tradesman. Maybe you can show me how to get solitaire on the system...

BTW, 2 yrs and I still havent got my unlock codes. Centroid wont help since I bought it from a dealer. Dealer is a tool and kept my money but generates lots of excuses.

[Arrow Control]

The Centroid does use encoder feedback , as long as encoder has A / A* B / B* Z / Z* , works on 5V and has a decent pulse count >1000 . There motion control card outputs its servo commands out a fiber optic cable which goes directly into there servo drive, You can use ther OPTIC1 card to take the fiber signal and convert it to a 0 to +- 10V torque command. the motion control card will output a signal that relates to current not velocity/speed. If the control senses it is out of postion it will increase current output until it is satisfied. If during a rapid move once machine has caught up to control the current output will pull back down, whereas on a velocity system it would stay up.

[Arrow Control]

[QUOTE=NC Cams]Dear Arr5ow Control:


I advised that I had 90vdc motors that drew 12- 15 amps - "they won't work with our amplifiers".

I advised that I had 2000ppr ENCODERS, not resolvers - "won't work, you need new encoders" DUH, why????

It does not take a rocket scientist to take the 3 phase isolated AC power to the 5T servo amp, add a rectifier bridge and a capacitor to make DC suitable to drive my DC servos - you can't convice me that your DC servo amps would never know the difference if I did....

Re: too high a bus voltage. There are voltage taps on the isolation transformer to adjust the voltage to nearly whatever you want - my system was an international model and if you ended up with 110vdc, reset the tap to a lower turns ratio and problem solved. If you retrofited 5T's, you or the folks at Centroid should already KNOW that.

PROM lader logic - sorry, that doesn't wash. My 5T has ALL the Fanuc signals dumped out to a huge relay lader logic board. You could run the darn thing with push buttons it you could push them fast enough or in the right sequence - in fact that's all the 5T does at its outputs - trigger PCB mounted relays. if your ladder logic worked and put out the appropriate signal via NO/NC relays, my system would never know the difference.

Spindle speed - simple D-A converter. Give the spindle drive a voltage and a direction via closed relay, off it goes.
************************************************** ***
Not sure why you were told Motors and encoders wont work - they will.
I cant answer for what you were told just will be happy to answer what ever I can.

I am fully aware of the servo transformes ability to have taps changed , the only thing that would affect on the Centroid drive - is they are expecting a higher voltage coming in and if you drop the voltage too much it will trip an undervoltage alarm. Possibly system could be ordered with modified resistors to trip at a lower voltage. Other than that a lower voltage is fine.

Spindle speed out of fanuc controls are actually 12 bits, The Centroid outputs an analog 0-10 out of the PLC to run spindle - it is only + though , so spindle would have to be able of reversing off an input as opposed to speed command going - . If the spindle drive uses the 12 bits and converts it with a D/A in drive ,then Centroid can be built with a card to provide those 12 bits - I actually was one who setup Centroid with info to do this.

Most all 5T controls did have a ladder burnt into chips, some systems were made with all the ladder functions outside the control on a relay board. The Fanuc then would just output M / T codes M11/12/14/18/21/22/24/28 ect and relay panel had to decode them and handle finish codes. The Centroid does not output these codes in BCD format as FANUC did. But with some programming in control could simulate it. My opinion has always been it is best to remove that relay panel and do it all in software for a reliability issue.

You goal of a 2-4K cost system is pretty low and would be tuff to accomplish

What kind of a machine is it your trying to do?

Terry

[Arrow Control]

One other thought about your servo motors , you never said what size the other one was. One was a size 0 , if the other one was a 0 or 5 OK , If it is a size 10 / 20 /30 then it runs on a higher voltage and since AJAX lowest cost drive is a 3 axis combined all axis must have same voltage.

[NC Cams]

Mr. Arr5ow Conrol: meet Mr. Shizzlemah, an obviously well satisfied Centroid customer.

Mr. Shizzlemah: your experiences were not wasted. I for one was inordinately thankful that you chimed in on this thread. You definitely saved me from reliving your experience(s). THANKS!!!! Ditto to others who conveyed their less than stellar experiences with lathe retrofits.

Mr. Lerman: and the world is replete with software that is full of promises but lacking on performance. By the way, have you by any chance met Shizzlemah???

He can vouch for the fact that even high buck machine control systems don't always do what they purport to being able to do. At least EMC would have been a lower cost disappointment as opposed to a high buck one. That would surely have beem a nice consolation...

The ONLY problem I have as a business owner with DIY/user supported/developed/open source systems is that when a machine doesn't work - I need service.

I simply can't rely or count on random support from a message board or development in process by what is essentially a hobby enterprise.

"Eventually it will work" features lists are neat to day dream by. Surely, DIY hobbiests can tolerate this -a functioning business simply can't.

It is bad enough that we all have had to deal with bug and security festooned software from Redmond (for decades now) that STILL isn't fully developed and/or bug free. Perhaps the trouble with computer software people is that they deal too much with cyberware and virtual reality.

A system crash to them is a rebootable inconvenience. A system crash for me that causes a CNC to botch up a cam billet or spit it out of the lathe could result in a $3000 boat anchor made out of tool steel - or severe bodily harm or death. BIG difference!!!

Cyberjunk versus expensive REAL junk in metal. Big difference. Perhaps EMC will someday do what it hopes to be able to do. Until then, it is as much vaporware as there ever has been created. If the folks who came up with it REALLY believed in it, they'd turn it into a real business instead of an iteresting hobby.

Until the creators become committed to the task by turning it into a real business, it remainds a part of what I call the irresponsible triumvirate:

The software guy blames the O/S guy who blames the hardware guy who blames the software guy for a NC malfunction when the stuff doesn't work.

Sadly, my customers won't buy the comuter industry's equivalent to "the dog ate my homework" excuses. Anyway, its about time to go kick the Fanuc 5T to wake it back up.

At least when it isn't suffering from old age and/or costing me small fortuns in cash therapy to keep it going, it does what it's supposed to do....

IF I weren't so fed up with the way the Japanese do business and the costliness of anything that has Fanuc written on it, I'd buy another one. sadly, my financial status in life relagates me to the lot I'm in in life and will just have to get bye as best I can with what I've got....

Oh for a cheap, reliable DOS based system that would simply WORK. There is charm in stuff that is clunky and reliable. Sadly, it falls into the 'no longer supported' pile of stuff that would still readily satisfy a LOT of the needs of many tight budgeted CNC'ers.

[NC Cams]

Arr5ow: The 3 systems in my shop are idential turn key Cadillac NC lathes with Fanuc/GN contols - 1979 vintage. Model 0 servos - full factory package CNC that is simply old. I"m tired of messing with it and was looking for a simple control, ala my DOS based Eztrak system that still does everything I need it to do.

You'd think that nearly 30 years after it (5T) was created, one could buy a contemporary computer that could would run it. All you need is to interface to the output relay ladder (it works and is was developed, so why change it???).

Output a voltage signal t0 the spindle, give it a closed relay direction signal and off it goes. Sadly, that's seems to be too hard. I don't need Windoze, I don't need multi-tasking, I don't need a LAN - my programs fit on a floppy for goodness sake.

All the lathes need to do is cut regularly spaced grooves into metal, over and over and over. With 3 lathes and 1 grinder, lathe thru-put is NOT my bottleneck. Hence, lightwarp cutting speeds are not essential.

Those familiar with Fanucs, especially 5T's, know that that EVERYTING in the cabinet is run off of its own isolated power multitap supply transformer. Each servo has its own transormer (2 in a lathe). It has to be this way due to the SCR driven AC direct to DC servo drives.

Thus, servo voltage is independantly adjustable as opposed to M/B as opposed to anything else. But as the Ajax website says, Ajax/Centroid has retrofitted Fanuc's so they already now that. Too bad they didn't teach it to the sales guy. His only mantra was 'it won't work'.

My DAC is just that: the Fanuc sends 8 or 12 digital bits (don't recall) to a separate D-A converter mounted to the side of the cabinet (aftermarket hang on) and a pure analog signal is then fed from it to the DC spindle drives (Randtronic or Cleveland Machine Control).

Apply voltage, close the relay, it runs CW or CCW. Can and did make spindle run with a 9V battery and a pair of clip leads. Too bad the Centroid can't - at least that's what the guy said on the phone.

All this info was supplied to the guy I talked to on the phone. He either didn't understand or his marching orders were to send folks like me over to Centroid or a dealer where they could make some REAL profit. BTW, still waiting for the local dealer to call which the sales guy said he'd arrange for. At this point, he needn't bother.

if Bridgeport could, in their infinite wisdom but poor business accument, come up with a DOS/PC based DX32 system that could run a lathe, mill, surface grinder and 4 axis VMC via a 386SX, 486 or Pentium 133 system in the mid/late 90's with an ISA card for goodness sake - that in many cases are still functioning today - why hasn't/can't somebody duplicate the effort in the way of an affordable and simple aftermarket machine control system???

Probably because it is "too simple-we can't make any money on it" (per the MDSI guy I talked to) and "nobody uses DOS anymore" that I"ve heard far too many times from computer geeks involed both in and outside the CNC industry.

Thanks for caring and trying to help. It's too bad the folks who's product you sell seemingly could care less about folks in my postion or with my needs/preferences.

[Arrow Control]

I know the machine - Have worked on Cadillacs with 5T's before - many years ago though. Probably still have a set of prints somewhere in cabinets. They are simple , turret setup leaves much to be desired all work off air valves that are prone to failure. The DAC board mounted on side of control cabinet has a meter on it right? Also look and see if there is a sticker on side of cabinet that says Machine sequence controller. The 5T has 2 transformers in it - one for control and one for servos, the servo transformer has seperate secondaries - one for each axis.
All I can say is, if you want me to help I can..

[NC Cams]

Arr5ow: got all the prints on pretty much EVERYTHING. I'd be dead without them. Thanks for the offer of help. Sadly, if Centroid gave their new hires OJT via answering Ajax calls, they lost themselves the sale of probably 3 systems....

Asside from a flakey BTR which I've recounted my escapades with and developed sollutions for, I'm in pretty good shape until the next time the darn things develop Alzheimers.

My "indexable turret" is a 4 position device made by a local guy in Jackson, MI - sort of a Dorian knock off. Simple and works. Nothing but a relay and an air solenoid. Give it filtered air and 24vdc - anybody who does air operated solenoids w/o filtered air deserves the grief.

Simple machine for simple tasks is all I need. Until I find a simple controller that doesn't need ME to program crap or dodge parts because of bogus machine control logic, runs generic G codes, I'll live with what I have.

i'm looking into some legacy stuff I stumbled upon but that's a spare time deal that can simply wait. It seems too simple for it not to work - so I suspect some corporate decision out there will prevent it from happening or make it financially impossible. Technically, it is readily doable.

Again, thanks for caring and trying. If you're in the market for Centroid, it's guys like this who you need to find and do busines with... Good luck in doing so...

[Shizzlemah]

" fiber optic cable "

Yep that's a great feature for literature. Fiber optic link to the servo unit.

That doesn't do a darned thing if you are picking up RF on the keyboard cable... But it's nice to brag about. Hmn, let me try that here :

I am the only guy on the block with a fiber optic boat anchor!

[NC Cams]

I wonder if the machine tool ontrol guys even heard of FCC class B noise conformance. Probably not. This spec is why PC's work so well at home. Yet some retrofit guys kluge up the cabinets, run pulsed DC all over the place and don't even follow rudimentary practices with regard to noise containment.

Want to know why Fanuc stuff is SO expensive??? The cabinets are replete with MIL spec shielded connectors, grounds and all. Not a DB to be found at interconnection between machine and controller. And people wonder why DIY stuff goes nuts. Simple hodge podge kluge of home PC stuff grafted to noisely industrial hardware. A wonder it works at all!!!

What do you expect a 15 amp, 100 volt unshielded DC pulsed signal at 20 to 100Khz is going to a servo motor will to to unshielded 5 volt low current PC signales anywhere's nearby???

If you can't listen to a $3 transistor AM radio tuned for max noise across the band, don't expect your unshielded 5vdc computer signals to remain unaffected.

Don't tell me: the system vendor told you they never encountered THAT before??? Yeah, encounter this....