If you take the computer and stick it in the same case with your control hardware. Then you take your computer Power supplie and ground it with your other supplies all at the same Star ground. Wouldn't this be removing the whole issue?
If you take the computer and stick it in the same case with your control hardware. Then you take your computer Power supplie and ground it with your other supplies all at the same Star ground. Wouldn't this be removing the whole issue?
thanks
Michael T.
"If you don't stand for something, chances are, you'll fall for anything!"
miljnor,
Depends on the motor power supply. Maybe, maybe not.
As NC Cams pointed out, depends on gnd. One thing can be said for sure, if you don't have opto isolation. Plug the computer, drive power supply and driver electronics to the same circuit. For low current a breaked outlet strip, as current goes up you may have to make your own.
Phil, Still too many interests, too many projects, and not enough time!!!!!!!!
Vist my websites - http://pminmo.com & http://millpcbs.com
Finding a 110v source in a prewired late 1970's vintage CNC controller that is packed to the hilt with P/S's, servo amps and all sorts of other stuff isn't always an option.
Besides, is far easier to plug the PC into the nearest outlet that most certainly is going to be wired to a totally different 1 phase transformer than the CNC which was connected to 440 volt 3 phase power, stepped down thru a 440 to 220 step down transformer by a psycho who twist connnected 18 gage control wire to lamp cord (literally).
THe guy who I bought the lathes from didn't have a clue about wiring codes or grounding. Why should he? It all ran after he hooked it up and it didn't smoke.
Until that one fateful day and WHAM and he's got this deer in headlights look as to "Why did it quit???" Oh, and this happened 3 times.
Result: I got 3 $40k (when new) CNC lathes for peanuts because of his unfortunate problems. His misfortune became my "opportunity". Spent a year plus fixing everything but 2 are now like new and the other has a lot of "shop character" - and all have opto isolated BTR's hooked to recycled DOS pc's.
So what your actually saying is Dont opto isolate so you can gets some good deals on used machinery!![]()
thanks
Michael T.
"If you don't stand for something, chances are, you'll fall for anything!"
Hi Phil,
regarding your dizgram (post no 5 of this thread). Could you post a quick picture of the correct "star" grounding pattern to use in that same setup (optoisolated BOB, 3 driver boards, power supply and computer).
Thanks
Steven
miljnor: I''m always amazed at the exceptions, in some cases arguments, that people take to "hard earned" lessons shared to prevent folks from making the same mistake I did.
You can do what you want - opto or not. Just don't be surprised when/if those untrustworty electrons go where they shouldn't and let smoke out of IC's. If you wire the whole thing in a home shop environment with 1 phase 110/220, you can get away with relative murder.
HOwever, get into an industrial with 200 amp 3 phase and you can get into real trouble real fast. Been there done that and still fixing stuff (see untrustworthy electron comment elsewhere).
Life is full of opportunities. Some you are creator victim of and some are/were created by others for an astute (perhaps better educated person) that you might be in a position to take advantage of.
Yes, I got a GREAT deals on some used machinery due to someone else's misfortune. HOwever, I"ve spent a couple years redoing the wiring on the beasts (another long and sad story about the guy who created my opportunities) and finding folks who can/will service literally like new CNC lathes with 1979 controllers.
It is even more fun to fix the stuff, and hear the guy you bought if fromsay "you can't do it, I tried" (all the while he's certain that he screwed you). The look of bewilderment on the "victim's" and your business partner's face when you do make it go is absolutely priceless.
Wanna bet he's now complaining about how I SCREWED HIM???
Treachery, old age and experience will always overcome youth and enthusiasm.
sdantonio, crude but here -> http://www.pminmo.com/phpBB/viewtopic.php?p=245#245
Phil, Still too many interests, too many projects, and not enough time!!!!!!!!
Vist my websites - http://pminmo.com & http://millpcbs.com
jeez. I was just joking man.
cheers :cheers:
thanks
Michael T.
"If you don't stand for something, chances are, you'll fall for anything!"
Sorry, miljnor, I was using your reply as a stepping stone/inspiration to address other prior posts that seemed to diminish the critical importance of proper grounding.
Did not mean or intend to dis you and appologize if I did.
BTW, there were some real heated "discussions" on this end during a very trying time when we went thru this grounding fiasco. The nerves are still a bit raw as you can probably see. Even now, there is a ramification that pops up and the original fiasco took place nearly 4 years ago.
Yes, if you do the wiring yourself and take the proper precautions (IE: star ground, powering from common current source, laptop, etc) you don't have to be so fussy.
But in an industrial environment (and there are folks here in that situation) you can't rely on "secret handshakes" or undocumented wiring crutches. You really need and SHOULD go the extra mile to protect your equipment and the safety of you and/or your help.
Why?
There is always the case where SOMEBODY will drag out an extension cord from across the shop to run a PC or meter, run a grounded plug instead of non grounded one, whatever. The person doesn't know about the "secret handshake" that you may have used with your grounding scenario to get it working (the situation we faced). Result can get expen$ive.
We were fortunate enough to have diagrams and schematics supplied with the equipment. Sadly, there were undocumented and disconnected grounds and other hacked up "fixes' on our interconnect diagrams. We ultimately tossed them due because they literally fell apart during the equipment move or "shouldn't be there" - little did we know what would happen when we repowered the equipment.
I do hope that this thread has been enlightening and, again, appologize for any tirade or dis'ing I may have done.
And yes
8-) ,
if you find cases where equipment has been miswired, are adept at solving electronic puzzles and can do the research needed to find obscure service manuals and/or service people to fix "obsolete" stuff inexpensively, such opportunities can provide you with very effective sources of inexpensive albeit slightly fried/toasted equipment....
Anyone have info to make a opto isolator print port protector (http://cgi.ebay.com/PRINT-PORT-PROTE...QQcmdZViewItem)
like these commercial model or similar? (I mean, schematics, copper, layout, etc.). Im building interface ant driver boards from info I get from Phils site (hi Phil) and I am THESE GUY IN THE WORLD WHO NEEDS OPTO ISOLATION because I dont know nothing about electronics and all errors are possible.
Thank you all very much for any information. Alberto.
I don't believe these are opto isolators. And suspect they are subject to the same ground fault condition that the vast majority of non opto isolators are, as well as there are faults that will still fail a parallel port.
Phil, Still too many interests, too many projects, and not enough time!!!!!!!!
Vist my websites - http://pminmo.com & http://millpcbs.com
Great Phil, thanks for your warning. Then: what can I do? Im scared hearing some users and encouraged hearing others. I thought those device was the ultimate solution. Please, recomend me some guidelines to connect safe or tell me where found info to build a working isolator device to connect between the PC port and the interface board.
One more time: thank you very much Phil.
Complete details can be found in following book:
"Real World Interfacing With Your PC" by James Barbarello
ISBN 0-7906-1145-7
From:
Prompt Pubilications
Howard W Sams Co.
2647 Waterford Parkway, E. Drive
Indianapolis, IN 46214-2041
1-800-428-7267
http://www.hwsams.com
Complete schematics and part numbers needed make it possible for even those with minimal electronics background to do proper isolation.
Also try "parallel port isolation" search on google.
If you can not understand the electronics needed to properly isolate the port, seek competent help. Some things simply can not be "taught" via message board description, ESPECIALLY if you do not have sufficient basic knowledge to build upon.
What speed of opto-coupler is neccessary for CNC work with EMC or the Mach 1, 2, or 3 controllers.
4N25 - 4n26 - 4N27 - 4N28
have Non-saturatied turn on time 2 microseconds
4N35 - 4N36 - 4N37 have turn on times of 2 microseconds with a max of 10 microseconds
Turn off times are the same with a 100 ohm load
HCPL-2300 chips give 25 nanosecond response
PS2501 series chips switch from 3 to 5 microseconds
MOC205 chips switch from 5 to 7.5 microseconds
CNY17 chips can turn off as slow as 34 microseconds
How fast of an optocoupler do you need for a good CNC isolation port?
Speed and electrical isolation are two different things.
The speed of the opto needed is a function of how fast you intend to send date to/from the port.
Electical isolation is how much voltage potential can be isolated between the input and output side of the chips.
A bit more spec sheet studying is in order.
Ok,
A little insite. Below is the front end of my driver designs. R1 serves two purposes, one it's part of a filter, two it provides 10,000 ohms of isolation to the output of the PC parallel port. In some electronic boards, there is no series resistance. If there is a failure on the cnc electonics side, the IC (IC1B in the graphic) can fail shorted and provide a significant current path back through the PC. (pin 2 on the graphic) Pop out comes smoke from the PC. For many years, that LPT port has been on the motherboard, thus you just fried your computer. That is one failure mode, and in my circuit will provide reasonable current limiting to protect the PC.
Another failure mode is via gnd wiring (blue on the graphic). Since many circuits are single ended, they rely on that 0v common to provide a reference current path which isnt' a problem, but sets up the potential for a problem via the AC utility wiring. That potential problem via the AC grounding path when the PC or machine side safety gnd isn't at the same 0V potential. That can happen because of high current loads on one of the AC circuits and faulty or absent safety gnds.
The ultimate solution for isolation is optical. There you have very high isolation characteristics, usually at voltage levels you shouldn't even get a fraction of. But there are limited Opto isolators that are fast enough to work with signal timing used by desktop cnc software. Two as example are the HCPL-2531 and HPCL-2631 (also versioned as 6n137 and 8) With optical isolation, when properly wired not to include a ground loop path, you are virtually foolproof for failure.
Potential exists for failure on the input pins of the LPT port also. But again, based on how the inputs are wired, that is much less of an issue to protect.
Another method to isolate is using differential receivers in an appropriate design configuration. Doesn't seem to be a popular method with the DIY'ers, not sure why.
Phil, Still too many interests, too many projects, and not enough time!!!!!!!!
Vist my websites - http://pminmo.com & http://millpcbs.com
I learned my lesson about optoisolation on my Fanuc's.
The lesson: Use OPTOISOLATION.
Can be a PITA to do as you have to be careful NOT to connnect grounds 'tween the PC and the device being connected to. It may also be difficult to get 5vdc "out" to power the opto on either side.
You can use diode array to pull power from the signals on LPT side and dump to a capacitor - plenty enough power to run the opto.
With a bit of creativity, you can also get power on "outside" via wall transformer to power the isolation circuit. Worst case, a 9V battery will last years.
I learned subsequently that it costs roughly $60 to have 3 a opto isolated boards etched and drilled from ExpressPCB. The design and layout software is FREE. Add another $5 for parts - figure $20 worst case.
In comparison, my Fanucs' have several THOUSAND in repairs due to NOT using proper isolation of LPT port and the Fanuc M/B's.
Don't even TRY to persuade me that you can't afford to do it right.
Having done it myself and learned in the process, if you fry a computer, the LPT port or an expensive CNC controller, you can find SYMPATHY in the dictionary somewhere between CRAP and VENERIAL DISEASE......
Is it possible to purchase a parrallel port opto-isolator circuit with a db25 connector on each end, a power supply and all the relevant opto electronic gubbins in the middle to make it all work in a simple plug and play turnkey solution ?
Dom
http://www.ukrobotics.com/projects
I just read through this entire post with a little fear and some hope. I am slowly putting together the hardware for my first CNC machine and have been trying to absorb and learn as much as I can. I think I will pick up the book that NC Cams recommended.
Right now all I have is a computer with controller software. I decided to buy the parallel interface board from CNC4PC. http://www.cnc4pc.com/Parallel_Port_Interface_Card.htm I just ordered it a few days ago, so I have not recieved it yet. I haven't decided what servo drives to use, that's why I've been reading in the Rutex forum. I am going to use MCG ID33003 brush motors.
Anyway, there has been alot of food for thought here.
Some of the stepper and servo controllers come with input signal opto-isolation (e.g. Gecko's). In that case you don't need to isolate again those pins on the port. Make sure you isolate all the inputs (limit and home switches) as well as the rest of the outputs from the parallel port (spindle on-off, cooling control, etc.)
For those of you who can manufacture your own printed circuit boards there are schematics on the Internet. Just make sure you keep separate PC port ground and pins from CNC equipment ground. Also you can provide +5volt from the USB port to supply the PC part of the interface, keep it separated physically on the board from +5 and/or +12 volt necessary to supply the CNC part of the interface.
Use high speed opto-isolators for the Step and Dir signals. Spindle on-off, coolant control, etc can use low speed opto-isolators , solid state or electro-mechanical relays.
Hugo.