Parallel port Isolation

[rkonnen]

I talked to Tom about parallel port isolation, he has said that it does not seem to matter. He did comment that he has heard that the Campbell board could give some issues. If others are using breakout boards or some type of parallel port isolation, if you could post your results. This would help myself and others for just the knowledge of what is being used.
Thanks,
Richard Konnen

[pminmo]

With properly laid out system and wiring, you don't have to have opto isolation. The key is the negative/neutral/motor/driver common and current paths. Also what you plug your pc 120V into vs your drive electronics 120VAC and motor power supply 120VAC plugs into. I would venture to bet there are a number of setups that do have opto isolation, but have defeated it in their wiring via negative/common wiring.

[Mcgyver]

Phil, can you elaborate a bit or suggest a solution or where people go wrong with it?

I'm not sure I reading this right - are you saying that spikes etc (or whatever you're trying to isolate from) could reach the pc through the neutral line of the house/shop wiring? wouldn't any electrical motor with its collapsing field potentially cause the same spike and isn't it the pc's power supply's job to deal with these?

[rutexus]

Gentlemen,

The opto isolated mother board is primarily targeted towards PC's that have less than a 5 volt output on the parallel port.

Every user should use a transformer to get the VM for the motor power. As long as the user does not connect the secondary of this transformer to anything there should be no issues of spikes or anything else that would create a need for isolation.

If a user were to rectify power straight from an outlet for their VM there could be ground loops that could damage equipment. It should NEVER be done this way.

Tom Eldredge
Rutex LLC

[pminmo]

One of the issues with non isolation is where is 0 volts. Sometimes I like to use extremes to illustrate a point. Fundamentally, wire has resistance and ohms law applies.
DON"T DO WHAT I DEPICT HERE STRICTLY HYPOTHETICAL TO EXPLAIN
Consider you have one wire from the motor power supply negative to drive X, from drive X to drive Y, from drive Y to drive Z then to the PC. Each drive handle 6A worth of motor. Each drive wire segment is .2 ohms. Ohms law is E=I*R so what you have in drive currents is 6A from the Z segment to the Y segment, 12A in the Y segment to X segment and 18A in the X segment to the Power supply. That means that the 0V at the power supply negative terminal, is actually 3.6V at the first drives negative terminal, 6V at the second drives negative terminal, and 7.2V at the Z drives negative terminal, which is wired to the PC PPort common. In all probablility the PC common is tied to safety gnd, and in good practive the power supply negative terminal is wired to safety gnd. So now now have a loop via you ac wiring safety gnd. The wire segment from the Z drive to the pc is probably going to pull a hefty current as there maybe parallel wiring in the cable and connections to/from the db25. You think this is just signal wiring, NBD and the next thing you know you power up, smell smoke, see you parallel cable wire smoking or fused open cause it was too small to handle the current.
If the drives are optically isolated and treated properly that last wire segment (Dark green)doesn't exist.

Tom I respectfully disagree. Optoisolation has nothing to do with the logic level output of the pc. 2nd leaving a transformer dc output not referenced to a saftey ground requires special physical isolation and insulation handling if it is to meet NEC an UL code requirements. Anytime you have a insulation fault condition that can expose 50V (I think that is the current threshold) to a conductive enclosure it has to fault.

[miljnor]

Is this is why they do star pattern wiring in grounds?????

[Mcgyver]

so its D-E that shouldn't be there, and why would it if its opto isolated, correct?

opto isolator boards (yours being the one I know the best) have the port's negative pins only connected (through the opto isolators) with other port pins. this is being the case, and so long as someone doesn't connect the parallel port ground to the ground in any way connected the with controllers/motor/etc power supply ground, there should not be an issue.

I want to make sure I understand exactly what the pitfall is you were referring to in the first post, else I am a strong candidate for falling in it thanks for the help

[pminmo]

Miljnor,
Correct, star grounding reduces the opportunity for ground current influence. Everything comes from the center of the star 0V.

Mcgyver,
Correct, how you defeat the isolation is having that segment in via a path from one of the db25 ground pins to your electronics/motor/powersupply 0V. With an opto isolator the led source of the isolators current path/loop if from an isolated DC source through the led, through the parallel port output. The isolated dc source is referenced to the PC pport ground.

[rutexus]

Thank you Miljnor for clarifying this for our users. Yes, this is why the Star powering is so important. I had to learn the hard way years ago. I just could not believe it made that much difference in a foot of nice heavy copper wire, but it can.

Tom Eldredge
Rutex LLC

[rutexus]

Gentlemen,

I now have the wiring diagram for the R2030 drive/Sanyo D. drive connections from Vladimir, which should save time for anyone hooking up any of the Sanyo D. drives. I don't know how to get it into this forum, so if you want a copy of the pdf just email me at tom@rutex.com and I'll attach it with a return email.

Tom Eldredge
Rutex LLC

[rutexus]

Gentlemen,

By the way, I have been shipping the R2030 drives and hope to have some feedback from the first users soon.

Tom Eldredge
Rutex LLC

[NC Cams]

DO NOT BELIEVE THAT OPTO ISOLATION IS NOT CRICICAL!!!!!

I have a Fanuc 5 controller on my nc lathe (3 actually). The idiot who I bought it from crafted up a BTR between it and the PC. Ground was ground as far as he was concerned so he tied the ground pins of the parallel to the ground interface for the BTR connect of the Fanuc. Turns out that he did the same for all 3.

The transformer the NC was on was at a different ground potential than the "ground" the PC was on for god knows what reason -(old building, kluged up wiring, whatever).

Result: the PC fried a number of the horribly obsolete TTL interface chips on the Fanuc due to a ground current issue. Let the smoke out of a number of chips REAL quick when that happens...

Trying to get an obsolete 1979 motherboard serviced in 2005 ain't easy. Fanuc wanted something like $4500 for the board (yeah sure - paid that much for the whole lathe). Found a guy in Texas who services them at a very reasonable price (considering) but it still cost several thou to fix them.

Yes, you can run non isolated interface 'tween the PC and whatever. I learned my lesson and WON'T EVER DO IT AGAIN. You are free to ignore this suggestion but you'd probably be less than intelligent/well informed to ignore it.