Problems with proximity of AC power & stepper wires?

[jhowelb]

What kinds of problems have surfaced from close proximity of AC power lead for spindle router and unshielded stepper motor wiring?

[martinw]

Dear jhowelb,

I really cannot imagine that there would ever be any problem at all in running unshielded cables for both your mains supply, and your steppers, even if both cables are "in intimate contact".

Low voltage and logic control wiring for PC, limit switches, DC supplies etc. are an entirely different matter, but I really do think that mains and stepper coil stuff can snuggle up nicely.

BTW, I have a strange suspicion that I am going to be "shot down in flames".....

Best wishes,

Martin

[jhowelb]

Aww, Martin ole buddy, they wouldn't REALLY flame ya would they? LOL!

I do know what you mean, seams these days that ANY statement on this board results is some difugalty!! Makes it difficult to learn anything cause nobody wants the flak!

I'm trying to unravel the reason for a G251 to bite the dust and a G203V to act nutty! It's got me barking at the moon and chasing ghosts.

BTW, good to hear from you again old friend!

jb

[123CNC]

If unshielded the mains and stepper motor leads are potential sources for EMI/RF noise. Especially when considering there overall length and paths. You may get lucky in not 'picking it up' on any of your control wiring, but if you have any inclination that it might be causing havoc, try to eliminate or add sources one at a time to 'try' and pinpoint the problem.

Chasing noise is a lot like catching ghosts, if your even sure they exist.

No flames, just noise.

[martinw]

Dear jhowelb,

Those G numbers sound like they may be from Gecko.

Just ask their technical support people, and forget your past scraps with other people on CNCZone. By all accounts, Marriss has about the finest outfit that there is in the field.

errgh...Peace , Man..

Best wishes,

Martin

[jhowelb]

Martin,
Yep, just shrug off the loony toons and glean what you can of value. Still it casts a pallor on the spirit enough to cause a good man such as yourself to give pause for fear of being flamed! Such a shame!

They are indeed Geckos and Mariss has already sent me a "stuff happens" replacement for the 251 but his notion that the problem was as result of shorted motor leads is off base. The little drive had been operating for weeks and decided to dump about an hour into an operation and thus I'm left to dare fate in the pursuit of the cause.

123CNC,
rf noise might cause jitters but how about burned mosfits? It seems that all the router tables pictured have the power lead running adjacent to the motor cables. If shielded, where would you ground it?

[123CNC]

Blowing mosfets, no. In the output of the g251?

Where to ground the shield? Many schools of thought. Common one, to ground single end at the signal source for signal wiring, this should be the same or single point as ground for the signal generator.

For a spindle motor leads, the shield or metallic conduit would be grounded at both ends.

For power, generally at the point of cabinet entry to the common ground point/bus. If running single leads, good idea to twist them together to help cancel themselves out.

For stepper motors, I'm not sure if you should treat it similar as the spindle or not, what I typically do is single ended ground in the cabinet where my amp/Gecko is mounted.

[martinw]

Martin,
Still it casts a pallor on the spirit enough to cause a good man such as yourself to give pause for fear of being flamed! Such a shame!

If shielded, where would you ground it?

Dear jhowelb,

Thank-you for the kind words, but actually the "shot down in flames" suggestion was not entirely serious. There are plenty of people who know a whole lot more than I do hereabouts on almost every forum that I clog up with my ramblings. That is just fine with me. (Probably a bit painful for them though... heh, heh.)

OK....I believe that the generally agreed principle is that any shielded cables should be grounded at one end only, and that end should go to a single ground point that is directly connected to the ground connection on the main incoming house/workshop supply. This is known as star grounding. Just one ground point.


Best wishes,

Martin

[jhowelb]

Thanks for your thoughts guys, guess I'll keep on blundering about and maybe I'll get lucky.

jb

[Mariss Freimanis]

jhowelb,

Funny things can happen between the drive outputs, the motor and back again.:-) Funny as in "I cannot believe there was a short-circuit in there someplace!" Once the impossible is eliminated all that's left is the improbable. Case in point:

Dan Mauch of Camtronics-CNC fame is the most technically competent and conscientious person I know. Why? Because I never had to deal with a packaged systems he turns out. He tests and burns-in every single unit before he ships. What he ships always works.

A few weeks ago he was testing a G540 using Mach3's 'Roadrunner' test program. If hardware passes the 'Roadrunner' test then it has to be good; it has the reputation of stressing hardware to the limit. Dan calls and says the G540 faults about 2/3 of the way through roadrunner every time. We go through everything; good supply, proper STP/DIR signals, functioning G540, new motors. Nothing seems wrong at all except the G540 bombs-out 2/3 of the way to done. During the consult a lot of things are checked, tested and swapped-out to zero-in on the cause. No joy; still bombs-out before finishing and I'm fresh out of suggestions and ideas.

All that's left are the motors and they are brand new. All that's left are the motors and since we had tried everything else, I said let's swap-out the motors. Dan did and problem went away. The problem was a brand-new, fresh out of the box motor had an internal intermittent short-circuit. It was narrowed down and the bad-boy motor was conclusively identified through further tests. Dan was very surprised, I was mildly so. The reason was once everything else was logically eliminated it left the improbable; it simply had to be the motor.

Why do I mention the G540? The G540 has short-circuit protection and it uses G250 drives. The G251s are identical to the G250 except for the connector. The G540 protection lets the drives live again and again if there is a short, outside and a single event will kill the drive.

Your drive had 2 blown MOSFETs and a blown 0.056 Ohm current-sense resistor. The damage was entirely consistent with a 75+ Amp current pulse through these parts that could only be caused by an external short-circuit. It made no difference where; in your wiring or inside the motor.

About Dan's motor, the analysis: The ill-fated motor was OK while it was cold. Dan ran the 'roadrunner' program. It stressed all motors and made them warm to hot. As the ill-fated motor got hot, it expanded as all hot things do. At some point (2/3 of the way through the program), it expanded enough to short a flaw in the motor winding insulation and caused a few turns of wire in the coil to short together. In turn the inductance plummeted to zero and it resulted in the current in the winding to head north of 75A in an instant. The G540 sensed it in 3uS and shut everything off and preserved the drives. Something like that happened with your drive without the "shut everything off" part.

If you are absolutely sure there was no accidental "Whoops" situation with your wiring then you should worry about a serial-killer motor on your hands.

About motors: Everyone wants cheap motors. You want to pay $0.99 for a motor and a penny more says you are being screwed. Consider what this philosophy does: It drives quality motors out from the marketplace and what fills the vacuum is cheap junk. It used to be unheard of (20 years ago) to have a motor with internal, intermittent shorts; it is becoming increasingly commonplace now. What is also becoming commonplace is crap motors that advertise terrific holding torque (>300 in-oz NEMA-23, >900 in-oz NEMA-34) but have absolutely miserable linearity. How does that matter? Awful low-speed resonance behavior is how. It's to the point no one realizes anymore a step motor should turn as smoothly as a good quality servomotor and not short-out when it gets warm.

If you are into 425 in-oz $29 NEMA-23 motors, keep doing what you are doing. Be prepared all are going to run far rougher at low speeds than you should ever put up with and be prepared every 10th motor or so will short-out and kill an unprotected drive. If you care about motors and you are willing to spend a few bucks more, get Japanese or US built motors. Vexta PK series, Sanyo Denki 103 series, MCG (if you can find any) or anything from Pacific Scientific comes to mind. Think of it as a fine stereo system, you buy Geckodrive for the best amplifiers, why attach $1.99 speakers to them? What do you think the whole mess will sound like? Good amps need good motors.

Sorry for the rant at the end. I didn't know it was there until I wrote it.:-)

Mariss

[jhowelb]

Thanks for weighing in, Mariss.

It is a 425 oz-in motor but priced at $49.00 and God only knows where it was produced. Probably in China like most of the things offered these days.

The same vendor provided the break out board over which there was much discussion about shipping a product other than that which was ordered and something as simple as documentation for the product provided. I shudder to think of the discussion over an intermittent "phantom" short in a motor.

The motor has been swapped out for a 23-305-DS8 for which there is precious little documentation available. It is stated to be 3 volt 3 amp and sold as a unipolar motor. I set it up as a bipolar parallel at 2.5 amp just for safety. (attached is a diagram I found by search engine query)

I was ready to toss the C-10 breakout board as well when I started this thread. Mean while I discovered that this time the problem was with a set screw on a screw shaft that had worked loose producing the same symptom.

My response to all this was ultimately to order more 203V's at the "appreciation" price. Quality maters.

I don't know where to purchase Japanese motors.

Thanks,
jb

[123CNC]

Try here, look at the PK- range for the 251s.

http://www.orientalmotor.com/product...c-step-motors/