Questions about wiring this up

[Jackal]

The  critical wiring has started.

Encoders, servos,limit switches.

1) Is it ok to run shielded wire from the Gecko drives to the servo motors for the power? If so, should the shielding be grounded at the Gecko end?




2) Will regular wiring be ok from the bridge rectifier  to the Geckos?  I have them twisted in pairs to maybe cut down on noise or (EMF?). They are close to each other.


 


3)When putting the  power wires and encoder wires in separate conduit. Is it ok to use metal flex conduit that isn't fluid sealed?


4) My encoders are  the AMT-102 type  from Digi-ikey.  http://search.digikey.com/scripts/DkSearch/dksus.dll?vendor=0&keywords=AMT102-V

What setting should I use for starting out? Here are the switch settings and different resolutions. http://products.cui.com/CUI_AMT102-V_Dip_Switch_Settings.pdf?fileID=3816

Everyone says to use differential encoders for less noise. This particular encoder doesn't have A, A-, B, B-, etc. But, this is the encoder that Mariss Fremanis  recommended for the Geckos.


5) On grounding the shielding for all of these shielded wires: They have  a pigtail soldered to the shielding at the end connected to the motherboard or breakout board. Should all of these be daisy chained together and then  grounded   to  the cabinet?

Sorry for all of the questions in one post. These are some things that i have been unsure about. After searching and reading  didn't clear this up, I asked here.

With these answers I'll fell more confident and a lot of other things will make more sense.

Thanks a lot,
JAckal

[Hood]

1) Is it ok to run shielded wire from the Gecko drives to the servo motors for the power? If so, should the shielding be grounded at the Gecko end?

Definitely something I would do and shielding would normally be connected at a single point in the cabinet where all shields connect to.

2) Will regular wiring be ok from the bridge rectifier  to the Geckos?  I have them twisted in pairs to maybe cut down on noise or (EMF?). They are close to each other.

Should be fine but try and keep as far away as you can from any low voltage signal wires.


 

3)When putting the  power wires and encoder wires in separate conduit. Is it ok to use metal flex conduit that isn't fluid sealed?

Again should be fine as long as any coolant that does get in doesnt find its way into anything electronic or  any wiring connections.

4) My encoders are  the AMT-102 type  from Digi-ikey.  http://search.digikey.com/scripts/DkSearch/dksus.dll?vendor=0&keywords=AMT102-V

What setting should I use for starting out? Here are the switch settings and different resolutions. http://products.cui.com/CUI_AMT102-V_Dip_Switch_Settings.pdf?fileID=3816

If using the parallel port then probably about 500 lines per rev (2000 as Mach sees it) would be ok. If using the smoothstepper or similar then probably 2000 lines per rev (8000 to Mach) would be fine

Everyone says to use differential encoders for less noise. This particular encoder doesn't have A, A-, B, B-, etc. But, this is the encoder that Mariss Fremanis  recommended for the Geckos.

Differential is the ideal as it helps guard against noise, you could get a line driver and a receiver so that the majority of the wiring is differential but you will probably be ok as long as you take care to shield properly.

5) On grounding the shielding for all of these shielded wires: They have  a pigtail soldered to the shielding at the end connected to the motherboard or breakout board. Should all of these be daisy chained together and then  grounded   to  the cabinet?

Daisychain is a bad idea, all shielding should be taken to a single point if possible (star grounding) although there are many opinions on shielding but star is definitely what most would recommend.

Hood

[NosmoKing]

There has been a change in the thinking in the past few years that shields should be connected both ends, but this should also  be accompanied with the equally important or more important something called  Equi-potential bonding to Earth ground.
This is an effort to ensure that all parts of the machine and relevant grounded control systems are all common to the same earth ground point.
This is especially important when mixing or combining different system such as exists in CNC control.

http://www.automation.siemens.com/doconweb/pdf/840C_1101_E/emv_r.pdf?p=1
http://te-dep-epc.web.cern.ch/te-dep-epc/Infos/Conferences/EMC%20Workshop/BaldorInstallationInstructions.pdf
I'm not sure why Gecko seemed to buck the industry standard of differential encoder for the sake of a few more inputs?
N.

[Jackal]

Thanks guys for the good info. I saved copies of the PDF files on wiring. That will be some very good reading.

According to the PMDX manual about not connecting different grounds & commons to the PC or B_O_B, it was confusing about where to ground the shielding.
I didn't want to create a ground loop or ruin my opto-isolation of the B_O_B.

More info is always welcome.


Thanks,
JAckal

[NosmoKing]

One of the reasons, or probably main reasons for advising the grounding of one end of shielding was to avoid ground loop, this would be where the earth ground of one piece of equipment was at a different potential to the grounded point of another, and current was liable to flow in the shield, now with the advent of equi-potential  bonding carried out correctly it would avoid any ground loops.
The whole concept is to make the machine one homogenous mass common to earth ground.
Your PC case should be a part of this due to the PC power supply common at ground.
N. 

[Tweakie.CNC]

Now to put the cat amongst the pigeons. 

If you bolt a wire, using a T nut, to one end of a cast iron mill table and then bolt a wire, using a T nut, to the other end then pass a current through the table you will be able to measure a potential difference at various points along the tables surface. Take a look at the size of the conductor in an ammeter shunt and see where the potential difference is measured.
Whilst it would be nice to have everything on a machine at the same potential the deviousness of science will just not allow it. If potential difference exists in an homogonous piece of metal then imagine the situation where various types of metal, plating finishes and fastener materials all having a different PD are combined into one machine.

Earth bonding is a requirement of law within industry and it makes sense that our home workshop meets the same safety standards as industry therefore earth bonding must be carried out in a professional way and there are many good documents suggesting its implementation.
However, we should not confuse earth bonding (which is essentially a safety issue) with signal shielding, they are two separate issues entirely.

For those who would like to re-invent the wheel, by all means connect the shield of your signal cables at both ends but you would be wise to follow the path where others have had success and only connect the shield at the controller end.

It was Eleanor Roosevelt who once said; “Learn from the mistakes of others. You can’t live long enough to make them all yourself”.

Tweakie.

[Jackal]

Guys,

 I appreciate  all the   views & opinions. By you sharing your opposing  situations & remedies, it teaches me to think, instead of you just solving my problems.

Thanks  for all the different aspects  & theories.

Keep the info coming.

JAckal

[NosmoKing]

Uni-Potential Bonding:
The only thing I would add is quite a few 'Noise' problems have been solved after this advice was given, problems such as limit switch 'tripping' when turning on spindle motor etc.
Also what has worked for me personally is grounding all system power supplies to the common earth ground star point, The PC and the AC Spindle motor is already connected in this manner by default.
BTW, as well as the Siemens paper on grounding, there is also one by Kollmorgen with the same advice.
N.

[Jackal]

Something else. I have a lot of this metal flex conduit.

Would it be  better to put the encoder, power, and limit switch wires  in separate conduit?

X= limit, power, + encoder (3) flexible conduits

Y= limit, power, + encoder (3) flexible conduits

Z= limit, power, + encoder (3) flexible conduits

The conduit is  3/8"  ( 10 mm) dia., so it wouldn't be a big messy hassle.

If it is a waste of effort, tell me that, too.

Thanks,
JAckal

[NosmoKing]

By Power, do you mean the power for the limits and/or encoder?
Define what the power is for.
N.