506,373 active members
2,692 visitors online
Register for free
Login
Page 1 of 2 12
Results 1 to 12 of 13
  1. #1
    Registered
    Join Date
    Dec 2004
    Posts
    132

    Apparent noise issue

    My machine has been running for some time now happily without any electrical noise issues. Everything is grounded to a common star ground point in my electrical cabinet, and this includes the shielded cable running from my stepper motors to the enclosure. From the connectors on the enclosure to the stepper drives, the same shielded cable is used, and the shielding is grounded to the big heatsink that the stepper drives are mounted to. My Kflop is grounded via the USB cable (I am assuming).

    A few weeks ago I installed a Kanalog so that I could hook up my proximity sensors and therefore have a reliable means of zeroing and squaring up my machine. Because of this switch, I had to change how I connected to my Kflop, using JP5 instead of JP7. To make this connection, I cut a short length of I bought a RJ45-->Serial adaptor (similar to this: RJ45 to DB9 Female Modular Adapter - Black) , and then stuck the pins into one end of a Pheonix Contact that I had laying aorund.. similar to this: (1803675 Phoenix Contact | Mouser).

    Everything works as I expect it to, however I now have a consistent issue with my motors clicking slightly while the machine is at rest. Sometimes when I'm running a file I can hear a clunk, which to sounds to me as if I have just lost steps (this is what my old machine used to sound like when it lost steps). Sure enough, when I go to check my 0,0 point that I have marked for reference on the table, it is slightly off. (the last little file I ran I ended up being .011 off in x and .02 off in Y)

    Messing around with my stepper motor cabling, I have been able to induce noise by shaking them around, and so I have paid a lot of attention to ensuring that I have a good metal to metal connection between the shielded microphone connectors that I am using to connect the cable to the cabinet with. This morning I ground away all of the paint surrounding the connectors to be sure I was getting decent contact. I tested each connector with a continuity tester, and I have continuity between the all the the connectors and the box, and thus the main ground. This seems to make the clicking go away for the most part, but it still occurs after a while. I'm not sure where to go from here as far as dealing with this, and I'm wondering if maybe there is something I can do on the Kflop end to mitigate this?

    Any ideas are greatly appreciated.

    -Spank

  2. #2
    Junior Member
    Join Date
    May 2006
    Posts
    3157
    Hi Spank,

    Here are some thoughts:

    If the steppers are clicking consistently while at rest I would focus on that. Does the position drift even while sitting at rest?

    What kind of drives do you have and how are you driving them? TTL mode?

    The KFLOP JP5 doesn't have an GND for the Step/Dir signals like JP7 does so you need to pick GND up from somewhere. How and where are you doing that from?

    Microphone connectors seem strange for Motors. Microphones are very low current devices.

    Shaking cables causing a steps would normally indicate a bad connection or cable somewhere.

    Not sure where your shields are connected. But Motor shields collect noise from the motor signals. Step/Dir cable shields are to protect the signals. Connecting the two together can feed motor noise into the control signals. Eventually the two will usually always ultimately be connected but how they are connected might make a difference. You might connect the motor shield to some far away earth ground to take the noise away. But connect the signal shields to some hopefully quiet close DC GND.

    HTH
    Regards
    TK
    TK
    http://dynomotion.com

  3. #3
    Registered
    Join Date
    Dec 2004
    Posts
    132
    Sorry for taking a while to respond here. I have been away from my machine since I posted. To answer you questions, JP5 has not been grounded... how would you suggest doing this. My guess is that this might cure my problems, as I didn't have this issue at all when I was connected to JP7. I say the connections are microphone connectors, because that's what they look like, however I did purchase them through PMDX (PMDX.COM - Products for CNC and motion control applications), and they are sold specifically for this purpose on his site. I've had good luck with other PMDX offerings, so I went with these as they seemed cost effective and suitable for my needs.

    I've lost you where you talk about motor shields vs cable shields. The only shields that I am aware of are the shields that I have on the cables running from my stepper drives (Centent CNO165 Centent CN0165) to the motors. They are connected to ground at the control box by way of the large heat sink that the drives are mounted on. The Motors are grounded to the same star ground, as is the rest of the machine. Kflop is grounded through the 5v DC power supply.

    I guess all of this brings me back to your original point that JP5 does not have a ground. Do I just run a ground wire from one of the pins to the 5v ground?

  4. #4
    Junior Member
    Join Date
    May 2006
    Posts
    3157
    Hi Spank,

    It depends on your drives whether you need to have a common GND connection or something else like +5V connection. I would make the same connections you had before when it worked. If you had a GND connection before connect one close to KFLOP like to the KFLOP GND lug hole marked GND.

    Regards
    TK
    http://dynomotion.com

  5. #5
    Registered
    Join Date
    Dec 2004
    Posts
    132
    "I would make the same connections you had before when it worked. If you had a GND connection before connect one close to KFLOP like to the KFLOP GND lug hole marked GND."

    This is where you've lost me. The only thing different is that I'm using JP5. JP7 has a ground connection... so are you saying that I just choose an empty pin on JP5 and then connect it to a 5V ground source? My drives are all grounded through their DC power supply (76V), and they are all obviously sharing the same ground through the supply. The ground lug on Kflop is confusing to me, because it doesn't seem to be connected to anything? If I connect a pin on JP5 to the Kflop ground lug, will I then have a 5V ground connection (Kflop is powered through it's own 5v power supply)?
    Thanks Spank

  6. #6
    Registered
    Join Date
    Dec 2004
    Posts
    132
    I already answered my own question with a continuity tester. I'm now faced with a new issue though: how can I ground JP5 when I am already using all of the available pins for step and dir signals?

  7. #7
    Junior Member
    Join Date
    May 2006
    Posts
    3157
    Hi Spank,

    Do not ground any of the JP5 output signals. That would cause a short circuit and cause possible damage.

    I don't see that you described or provided any information on what type of drives you have or how their inputs should be driven or how you had it wired originally. This is basic electronics. Current flows in a circuit loop. Imagine you have a battery, a switch, and a light bulb connected in a loop. The light can be controlled by putting the switch between the + of the battery and the bulb, or placed between the - of the battery and the bulb. The terminology used to describe the methods is that in the first case the switch is "sourcing" current to the bulb, and the second it is "sinking" return current from the bulb. Most drive's inputs need to be driven like a light bulb (LED). Often in order to save pins/terminals several LEDs have one of their leads all tied together in common. Whether the + or -'s are tied together determines how they must be driven.

    Regards
    TK
    http://dynomotion.com

  8. #8
    Registered
    Join Date
    Dec 2004
    Posts
    132
    Tom,
    Thanks for your response. I provided information regarding my drives back in post 3 of this thread. Again, they are Centent CNO 165 drives. Here is the user manual for the drive: http://centent.com/Manuals/cn0165.pdf. These are very similar to Gecko drives. I understand basic DC electronics, and I understand what a circuit loop is. All four of the drives are supplied with 76 VDC from the + side my power supply, and the negative of all drives returns back to the (-) side of the power supply. There is an additional an additional ground terminal on the drives that is used solely for a current set resistor.

    Now in your initial response to my question is this thread you said:

    "The KFLOP JP5 doesn't have an GND for the Step/Dir signals like JP7 does so you need to pick GND up from somewhere. How and where are you doing that from?"

    Once again, the ONLY difference between my setup before installing the Kanalog and now is that I am sending step/dir signals via JP5, which, as you state doesn't have a GND and yet my signals still need a GND from somewhere.

    You then say:
    Do not ground any of the JP5 output signals. That would cause a short circuit and cause possible damage..... I understand that, and I wasn't about to do that. What I am still asking and confused about is where and how do I ground these SIGNALS that were grounded before when then were run to my breakout board via JP7? If I must run the signals via JP5, and they also must in turn be grounded... how is this done? Should I run my signals from somewhere else?

  9. #9
    Junior Member
    Join Date
    May 2006
    Posts
    3157
    Hi Spank,

    Sorry I wasn't aware of your background and I hadn't noticed the link.

    It seems the Opto LEDs in your drives are common anode (have the + connected together). They do not have or need a GND connection regarding the Step/Dir input. If you have a connection between the High Voltage Supply Ground and KFLOP DC GND then the opto isolation is defeated.

    I find the specification on page 41 confusing. I think when they say 1.8V min that they mean that the step or direction pin must be at least 1.8V lower than the +5V input pin

    From page 41
    Step Pulse Input
    Logic '1' voltage 1.8V min 5.0V max
    Logic '0' current 3mA min 20 mA max


    You didn't describe how you had things wired before using JP7 or if you configured KFLOP to use LVTTL or open collector output mode. I assume you had a connection from +5V on JP7 to the drive's pin labeled +5V DC? Or maybe you were wiring directly from some +5V supply?

    You would of course need some positive supply connected to the drive's pin labeled +5V DC. I'm thinking it would be better to connect +3.3V to the drive's pin labeled +5V DC. The problem with using +5V is that when KFLOP's step or direction output turns off the output is still diode clamped to the 3.3V power supply. This means the pin will be at about 3.85V. The Drives Opto LED will still see 5V-3.85V = 1.25V which is dangerously close to the 1.8V min required to turn on. Small noise spikes on the +5V supply might cause the LED to turn on. LVTTL mode would be worse (5V-3.3V = 1.7V). Furthermore the spec has no information on the max voltage or current to be guaranteed to turn off. Just because 1.8V is guaranteed to turn on the LED doesn't mean that 1.6V might not also turn it on.

    Using 3.3V instead of 5V should work well because when KFLOP sinks low to below 0.5V the Drive should see (3.3V-0.5V = 2.8V) which is well above the 1.8V minimum. When KFLOP goes high (either in LVTTL mode or open collector) the pin should go very close to 3.3V so the drive should see (3.3V-3.3V = 0v) and definitely turn off.

    So in summary: Try connecting 3.3V from KFLOP JP2, JP4, JP6 or Kanalog JP8 to your Drives +5V DC pin.

    HTH
    Regards
    TK
    http://dynomotion.com

  10. #10
    Registered
    Join Date
    Dec 2004
    Posts
    132
    Well, I think my board might be fried. I tried what you suggested above, and initially I basically had 3.3v from Kanalog on the same lug as my 5v supply (while also being connected to the drives) mistakenly (me being sloppy). The machine worked like this briefly, but then didn't. Once saw my mistake, I disconnected the 3.3v from Kanalog and reran the 5v as I had it before to the drives and 5v to Kflop. Everything actually worked really well with no noise. So I started cutting a job, and at some point in the middle, Kflop disconnected. I looked into my enclosure, and there was no power going to Kflop, so I went back and tested my connections and I wasn't getting 5v from my supply. I fiddled a bit more, and then lo and behold I was getting 5v at my terminal strip that supplies my 5vdc to the components in my enclosure that need it. So I turned connected everything back up again and repowered, and Kflop was working, and the machine was smooth without any noise issues. I restarted the program that had stopped and was able to finish the program. I moved the gantry out of the way so I could reposition the work piece to cut the other side, and then Kflop disconnected again... again, no power. So I figure at this point that my mistake with supplying 3.3v in addition to 5v might have messed up the supply somehow, so I take that off and swap it out with another 5v supply that I had lying around. Once that's all connected, I power back up, and POP... something shorts out. I figure it's the power supply, and indeed it's not putting out any power, so I connect yet another 5v supply that I have. This one is putting out a clean 5v that I test prior to getting it connected. When I repower, nothing happens with Kflop. So I test the 5v strip, and it says that the supply is only outputting 3v... which is strange, but I take that for what it is, and figure that Kflop isn't going to run if it doesn't get 5v. I then run out to the store and pick up another 5v supply, and get it all connected, and nothing happens. This time I am getting 5v at the molex connector going to Kflop... so it doesn't look good. I put the jumper back on J3, and tried running with just USB... nothing. This is happening at a really bad time for me, as I'm trying to fulfill orders that I have for Christmas. I'm pretty sure it's fried...thoughts?

    -Spank

  11. #11
    Junior Member
    Join Date
    May 2006
    Posts
    3157
    Hi Spank,

    That doesn't sound good. If you send us the board we can test/diagnose it. If something simple is wrong we may be able to repair it. But applying the wrong power to either 3.3V or 5V pins is likely to cause major damage. PM me if you want an RMA number for us to test it. If you disconnect everything and only apply +5V and the LEDs do not flash then the board has been damaged.

    Regards
    TK
    http://dynomotion.com

  12. #12
    Registered
    Join Date
    Dec 2004
    Posts
    132
    Thanks Tom,
    I'll PM you. I'm not sure if I sent the wrong voltage to the board as I was supplying the board 5v and then mistakenly connected 3.3v from kanalog back to the same terminal block that had the 5V supply. At any rate, KFlop still worked after that, however I am thinking that when the second power supply shorted out, that it might have caused a voltage spike that damaged the board.
    Spank

Page 1 of 2 12

Similar Threads

  1. Need help with electrical noise issue
    By b42000 in forum DIY CNC Router Table Machines
    Replies: 10
    Last Post: 07-18-2013, 02:47 AM
  2. Noise reduction issue!
    By Barefootboy4 in forum Mach Software (ArtSoft software)
    Replies: 0
    Last Post: 09-15-2011, 12:40 PM
  3. Need help please, crazy noise issue.....
    By pete from TN in forum Benchtop Machines
    Replies: 9
    Last Post: 04-13-2010, 11:17 PM
  4. Motor Noise Issue
    By dsnellen in forum General CNC Machine Related Electronics
    Replies: 4
    Last Post: 01-26-2010, 03:36 PM
  5. Apparent issue with SJ200 & Leeson 3ph 2HP motor
    By haysys in forum Phase Converters
    Replies: 31
    Last Post: 01-14-2008, 01:04 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •