584,798 active members*
4,466 visitors online*
Register for free
Login
Page 12 of 13 210111213
Results 221 to 240 of 254
  1. #221
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    Retested spindle input wire to Kanalog ground through meter and 120ohm resistor (closest I have to 100ohm). Checked 29mV static, then 14mV when jogged. Same results regardless of jog direction. Spindle at rest, then rotated CW at 4.5 RPM, regardless of jog direction.

    Please reread my post about the 70 mV change. Maybe I was not clear enough. The 70mV change was with DAC=100, then jogging.
    Thanks

  2. #222
    Join Date
    May 2006
    Posts
    4043

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Charles,

    So from Post 214:

    Connected voltmeter to spindle drive high and low inputs. Commanded DAC4=100. Voltage at drive input checked -454mV, same as at Kanalog.

    Jogged Z- from KMotionCNC. Drive input voltage went from -454mV to -528mV, then back to -454mV when jogging stopped.

    Jogged Z+, voltage -531mV.
    Similar results from X.
    Spindle sped up slightly, no jerking.
    This is where Jogging X Z causes Spindle Command Voltage to shift by 77mV.

    Later I understood the voltage at the Kanalog end (DAC4 relative to Kanalog GND) also shifts by the same 77mv when Jogging.

    I thought we performed a similar test with DAC4 commanded to zero and saw a similar shift magnitude when Jogging.

    So the question becomes why is the DAC4's voltage shifting?

    The test from your previous post shows no significant current being introduced. 29mV/120Ohms=0.24mA. Did the DAC Voltage also shift 70mV?

    So maybe repeat the test with the DAC commanded to 0 and then 100 while verifying Jogging does cause a 70mV voltage shift and there is no significant current being injected.
    TK
    http://dynomotion.com

  3. #223
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    Test 1. With spindle command wire connected to Dac4, voltage reading from Dac4 to Kanalog ground, and Dac4=0, I get 23mV, CCW rotation at 10 RPM.

    While jogging Z- or +, I get -39.6mV , CW rotation at 30 RPM.

    Test 2. Dac4=100, -456mV, CW, 225 RPM.

    Jogging Z, -516mV, CW, 260 RPM.

    Changed meter setup to check current with meter and 120 ohm resistor in series with command wire.

    Test 3. Dac4=0, -37.9mV

    Jogging Z, 24.5mV

    Test 4. Dac4=100, 430.8mV

    Jogging X, 503.2mV

    Thanks

  4. #224
    Join Date
    May 2006
    Posts
    4043

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Charles,

    Based on Test 4 it looks like you are measuring the Dac voltage relative to Kanalog GND rather than across the resistor. Is that correct?

    I would put in the series resistor and leave it in for both tests.

    Set DAC4=0 measure voltage, current, and RPM before Jogging X and while Jogging X
    Set DAC4=100 measure voltage, current, and RPM before Jogging X and while Jogging X
    TK
    http://dynomotion.com

  5. #225
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    You are correct. I was measuring to ground.

    Reran tests as you suggested.

    120 ohm resistor between Dac4 and spindle drive command wire. For voltage tests I measured from Kanalog ground to Kanalog side of resistor as well as from drive side of resistor. For current tests I measured across the resistor. All values in mV except RPM and direction of rotation.

    Dac4=0, no jog, voltage Kan side 23.4, drive side 23.0, current -.2, 8 RPM CCW

    Dac4=0, jog X, volt Kan side 14.7, drive side 6.7, current .1, 12 RPM CW

    Dac4=100, no jog, volt Kan side -455.7, drive side -452.9, current 2.5, 218 RPM CW

    Dac4=100, jog X, volt Kan side -467.9, drive side -466.2, current 2.7, 240 RPM CW

    Since X numbers look very different from what I have seen in Z, I retested the same way in Z axis.

    Dac4=0, no jog, volt Kan side 20.8, drive side 19.5, current -.1, 3 RPM CCW

    Dac4=0, jog Z, volt Kan side -38.3, drive side -37.6, current .1, 30 RPM CW

    Dac4=100, no jog, volt Kan side 457.4, drive side 453.0, current 2.7, 227 RPM CW

    Dac4=100, jog Z, volt Kan side 513.6, drive side 509.7, current 2.9, 260 RPM CW

    Just for fun I checked Dac4 to Kanalog ground voltage with the drive command wire disconnected.

    Dac4=0, no jog, 22.1
    Jog Z, -35.1

    Dac4=100, no jog, 458.1
    Jog Z, -512.3

    Hope that is readable.

    Thanks

  6. #226
    Join Date
    May 2006
    Posts
    4043

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Charles,

    Thanks for the clear test data. I think this shows there is no significant DC current flowing through the command wire.

    I'm glad you did the extra test as I thought earlier we established earlier the DAC did not shift unless the command wire was connected. That was the idea behind checking the current through the command wire. I'm sorry if I lead us on a wild goose chase,

    So I believe we have established that the Kanalog DAC outputs all shift by ~60mV whenever X or Z Jog regardless of whether they are connected or open or what Voltage they are commanded to.

    If everything is connected and powered but the motor being Jogged is disconnected so there are no large motor currents then we do not see the shift. Correct?

    If so then I think that would point to a grounding issue. Although there is still no excuse for Kanalog's DAC output to shift relative to it's own GND. You are placing the Meter lead right on the Kanalog GND at a Kanalog Screw Terminal correct? Kanalog has a complete copper GND Plane layer. So I would expect it would require a huge current to be passing through Kanalog GNDs to cause a 60mv shift.

    60mV is not much and may not really cause a problem but it should not happen and I normally like to figure out things that should not be before moving forward. The plots back in post 201 show the DAC needed to change by the equivalent of ~60mV (10 DAC counts) to hold position when Jogging. And the Position needed to change by ~100 counts (P gain 0.11) to achieve this offset. Is this the "Jerk" that you often refer to?

    I suppose its time to try a different Kanalog.
    TK
    http://dynomotion.com

  7. #227
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    "If everything is connected and powered but the motor being Jogged is disconnected so there are no large motor currents then we do not see the shift. Correct?"

    I just retested. With all Dac's connected and ground wires connected, spindle fuses out, meter leads on Dac4 and Kanalog ground, no jogging, Dac4 checks approx. 20 mV. While jogging Z-, -35.6,. Z+, -64.9,. X-, 5.1,. X+, 9.0.

    X and Z check very different

    I have a 4" wire connected to the Kanalog ground screw terminal, to which I clamp the meter common.

    Jerk is what I refer to as any quick, maybe violent movement.

    I sent an email to support@dynomotion requesting a replacement Kanalog per your suggestion. Hopefully we can soon move on to the next adventure! Thanks Tom

  8. #228
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    Sorry to say, but the new Kanalog did not help. Spindle still responds to X and Z jogging. I have spent all day going over grounding, possible interference, and anything I could think of that may be causing the spindle problem.

    I have 4 pages of notes that I don't want to bore you with. A lot of it was rechecking. I checked every wire on the machine ground block. I think it's called a 'star' configuration. All wires were disconnected and checked to the ground block at .3 to .8 ohms except the two going to the original control transformer and the one for the new Antek transformer, which were open.

    I have condensed most of my notes to the following:

    Jogging with the spindle command wire disconnected at Kanalog affects the spindle.

    Jogging with X and Z command wires disconnected at Kanalog does not affect the spindle.

    Jogging with the spindle encoder disconnected at the motor affects the spindle.

    Jogging with power disconnected from the axis power supply does not affect the spindle.

    Jogging with a 1.5V battery connected to Kanalog ground and spindle command wire, disconnected from Dac4, does not affect the spindle.

    Thanks

  9. #229
    Join Date
    May 2006
    Posts
    4043

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Charles,

    Sorry to say, but the new Kanalog did not help. Spindle still responds to X and Z jogging.
    I didn't really think it was the Kanalog but thought it would be worth a test. I suppose both Kanalogs could have a problem but very unlikely.


    Jogging with the spindle command wire disconnected at Kanalog affects the spindle.
    That would indicate some interference in the wiring/grounding/Drives/Supplies. You might try that test again with the command wire connected to Kanalog GND. And/or with the command wire disconnected at the Spindle Drive end.


    Jogging with X and Z command wires disconnected at Kanalog does not affect the spindle.
    In this case without any motors driving there would be any high currents so no effect. This would indicate some interference in the wiring/grounding/Drives/Supplies.


    Jogging with the spindle encoder disconnected at the motor affects the spindle.
    That's good to know as it shows the interference is not in the feedback or servo tuning.


    Jogging with power disconnected from the axis power supply does not affect the spindle.
    Again points to high currents flowing somewhere.


    Jogging with a 1.5V battery connected to Kanalog ground and spindle command wire, disconnected from Dac4, does not affect the spindle.
    This is inconsistent with the above indicating a problem with Dac4. But maybe you just don't notice the interference as it is small compared to the 1.5V.


    Do you have access to an oscilloscope? There could be higher amplitude AC signals on commands and GNDs that we don't see with a voltmeter.
    TK
    http://dynomotion.com

  10. #230
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    Thanks for sticking with me on this.

    More tests:

    Connected spindle drive command wire to Kanalog ground (both wires). Spindle rotates at about 3 RPM. Jogging X or Z causes it to go to about 8 RPM.

    Disconnecting the command wire at the drive makes the spindle rotate about 11 RPM. (Seems to creep up). Jogging causes it to increase to about 15 RPM, then drop back when jogging stops.

    Disconnecting the ground wire from the drive makes the spindle jerk or shutter. I think this may be because the counts are building up due to feedback, then trying to catch up.

    I have rerouted and isolated the spindle tach cable, the spindle encoder cable, the spindle command cable.

    I have checked the same cables with an ohm meter. No cross over, continuity is good.
    Added a ground wire to the X axis motor.

    When the spindle command wires are both connected to Kanalog, jogging affects the spindle. When both are not connected, jogging does not affect the spindle. When either one is connected, jogging affects the spindle.

    I connected a 0-9VDC variable battery box to the spindle command cable.

    0 Volts, 0.4 RPM, jogging does not affect the spindle.
    50.2 mV, 24.4 RPM, no effect
    102.0 mV, 49.1 RPM, no effect
    204.1 mV, 98.1 RPM, no effect
    400.3 mV, 192.0 RPM, no effect
    1.003 V, 480.9 RPM, no effect
    2.012 V, 964.9 RPM, no effect
    4.009 V, 1926.1 RPM, no effect

    Spindle ran smooth, quiet, steady within .3 RPM.

    I know some of this suggests grounding, maybe too much grounding causing loops? Feedback and control cables are shield grounded on one end only.

    There does seem to be something going on between Kanalog and the spindle drive, but it appears to be on the Kanalog end.

    The spindle runs perfectly on the battery box, unaffected by axis jogging.

    Can you think of anything that I could have wrong in my Setups or C code? I will post copies of what I have.

    Is it possible that the KFLOP is bad or damaged?

    Thanks for your patience.

  11. #231
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    My Init file

  12. #232
    Join Date
    May 2006
    Posts
    4043

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Charles,

    Connected spindle drive command wire to Kanalog ground (both wires). Spindle rotates at about 3 RPM. Jogging X or Z causes it to go to about 8 RPM.
    This shows there is definitely some interaction albeit small. Looks like the Spindle drive does about 500RPM / Volt. So 5 RPM change would be ~ 10mV.

    Disconnecting the command wire at the drive makes the spindle rotate about 11 RPM. (Seems to creep up). Jogging causes it to increase to about 15 RPM, then drop back when jogging stops.

    Disconnecting the ground wire from the drive makes the spindle jerk or shutter. I think this may be because the counts are building up due to feedback, then trying to catch up.
    Floating inputs and/or without a GND reference are likely to be erratic so this is probably normal.

    I connected a 0-9VDC variable battery box to the spindle command cable.
    what is a battery box? Where was it connected. Was the Spindle command reference GND connected to Kanalog GND? If not is there interaction with the battery box if it is?

    I don't see any issue with the Init C Program. You might want to Flash New Version to make sure nothing else is running or configured unusually. Although I think I recall you already did this?

    Is it possible that the KFLOP is bad or damaged?
    I suppose anything is possible. But its hard for me to imagine how KFLOP damaged would result in this interaction.


    I don't really have a feel for these "effects" you are observing. I wish we had some way of quantifying them. I wouldn't expect these tens of millivolt changes to cause a violent reaction. The plot showed very small changes in output and encoder movement. I suppose we could code a loop to pulse a DAC a certain amount at a certain frequency to see what it take to get something similar to what you are observing.


    If I had to bet I would think there is major ground noise. An oscilloscope would be very helpful.


    A USB Isolator to allow Kanalog GND to be isolated from Earth GND might be a good test. Or we could write a C Program to Jog XZ back and forth so we could then temporarily unplug the USB to eliminate any earth GNDs to see if the interaction goes away.


    Thanks for your patience
    TK
    http://dynomotion.com

  13. #233
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    The battery box is homemade device that a friend used at his former job. It is simply a box containing a 9 volt battery, a variable resistor, a power switch, a polarity reversing switch, and test lead terminals. See picture. Works great for feeding input to drives for testing or just moving axis' or spindles at various speeds and directions.

    Reran same "battery box" test with ground wire connected from box neg terminal to Kanalog ground and without.

    Bottom line, grounded causes increase in RPM, and jogging while grounded causes more increase in RPM. Speed varies up to 3 RPM while grounded, .3 variation while not grounded.

    Voltage. RPM grnded. Jog grnded. Not grnded
    0. 2. 6 .4
    50.2 mV. 26. 37. 24
    101.8 mV. 53. 64. 48
    203.4 mV. 103. 113. 96
    404.8 mV. 201. 213. 192
    1.01 V. 490. 499. 480
    2.01 V. 960. 973. 958
    4.02 V. 1919. 1928. 1917

    Jogging without grounding to Kanalog caused no effect on the spindle.

    Still looking for a scope. Will look for a USB isolator. I have a laptop connected to the USB, which is mostly not plugged into the charger, so it may not be earth grounded. It was not charging in the tests above.

    Still looking for ground issues.

    Thanks

  14. #234
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Sorry my chart cluttered up. I had it nicely spaced. I can redo it a different way if needed.

  15. #235
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    See attached picture of the "Battery Box"

    Click image for larger version. 

Name:	Battery Box.jpg 
Views:	0 
Size:	99.2 KB 
ID:	461908

  16. #236
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    I see USB isolators from $15 to $130. Do you have a preferred isolator for these systems?

    Thanks

  17. #237
    Join Date
    May 2006
    Posts
    4043

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Charles,

    I'm not aware of any that don't work. KFLOP uses the "Full Speed" 12 Mb/sec mode. If using the ADUM4160 chip it should work. See also here.
    TK
    http://dynomotion.com

  18. #238
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    Did you see the above tests with the battery box grounded to Kanalog?

  19. #239
    Join Date
    May 2006
    Posts
    4043

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Charles,

    Did you see the above tests with the battery box grounded to Kanalog?
    Did you mean to say 13 RPM variation when grounded rather than 3RPM?

    I assume:
    "grnded" = battery box connected to Kanalog GND with no jogging
    "Jog grnded" = battery box connected to Kanalog GND with jogging
    "Not grnded" = battery box floating with no jogging

    Seems like further evidence there is noise on the grounds being introduced into the command.

    Without a scope it hard to say the frequency or amplitude of the noise. It might be 60/120Hz or high frequency. You might be able to measure it with the AC setting on a Voltmeter.

    You might also try adding an RC Low pass filter at the input of the Spindle Drive. Something like a 0.1uF Ceramic capacitor across the input pins and then a series 2K Ohm resistor in the command before the capacitor, That would have a cutoff at about 300Hz.
    TK
    http://dynomotion.com

  20. #240
    Join Date
    Jun 2006
    Posts
    194

    Re: Warner/Swasey QCM-6 lathe with Kflop/Kanalog

    Hi Tom,

    The speed variation of 3 RPM while grounded and .3 not grounded refers to the RPM reading fluctuations in a current state. I simply noticed that without grounding, the RPM was more steady than when it was grounded.

    Your assumptions on the chart headings are correct.

    Thanks

Page 12 of 13 210111213

Similar Threads

  1. Warner Swasey WSC-6 Fanuc 0t
    By Metzenw in forum Fanuc
    Replies: 3
    Last Post: 10-23-2018, 04:43 PM
  2. wsc-6 warner swasey turret HOT
    By shmarkey in forum Uncategorised MetalWorking Machines
    Replies: 0
    Last Post: 03-05-2016, 01:15 PM
  3. Need Help, Warner & Swasey WSC - 15 cnc lathe
    By Mohak in forum Machinery Manuals / Brochures
    Replies: 0
    Last Post: 01-02-2012, 02:41 PM
  4. Warner Swasey
    By John Leighton in forum Want To Buy...Need help!
    Replies: 0
    Last Post: 05-26-2010, 10:01 AM

Tags for this Thread

Posting Permissions

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