Tree 325 retrofit-Expectations versus capabilties


Results 1 to 16 of 16

Thread: Tree 325 retrofit-Expectations versus capabilties

  1. #1
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default Tree 325 retrofit-Expectations versus capabilties

    Hi,

    So, I've decided I want the project of upgrading my control anyway, turns out my control was fine and it was just a card that needed re-seating. Everyone needs something challenging to do, right?
    Here's what I want to interface with my Kflop, Kanalog. I'm trying to decide between KmotionCNC and Mach3... so, let me outline what I'm trying to have in this retrofit:

    Also, if you could provide a very high level explanation for how to do each of these things, that would be much appreciated. (high level meaning: "Mach 3 has some options to encorporate thisby setting options on XX dialog box, OR, you'll need to write a C program to do that logic.") I'll get into the details later. I'm just trying to make sure I get the right hardware at this point.


    -E-stop loop: Hardwired to break NC loop if there are servo overload errors, spindle VFD errors, over-travels, or if the pushbutton is pressed. Will have FET EnableAll switch in this loop as well. Does KMCNC or mach3 have a directly controllable E-stop bit?

    -Feed rate override, spindle speed override (probably use GPIO ADc's and LVTTL inputs for these. digital Inputs are for the "off" detent in the FRO/SSO knobs.) I want it to integrate directly into either mach3 or KMCNC, not use a keyboard emulator.

    -Cycle start, Cycle hold (Opto-ins)

    -MPG - There is an A and B channel, no A- or B-. I'm going to use pins 9 and 11 of JP2 for this. Pins 1, 3, 5, and 7 will pick up the index pulse for the servo encoders.
    MPG to have .0001, .001, and .010 graduation selector. GPIO inputs.
    Also to have X Y Z selector, GPIO inputs. Should have just enough for this.
    How do I get the MPG to interface to KMCNC and mach3, or can the MPG be handled on a KFLOP level, and just report the move to M3/Kmotion?

    Spindle: +/-10V spindle, also has reverse and forward relays. (IE, reverse on, negative 5 volts will give me a forward spindle motion @50% spindle speed maximum). Use analog out, and a couple of opto outs for this. I'd imagine this is straight forward enough?

    -Not allow cycle start if air pressure and lube pressure are not made. (likely to use opto-in's for this, since the switches are fed from 24V source.). Also display an alarm saying that air pressure is low or oil level is low.

    Tool change - tool change is manual on this mill. I'll need two opto In's (clamp and unclamp) and 1 relay driver (air solenoid) Where can I write the logic to prevent me from ejecting a tool while the spindle is running, etc?

    Servos: Analog outs.

    Does all of this sound reasonable enough for Kanalog, KFLOP, and KmotionCNC/Mach3 to handle? Or do I need to incorporate more hardware interlocks in this system, using relays and such?

    Similar Threads:
    Last edited by nickcornilsen; 10-03-2012 at 03:14 PM.


  2. #2
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default

    Hi Nick,

    Regarding:

    EStop – Mach3 has an Estop input that you can assign. With KMotionCNC it is usually sufficient that the amplifiers are disabled. Otherwise you could use a C Program to monitor a signal and do something else such as disable the Axis Channels.

    FRO/SRO – for Mach3 I suppose you could assign external trigger (Digital inputs) to screen buttons. For KMotionCNC there is a KFLOPtoPC C command to adjust the FRO on the KMotionCNC Screen. There isn’t a SRO on the display but the speed could be adjusted within KFLOP.

    Cycle Start – for Mach3 an external trigger could be used. For KMotionCNC there is a KFLOPtoPC C Command for Cycle start.

    Cycle hold/feedhold – for both Mach3 and KMotionCNC it is usually best to do a hardware feedhold in KFLOP. This guarantees instant response without any potential Windows delays.

    MPG – for both Mach3 and KMotionCNC it is handled in KFLOP in a C program. See the MPGSmooth.c program. This allows instant guaranteed response and any type of selector switches you wish. KFLOP C programs are usually fast enough to handle MPG encoder inputs in software. This allows any input to be used and doesn’t consume a hardware encoder input. But one of the 8 hardware encoder inputs can also be used.

    Spindle Speed Control – A C program is used to do whatever is necessary (set DAC, relays, delay, change gears, etc).

    Air/lub checks – For Mach3 you might need to change the Cycle Start behavior using Cypress Basic to do the checks. For KMotionCNC not sure if it would be easy to intercept cycle start. A watchdog C program might be used to disable axes or display a message if the wrong conditions exist.

    Tool Change sequence – For Mach3 it could be in a Cypress Basic Macro or KFLOP C Program. For KMotionCNC it would be in a C Program. Safety checks should be similar to the Air/lub issue.

    Analog outputs – Kanalog allows Mach3/KMotionCNC + KFLOP to work with analog servos.

    External hardware safety protection is always more reliable than software protection.

    Hope this answers most of your questions.
    Regards

    Regards
    TK http://dynomotion.com


  3. #3
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default

    Tom, thanks for taking the time to write out that detailed answer to my questions! That will help me greatly in drawing out the electrical wiring diagram.


    Quote Originally Posted by TomKerekes View Post
    Hi Nick,

    Regarding:

    EStop – Mach3 has an Estop input that you can assign. With KMotionCNC it is usually sufficient that the amplifiers are disabled. Otherwise you could use a C Program to monitor a signal and do something else such as disable the Axis Channels.

    FRO/SRO – for Mach3 I suppose you could assign external trigger (Digital inputs) to screen buttons. For KMotionCNC there is a KFLOPtoPC C command to adjust the FRO on the KMotionCNC Screen. There isn’t a SRO on the display but the speed could be adjusted within KFLOP.

    Cycle Start – for Mach3 an external trigger could be used. For KMotionCNC there is a KFLOPtoPC C Command for Cycle start.

    Cycle hold/feedhold – for both Mach3 and KMotionCNC it is usually best to do a hardware feedhold in KFLOP. This guarantees instant response without any potential Windows delays.

    MPG – for both Mach3 and KMotionCNC it is handled in KFLOP in a C program. See the MPGSmooth.c program. This allows instant guaranteed response and any type of selector switches you wish. KFLOP C programs are usually fast enough to handle MPG encoder inputs in software. This allows any input to be used and doesn’t consume a hardware encoder input. But one of the 8 hardware encoder inputs can also be used.

    Spindle Speed Control – A C program is used to do whatever is necessary (set DAC, relays, delay, change gears, etc).

    Air/lub checks – For Mach3 you might need to change the Cycle Start behavior using Cypress Basic to do the checks. For KMotionCNC not sure if it would be easy to intercept cycle start. A watchdog C program might be used to disable axes or display a message if the wrong conditions exist.

    Tool Change sequence – For Mach3 it could be in a Cypress Basic Macro or KFLOP C Program. For KMotionCNC it would be in a C Program. Safety checks should be similar to the Air/lub issue.

    Analog outputs – Kanalog allows Mach3/KMotionCNC + KFLOP to work with analog servos.

    External hardware safety protection is always more reliable than software protection.

    Hope this answers most of your questions.
    Regards




  4. #4
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default

    Ok, I've pretty much got the wiring diagram done. Luckily I don't have to change much, infact I was able to reverse engineer the pin-out on a few of the existing D sub connectors and won't even have to touch them!

    The one thing I'm having a problem figuring out, however, is how to get a 0 to 10v signal to the JP6 analog input. I've attached a screenshot of how I plan to drop 24Vdc to 10V with one voltage divider, then use the pot as yet another voltage divider to get 0 to 10v, with a pull-down resistor to hold it at 0 since this switch has an "off" detent at one end of the scale... I assume I want to drop the 24v to 10V, instead of coming straight off the 5V rail, so I can double my resolution and reduce any noise effects, right? Or will it really matter? The potentiometers will be on the user interface panel, outside the main machine enclosure if that makes a difference.

    Will this work, or have I found one of the many ways to release the magic smoke from one or several components?

    Also, for the MPG... since I only plan to use 3 axis, and no index pulses, I have 4 encode inputs free (or, 8 as it were). I was thinking I'd run the single ended encoder outputs to pins 1 & 3 of JP2, and tie pins 2 and 4 to ground. Then, run the 5v axis and resolution select to the other + pins in JP2, and tie the other - pins to ground also. Will this work?

    Attached Thumbnails Attached Thumbnails Tree 325 retrofit-Expectations versus capabilties-frossoidea-jpg  


  5. #5
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default

    Hi Nick,

    I think you need a bit more voltage division. I calculate the output voltage range of your circuit will be about 11V.

    The 400K in parallel with the 50K pot is
    1/(1/400K+1/50K)) = 44K
    Then the voltage division of a 44K and 50K is
    24V x 44/(44+50) = 11.2

    But actually the input impedance of Kanalog is about 100K so it should workout less than 10V.

    I’m not sure what the detent does that you refer to, but the 1Mohm will not be sufficient to force the input to 0V. But I don’t think it is necessary. All the values are fairly arbitrary. You can code the program to look for whatever ADC value you get.

    The circuit looks pretty safe and unlikely to let out any smoke :}

    I’m not aware of anyone who has done this. I’m not sure how you are expecting this to work when there is a FRO on the Screen and a FRO external Knob. For example if the User moves them both in opposite directions what is supposed to happen?

    You can not connect single ended encoder to differential inputs by grounding the un used pin. See:

    Regards

    Regards
    TK http://dynomotion.com


  6. #6
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Why no sub the 50k res for a 200ohm to 500 ohm and replace the 400k with a 10v zener?
    You will have a constant 10v source, a 1µf to 10µf across the zener would also be advised.
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

    “Logic will get you from A to B. Imagination will take you everywhere.”
    Albert E.


  7. #7
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by Al_The_Man View Post
    Why no sub the 50k res for a 200ohm to 500 ohm and replace the 400k with a 10v zener?
    You will have a constant 10v source, a 1µf to 10µf across the zener would also be advised.
    Al.
    I've never heard of a zener diode... but I'll look into it! thanks for the tip..

    Tom, I'm not sure yet how the FRO and SSO will play with Mach3. I can edit the screens in mach3 I beleive... so I'll probably remove it entirely fro the software interface, and instead display only a FRO/SSO override %... The detent I was refering to is a 100% disconnect of the potentiometer circuit at the extreme left of the potentiometer scale. There is a detectable "click" when it goes into this position.

    As for the single ended encoder inputs - easy enough, I'll connect them to the RJ45 connectfor on KFLOP. How do I size the extra termination resistors for this?



  8. #8
    Community Moderator Al_The_Man's Avatar
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    24216
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by nickcornilsen View Post
    ... The detent I was refering to is a 100% disconnect of the potentiometer circuit at the extreme left of the potentiometer scale. There is a detectable "click" when it goes into this position.
    Are you sure it is not just an on/off switch? These are typical on volume controls and the high Pot value would also indicate it.
    If it IS a volume control then it is the wrong type, as these are logarithmic resistance taper, you need the linear taper for this application.
    Zener diode - Wikipedia, the free encyclopedia
    Al.

    CNC, Mechatronics Integration and Custom Machine Design

    “Logic will get you from A to B. Imagination will take you everywhere.”
    Albert E.


  9. #9
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default

    These are the pots the mill was originally equipped with, at 50% dial travel the reading is 26K, so I think they're linear.



  10. #10
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default

    Al's suggestion of the zener is a good one. It acts like a voltage regulator so any noise or voltage fluctuation on the 24V supply will be eliminated. I'd probably go a little higher (1K) on the series resistance so the resistor doesn't dissipate so much heat. 14V across a 1K is 0.196 watts.


    As for the single ended encoder inputs - easy enough, I'll connect them to the RJ45 connector on KFLOP. How do I size the extra termination resistors for this?
    Reply With Quote
    The termination resistance should be around 150ohms.

    Regards

    Regards
    TK http://dynomotion.com


  11. #11
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default

    Thanks again Tom.

    I've purchased the Zener Diodes, and will be picking up my Dynomotion boards at the post office this morning when they open!

    I've also got a pretty cool case coming from PAR-metals. 87 bucks for a case with holes punched for the D-sub connectors, mount holes for the Kanalog/KFLop, fully alodine coated... I thought it was a pretty good deal!

    Now, one question that is perplexing me.

    My SD1525 servo drive cards are capable of accepting a differential analog input, or a single ended input. I've been reading about how a single ended input is a bit more susceptible to noise, whereas differential analog is better at rejecting common mode noise. My servo cards are separated from my Kanalog by about 6 feet of cable. It's STP cable the whole way. It is currently wired to the SD1525's as differential. So my questions are these:

    -Would my current configuration, with STP wire over a short distance be good enough to run as single ended?

    -If not, how do I get differential analog output? IS there a converter, or do I use two analog outputs, and configure one to mirror the other?



  12. #12
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default

    Hi Nick,

    I think you will be ok with the single ended Kanalog outputs, although 6 ft seems pretty far. Run the - input of the amplifier to a ground at Kanalog. In this way if there is ground level variation between Kanalog and the Amplifiers it will be theoretically eliminated. You probably want to connect the shield also to Kanalog DC ground at only the Kanalog end as well. The analog signals are not that high of frequency (tens of KHz).

    Regards
    TK

    Regards
    TK http://dynomotion.com


  13. #13
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default

    Tom, thanks yet again for your input. I appreciate the level of support you provide for your hardware.

    This should be my last question on the hardware side of things.

    I don't have enough optically isolated inputs to get all functions connected, so I'm going to have to use the GPIO to receive 3 input signals. These are less "critical" than the others, but they will stop motion if they pick up interference and change state in an uncontrolled manner.

    Currently, these switches are setup to source 24VDC to the controller. If I were to leave them as 24VDC, then put a 2.4K resistor just before they go into the Kanalog board, would this provide me with a level of noise protection since there is a 3.3V clamping diode on the Kanalog board? Or would a voltage divider be better that is independent of the integrated clamping diode?

    Would I also be best served with a pull down resistor after the zener/voltage divider? I can't tell what the impedance of the GPIO inputs is, what size would it need to be?

    Of should I come up with a opto-isolator suing a 4N25 Int. circuit?

    The wiring to these switches and sensors is unshielded, and it would be a royal pain to pull them out and replace since they go to the other control cabinet that has the spindle drive.

    Thanks once again,
    -Nick



  14. #14
    Registered
    Join Date
    Apr 2006
    Location
    USA
    Posts
    76
    Downloads
    0
    Uploads
    0

    Default

    What is it that you hope to gain from a retrofit? What control do you have now?

    IMHO the Dynapath Delta 20 on my Tree is a very capable control, especially considering it's age. I would happily take it over Mach3 or any other hobby level (I've used it, it is hobby level) PC based control.

    There are a few features missing: No G98/G99 for canned cycles, no rigid tapping, tiny memory, pretty slow for 3D work.

    However, the Delta 20 is a real industrial control. It runs on 24V, it uses real hardware switches, It has closed loop control of spindle and servos, it has an MPG, graphics are good, and it's pretty darn reliable.

    Mach3 is just not in the same league, despite the pretty interface. It has a penchant for 12V electronics, and software switches (key board emulators and others), has limited I/O, has no true closed loop servo control, has a few bugs, and uses Windows.

    The Delta 20 is about perfectly matched to the capability of the J325. It is not a super rigid or fast machine, no matter the control.



  15. #15
    Registered
    Join Date
    Sep 2012
    Location
    United States
    Posts
    44
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by UWP_Wes View Post
    What is it that you hope to gain from a retrofit? What control do you have now?

    IMHO the Dynapath Delta 20 on my Tree is a very capable control, especially considering it's age. I would happily take it over Mach3 or any other hobby level (I've used it, it is hobby level) PC based control.

    There are a few features missing: No G98/G99 for canned cycles, no rigid tapping, tiny memory, pretty slow for 3D work.

    However, the Delta 20 is a real industrial control. It runs on 24V, it uses real hardware switches, It has closed loop control of spindle and servos, it has an MPG, graphics are good, and it's pretty darn reliable.

    Mach3 is just not in the same league, despite the pretty interface. It has a penchant for 12V electronics, and software switches (key board emulators and others), has limited I/O, has no true closed loop servo control, has a few bugs, and uses Windows.

    The Delta 20 is about perfectly matched to the capability of the J325. It is not a super rigid or fast machine, no matter the control.
    You make some good points. I'm fully aware that the Delta 20 control is a solid control.

    Now, limited IO - Kanalog and the KFLOP board take care of that, and provide me with closed loop servo motion. Granted, it isn't closed back to Mach3. I wish I had a few more optically isolated 24V inputs, but I'm thinking I can create my own inside the Kanalog breakout board enclosure.

    But, I will have a MPG. I will use the original air pressure, lube level, travel limit switches, the yaskawa drive, so it isn't a hobbiest grade retrofit.

    I may not even USE mach3... I may use the Kmotion CNC control, if I can't get what I need from Mach3.

    I also get this - I'm no longer held hostage to an obsolete control. I don't have to pay $300 bucks to simply find out what's wrong with the damned thing, let alone fix it!



  16. #16
    Member TomKerekes's Avatar
    Join Date
    May 2006
    Location
    USA
    Posts
    4043
    Downloads
    0
    Uploads
    0

    Default

    Currently, these switches are setup to source 24VDC to the controller. If I were to leave them as 24VDC, then put a 2.4K resistor just before they go into the Kanalog board, would this provide me with a level of noise protection since there is a 3.3V clamping diode on the Kanalog board? Or would a voltage divider be better that is independent of the integrated clamping diode?
    Hi Nick,

    Optical Isolation is always the best approach, but the signals coming in through a 2.4K ohm should work fine. I would place a 350ohm divider resistor to keep the voltage going into the Kanalog GPIO to less than 3.3V as I don't think it is a good idea to constantly feed current into the clamping diodes. The Kanalog GPIO inputs are MOSFET inputs so the impedance is very high (megohms). I would also place a 0.1uF capacitor across the 300ohm resistor to filter any noise.

    Regards

    Regards
    TK http://dynomotion.com


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
  •  


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Tree 325 retrofit-Expectations versus capabilties

Tree 325 retrofit-Expectations versus capabilties