Announcement

Collapse
No announcement yet.

Cnc Parallel Port Software Breakout Board

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Cnc Parallel Port Software Breakout Board

    Any of you in the "know" have an opinion about this?

    http://cgi.ebay.com/CNC-PARALLEL-POR...19105718412152

    Regards,
    -SD:

  • #2
    It looks pretty cheezy compared to this one:
    http://www.probotix.com/breakout_boa...reakout_board/

    I am using the Probotix board, along with a Smooth Stepper, to refit my DynaMyte2400 bed mill.


    Rex

    Comment


    • #3
      I am a bit wary of R.O.C. electronic purchases, I bought a OBD2 for extracting codes from current vehicles, piece of junk, ended up scrapping it.
      The thing that amuses me in Break out Boards like the one in the link, is they claim total isolation from PC to coolant pump or spindle, the truth is that if the motors are mains fed such as induction or universal motors, however much isolation you build in, they end up all being common by virtue of the PC power supply being referenced to ground and the motors via the grounded neutral, so all you end up with is transition from one voltage level to another, but NOT isolation.
      Max.

      Comment


      • #4
        It probably does what they say it does and it looks simple enough to be easily repairable.

        One concern is the tan colored circuit board. Fiberglass boards are generally not that color. It could be a cheaper, more easily breakable, board material.
        Paul A.

        Make it fit.
        You can't win and there is a penalty for trying!

        Comment


        • #5
          isolation

          Isolation on the stepper drives does serve a function although the two sides operate at almost the same potential. If you wire your grounds right, you shouldn't need isolation. If you wire them wrong, it is another matter.

          Your power supply may have three terminals: "+", "-", and ground. Do not connect "-" to ground on the power supply end (with the power supply having a grounded cord). If you do, what happens is that all the current flowing through the motors flows through the -/ground wire and causes a voltage drop across that wire. This forces the ground at your stepper drivers to be at some level above actual ground. Some of the current may decide to try to find an easier path to ground: through your computer. It can also mess up the logic levels. Run three wires and tie ground and - together in a star ground at the stepper motor controllers or just leave the ground terminal on the power supply disconnected (but there is still a ground throught the power strip). Either way, "-" is allowed to float a little below ground to make up for the voltage drop across the wire.

          For a spindle motor control, isolation is usually appropriate, or if you have really huge motors such as on a giant gantry.

          The board linked by the OP did not have optoisolation on most of the signals although it looked like it might have a few isolators (or just fets?) on some auxiliary signals. Instead it has what looks like a crappy transistor buffer. May hurt more than it helps. And, yeah, it looks like a crappy single sided phenolic board. And you have to take precautions because you have 117VAC exposed on part of the board.

          Also beware of the screw terminal strips. There are good ones and there are ones which will cause you grief. The ones with a little spring piece between the screw and the wire (sometimes called a wire retention clip) are awful; they have to be retightened.

          Comment


          • #6
            I use the opposite approach, I Earth ground ALL supplies and have have no problems and have actually solved many problems of users that have noise related problems, as I mentioned, if using AC Mains fed motors, there is no isolation however many opto isolators you use, you would need to modify the PC common and remove its ground reference in order to achieve this.
            Max.

            Comment


            • #7
              I have used a few different breakout boards. None of them use true isolation but they are buffered. This can help if you have a low parallel port voltage. Also if they use tri-state buffers you can disable pulses passing through to drives during boot and estop conditions. Windows will probe the parallel port and pins can go high and low causing potentially bad things to happen.

              PMDX.com has some real nice boards, I have used two of them in the past. First was a PMDX-120 (No longer made) and the most recent was a PMDX-122 on my hercus lathe. They have a watchdog circuit which receives a pulse train from mach. If it is interrupted the board will not pass signals. It also has a relay for turning stuff on/off.

              http://www.pmdx.com/PMDX-122

              I also have used the boards from CNC4PC.com, I used two in my CNC Mill. They will also disable their outputs if need be.

              http://www.cnc4pc.com/Store/osc/prod...products_id=45

              Comment


              • #8
                Originally posted by Paul Alciatore
                It probably does what they say it does and it looks simple enough to be easily repairable.

                One concern is the tan colored circuit board. Fiberglass boards are generally not that color. It could be a cheaper, more easily breakable, board material.
                It a phenolic board instead of a fiberglass. For a simple single layer board like this it really does not matter. Very common in consumer grade appliances.

                Comment


                • #9
                  I use the opposite approach, I Earth ground ALL supplies and have have no problems and have actually solved many problems of users that have noise related problems,...
                  Grounding is a source of many problems. You must be careful how you ground each and every power and signal line in the entire system. Earth grounding can turn into a huge problem if it isn't done right. It is most important that there only be one ground path between separate pieces of equipment. When a computer is connected to a controller the signal leads should be shielded with the shield connected to ground at the plug on the computer. At the other end the shield should NOT be DC connected to the ground plane of the breakout board or the case of the controller. Instead it should connected via a low value capacitor that will pass fast rise time spikes but not ac hum or DC offsets.

                  All machines and electronics should be grounded at the same single point. That means the controller and the computer should be grounded at the same power bar. The machine itself should not share ground with the controller or computer. All signal lines to/from the machine should be shielded and the shields should only be grounded at the controller end. Power lines to motors that originate in the controller (steppers, servos) should be shielded and grounded at the controller only.

                  Power lines that are supplied from elsewhere such as the spindle motor should be shielded at the machine end and grounded there. The spindle control should be entirely isolated in the controller using both a contactor and a solid state switch for power switching. There should be no contact between any part of the circuit and the controller. The most reliable way to do this is to put the motor control switching on the machine outside the controller and operate it with a low impedance, low voltage signal line.

                  Under no circumstances should the machine and the controller be grounded together if they do not share the same AC circuit. If they do share the same circuit they should not be grounded together but if they are it should be at the same single point as the computer.

                  The most important concept is that all grounding must have only one path to reach any other grounded part of the system. Grounds carry stray currents and if there is more than one path there will be a difference in the resistance and AC impedance of the paths that will result in unwanted coupling of stray signals to the wires the shields are supposed to protect.

                  It's called "Single Point Grounding" and it eliminates the problems of "Ground Loops".

                  It is entirely possible to totally isolate the computer from the rest of the system. All it requires is that all control lines, both input and output be optoisolated with no common ground. The computer itself is operated on an isolation transformer so it has no common ground with the rest of the system. For ultimate protection the control signals from the computer and the signals from the controller to the computer may be run with fibre optics. This is no longer esoteric or even unusual. Think of a laptop running on battery power.
                  Free software for calculating bolt circles and similar: Click Here

                  Comment


                  • #10
                    Rather than adding or removing grounds indiscriminately which can mask problems without actually solving them or even make them worse, you need to address the source of the problem.

                    The source of the problem is the motor return current. You need to keep that out of the ground system used by logic. You do this by allowing only one path where it matters, back to the negative terminal on the power supply. This goes through the return wire "-" only, not have a choice between going through the return wire or system ground.

                    Breaking the shield on the parallel cable does not help. There are situations where isolating the shield is appropriate, this isn't one of them. Your computer can still be fried through the signal return leads and the signal leads and keeping the shield connected is likely to reduce the probability of those happening even though the shield itself could be a path for damage. But the shield of the parallel cable is normally connected in such a way that it connects to frame ground on the computer rather than logic ground.

                    It is not correct that signals should only have one path to ground, although there are many times when this is desirable. What is important is what signals share a path to ground. It is a misconception that ground loops are created by having multiple paths to ground or physical loops in your ground connection. Those extra connections may cause problems because they may create a shared path but they may also improve things. You don't necessarily fix ground loops by bringing a bunch of wires together to a single point ground. In complex systems, true single point grounds are not achievable. Your typical home shop CNC system does not permit a true single point grounding system. There will be physical loops in your ground topology; there will also be at least minor "ground loops" (or severe ones if you screw up the wiring) which are not the same as the aforementioned physical loops, though they may be related. Each stepper driver IC REQUIRES a connection between the ground on the logic side and V- on the power side. This is where you want your ideal single point ground to be; unfortunately, there are more than one. So all you can do there is produce a reasonable facsimile of a single point ground with a lot of copper. One wire from each V- lead on each stepper driver to your star ground or two from separate screw terminals (because screw terminals can fail) if you can. Actually soldering a wire on to the board and bypassing the potentially unreliable screw terminals is actually a good idea. One wire from the logic ground of each stepper driver to your star ground if you don't already have two from the V- on the power side will probably do more good than harm but it is much better to have two on the power side. Make sure there is redundancy from the V- terminal of each driver to the star ground. One wire from the ground on your breakout board logic ground to your star ground. One or more wires from your star ground to the V- on your power supply. But this star ground is the ONLY place where ground and V- are allowed to meet in the external wiring (they also meet on each driver board). You may connect one wire from your star ground to the case of the computer butif you have optoisolators on your motor drivers, this may work against you. You may connect one wire from your star ground to the power strip ground. You may connect one wire from your star ground to your motor power supply ground terminal. There should be at least one connection from your star ground to the same ground used by the AC line.
                    You may connect a wire from your star ground to your machine tool frame (there are some pluses and minuses to this). You may not connect the ground terminal on your power supply or any other ground to the V- terminal on the power supply at the power supply end.

                    Keep the wires running to your star ground from your driver boards SHORT and heavy and use reliable connections at the star ground.

                    This configuration does create some minor ground loops, however they are better than the alternatives, and it avoids the ground loops that cause the most trouble. The power strip is not the right place to make your star ground - too much wire between there and the parts that matter if you rely on power cords. However, you may make your star ground using a bolt inserted through a hole drilled in the metal case of the power strip if, and only if, it is practical to mount the power strip immediately adjacent to the stepper drivers; if it is not located within inches of the driver boards, do not run long wires to the power strip ground in place of a star ground in immediate proximity to the driver boards. However, bear in mind that bolts may not make the most reliable electrical connections for multiple connections. Do not wrap bare wires around bolts. You may use soldered lugs; avoid crimp lugs especially insulated crimp lugs (the kind that are crimped through the insulation - these are notoriously unreliable when assembled using consumer grade crimp tools). Feel free to use bus bar located near the driver boards. Feel free to solder connections. You may use the type of <a href="http://store.solar-electric.com/tbb.html">busbar used in circuit breaker panels</a>, one wire per screw.

                    If there is a break in the V- connection between any motor driver and the star ground (or any other way you connect your V- connections together - this is in no way unique to the star ground), bad things may happen. That current will try to find an alternative path and you may not like the path it takes. It will try to go through the logic ground to another driver board and from there out the driver board's V- connection. Some of the traces in route may not be able to handle it. If they fail, it may then try to go through ICs on your computer motherboard, which will not survive.

                    Optoisolators can improve the situation because they isolate the computer ground from the driver ground even though these will ultimately be connected together at the power strip and be at almost the same potential in normal operation but they can be very different with bad wiring or in certain failure scenarios. Thus they eliminate damage in certain scenarios and can improve signal integrity (though some bad isolator boards may make signal integrity worse). However, a star ground as described here with redundant and reliable connections, which you should have anyway, can eliminate most of the same damage scenarios. The best place for the optoisolators is actually on the driver boards. Optoisolators stop two damage paths: through the ground leads (multiple) of the computer cable and through the data lines. Optoisolators will give better protection if your spindle motor shorts from the AC line to the frame of your machine. Also bear in mind that the risk to your computer is probably not going to be zero, especially as you start connecting more stuff. Vision systems, touch probes, DRO scales, tool height sensors, etc may create additional risks; the optoisolators on your drivers don't protect you against independent paths.
                    Certain failure modes such as the spindle motor short may find a path through your computer power cord ground, through the computer, through a cable to a peripheral, and from the peripheral to a grounded object.

                    When you use the star ground as described here, some items will also be connected to ground on your power strip as well. This creates some physical loops in the ground topology. Get used to the idea. The power strip itself normally makes a poor main star ground. Long wires, ground connections which may be unreliable, and lack of individual wires from each of the points that needs to be grounded.

                    Power supplies should be grounded on the AC input side.
                    Use a fat wire on V- to reduce the voltage drop across this wire.

                    Optoisolators are a good idea. However, they can be expensive insurance where the cost outweights the risk it is protecting against if you use due diligence in your wiring. If your wiring is poorly thought out or you have unreliable connections on V- pins, it may be worthwhile.

                    If you have a spindle motor controller connected to the computer it does generally need to be isolated (except for something like a low power DC operated dremel tool).

                    Comment


                    • #11
                      Rather than adding or removing grounds indiscriminately which can mask problems without actually solving them or even make them worse, you need to address the source of the problem.
                      Who is talking about removing/adding grounds indiscriminately? I have given very well defined rules for grounding.

                      Breaking the shield on the parallel cable does not help. There are situations where isolating the shield is appropriate, this isn't one of them.
                      I didn't recommend isolating it. I recommended using a capacitor to connect it at the one end so that DC offsets cannot cause problems. This is standard practice, not something I invented.

                      Your computer can still be fried through the signal return leads and the signal leads and keeping the shield connected is likely to reduce the probability of those happening even though the shield itself could be a path for damage.
                      It cannot be fried through the signal return leads if they are not connected to anything except the optoisolator LED or Phototransistor. An optoisolator is powered only by the signal and has no ground connection. The shield is intended to prevent induction of stray voltages that are changing rapidly. It will still provide a path to ground while connected by a capacitor for such impulse induction AND cannot provide a path for damaging offset currents that might appear due to poor grounding of the controller.

                      But the shield of the parallel cable is normally connected in such a way that it connects to frame ground on the computer rather than logic ground.
                      Nonsense. They are the SAME THING. If you don't believe me go measure for continuity between the black wire in a power supply harness (logic ground) and the power supply case (frame ground).

                      You don't necessarily fix ground loops by bringing a bunch of wires together to a single point ground.
                      Yes you do. It isn't a "bunch of wires". You bring the ground wires from each item to the same single point. The breakout board will have a ground plane. It should be mounted on insulated standoffs and the ground plane taken to the single point common ground in the controller. The same applies for each board such as the drivers, the power supplies and so on. This isn't a random exercise. I have been working on machine controllers since 1976.

                      Your typical home shop CNC system does not permit a true single point grounding system.
                      Yes it does. To do so requires careful attention to not creating multiple ground paths but there is nothing to stop it being done.

                      This configuration does create some minor ground loops, however they are better than the alternatives, and it avoids the ground loops that cause the most trouble. The power strip is not the right place to make your star ground
                      The power bar is where the grounds from separate devices meet. The devices within any particular enclosure are connected in a star topology within their respective enclosures. Then the enclosures are connected in a star topology at the only point of common ground which is where they are plugged in. Any other topology will create ground loops.

                      Optoisolators can improve the situation because they isolate the computer ground from the driver ground even though these will ultimately be connected together at the power strip
                      Optoisolators do not have a power connection.

                      However, a star ground as described here with redundant and reliable connections, which you should have anyway, can eliminate most of the same damage scenarios.
                      You do not understand the concept of single point grounding, that is clear. Redundant ground connections are to be avoided, not created.

                      I suggest that you do some studying on this topic before replying. See here:

                      http://www.google.ca/search?hl=en&q=...=&oq=&gs_rfai=

                      However, they can be expensive insurance where the cost outweights the risk it is protecting against...
                      A 4N32 industry standard optoisolator costs 35 cents in single unit quantity.
                      Free software for calculating bolt circles and similar: Click Here

                      Comment


                      • #12
                        I absolutely agree with Evan (Wow!) on these points:-

                        Star earth all secondary negatives on all the power supplies. Take all machine earths to the same point.
                        Shield all cables with braided copper and connect only one end to earth. (Haven't tried the capacitor idea.)

                        Both my Denford mill and Boxford lathe were converted to Mach3 using the above principles and have given no problems.

                        OP's question - no I have not used these particular boards but would recommend CNC4PC as well, I used their C11's. Look at their site here :-

                        http://www.cnc4pc.com/Store/osc/index.php

                        And Sir John's Sieg wiring here:-

                        http://cnc4pc.com/Tech_Docs/C11G_&_S...0storybook.pdf

                        IanR
                        Last edited by dvbydt; 09-01-2010, 10:39 AM.

                        Comment


                        • #13
                          Originally posted by dvbydt
                          Star earth all secondary negatives on all the power supplies. Take all machine earths to the same point.
                          Shield all cables with braided copper and connect only one end to earth. (Haven't tried the capacitor idea.)
                          I have used the common grounding method of all power supplies in the PC based systems I have built and installed over the last 25yrs, they are still giving trouble free service to this day in some rough environments.

                          The Siemens site has a 8meg PDf on the how-to of grounding and bonding in machine control and electrical enclosures.
                          Some appear under the impression that bonding straps introduce ground loops, when actually they are used to defeat them.
                          Which ever method you adopt, beit total isolation or common grounding it has to be done methodically.
                          Max.
                          Last edited by MaxHeadRoom; 09-01-2010, 11:05 AM.

                          Comment


                          • #14
                            Originally posted by macona
                            It a phenolic board instead of a fiberglass. For a simple single layer board like this it really does not matter. Very common in consumer grade appliances.
                            Personally, I feel that durability is a very important concept in a shop. In general, I do not like phenolic boards. But, yes they can be OK if kept inside proper outer enclosures and treated with some care when wiring to them. But you can beat a fiberglass board with a hammer without harming it. They are at least ten times stronger and more durable.

                            Evan is right on the proper grounding idea. And do use a ground conductor that is at lease as heavy as the current carrying ones, preferably a size or three larger when logic circuits are concerned. Optical or RF isolation is nice, but not actually necessary if proper care is taken. The board shown by the OP appears to have a row of transistors for the output side. They are most likely configured in the open collector type of circuit which means that they are the only components that are in any danger from any reasonable spikes that may be present in the circuits connected to the board. A spike going backwards through such a transistor will be attenuated instead of being amplified. The base resistors, which are almost certainly present, would further attenuate these spikes and almost nothing would be left to back feed into the computer or any logic chips on the board. If you have problems from spikes, they are probably very severe and you probably should run down the source of those spikes and eliminate them there as they would probably cause other problems as well.

                            An exception to this would be if you are controlling circuits with higher Voltages (115, 230, 440 Volts). Then a failure in those circuits could feed back a Voltage which would exceed the maximum ratings of those transistors and of any and all logic IC and that could cause short circuits that could continue backwards into the computer itself. In other words, if you have logic level signals controlling line level circuits, use isolation that can withstand the line Voltage with a generous safety factor. But if the circuits you are controlling are 12, 24, or even 48 Volts, I don't think you need to worry about it.

                            I would not have any hesitation to use such a board - with proper grounding of course.
                            Paul A.

                            Make it fit.
                            You can't win and there is a penalty for trying!

                            Comment


                            • #15
                              Originally posted by Paul Alciatore
                              Personally, I feel that durability is a very important concept in a shop. In general, I do not like phenolic boards. But, yes they can be OK if kept inside proper outer enclosures and treated with some care when wiring to them. But you can beat a fiberglass board with a hammer without harming it. They are at least ten times stronger and more durable.

                              You make a hobby of pounding on circuit boards????

                              If the board is mounted in an enclosure you will never notice a difference. Heck, for that matter the thing could be hanging in the breeze and it wouldn't make a difference. Electrons only care about the copper and if you are banging on a board the components arnt going to last any longer than a phenolic board!

                              Comment

                              Working...
                              X