Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: TunerTom on August 12, 2009, 03:57:58 PM
-
I see the Smooth Stepper discussion area is not taking new posts.
I have been working with Mach3 for about three months now and am convinced that this application is working well for most of you, but I continue to have problems with instability (losing steps) and am curious if anybody thinks adding a Smooth Stepper card might help me.
Right now I have a PC running Windows XP dedicated to this project. I have run the System Test and it reports the system tests OK, and the "spikes" are nice & short... I have also uninstalled all programs from that PC except for Mach3 a G-Code editor, a file comparison utility, and a Parallel port monitor (none of which which are running when Mach3 is running) I have the graphic toolpath display turned off.
My PC inputs data from LPT1 to a Probotix PBX-RF RF Isolated CNC Breakout Board (and a Probotix relay board for spindle on/off) and tailstock (extend/retract). The oputput ot the RF isolator is wired into a Slo-Syn SS2000-DP4 drive and two Slo-Syn KML092F07 Stepper motors. I'm using Mach3 Turn. Everything works, except that the stepper motors (particularly the Z-Axis) fail to return exactly to their starting points, as My g-code calls for at the end of a run.
I have shielded cable from the output of the Probotix card, but the card itself is not installed inside a grounded metal enclosure. The output of the drive connects to the stepper motors using shielded cable. The shield is tied to earth ground. There is a splice at each stepper motor, and from there the wiring is exposed. The shield is tied to earth ground. The exposed leads are about 3" long at each motor. The Z-Axis motor is particularly unstable; it misses the home position most of the time and the error ranges from (say) three or four pulses to maybe 25 pulses. Once in a "blue moon", Z will return to zero. X is the other way. X might miss 1 or two steps "once in a blue moon" but returns exactly to zero, most of the time.
I have tried numerous settings of maximum speed settings in Motor Tuning; I have tried numerous settings of Accel/Decel to no avail. I have tried using the Sherline 1/2 step mode, which helps occasionally. I have turned Sherline off and tried pulse widths as high as 14. (I read that 15 or higher can cause the system to lock up). A max speed of 300 mm/min would be nice to obtain.
I wrote a simple G-code program that exercises the saddle in the Z-Axis back & forth at full speed (G0) for 50mm four times... then the X-Axis in/out 50mm four times at full speed (G0) four times. This profile NEVER MISSES! The problem seems to show up when curring curves and both motors are running simultaneously at varying rates...
There has to be an answer to this problem.
Is there anything wrong with my choice of the Probotix RF isolator?
Would getting a Smooth-stepper card help? If so, does the Smooth Stepper provide the necessary isolation? Should I then install a simple breakout board on the Probotix's place?
I'm going to go to the shop tomorrow and run through the procedures in your report to optimixe Windows... Maybe that will help.
Any suggestions or answers would be greatly appreciated.
I like the idea of the Smooth Stepper using the USB port and that it's parallel pouts are probably more stable than the PCs LPT1 port.
It even suggests that we might be able to use a laptop at the lathe... Another plus for us...
My sincere thanks for any advice you might be able to share...
-
Where in the world are you, just with you saying mm.
Hood
-
I'm in Massillon, Ohio, USA
Are you near Loch Ness?
-
Ah ok, just if you were in the UK I could have loaned you a SS to see if it sorted you out.
Nope probably 150 plus miles away from Nessie, then again to you USA guys thats probably close ;D
Hood
-
Hood,
I will gladly buy the SS if someone thinks it will help.
If there is some glaring error in my present setup, I hope somebody will tell me...
Tom
-
Cant say whether it would help or not as dont know where the issue lies :( If it was a bad pulse then yes it would, if its a problem outwith the computer then likely it wouldnt.
Hood
-
Thanks...
-
OK, but we have a fairly new laptop running Windows Vista.
Is it worth trying an SS on that machine?
-
Not a Vista fan myself and never used it for Mach but I think quite a few people do.
Do you have a scope? If you do check the pulse that is going to the drives.
Hood
-
I don't like Vista either
I have the following scope on order... $55.00
Hope it will be adequate.
http://www.seeedstudio.com/depot/digital-storage-oscilloscope-with-panels-p-167.html
-
Dont know too much about them but looks like it will do the job.
Hood
-
What voltage are you running for the steppers?
-
The motors are 3.5 volt so I assume the drive is outputting something close to 3.5VDC? I didn't build that part of the lathe. I do have a scope coming... so I can measure the voltage then.
I can say, the motors have PLENTY of torque... I don't think I can stall them by holding the hand-wheels on hte lead screws.
-
Stepper motors should run on 3 to 25 times their rated voltage. I have had trouble with lost steps below about 10x voltages. Stepper problems do not manifest at a steady state. That is when they are the most powerful. It is at speed where the power drops off sharply that the problems occur.
You mentioned a parallel port monitor. That sounds like a bad idea. If the program is, or if it uses a TSR type of utility, it may be polling the port occasionally. That would be enough to disrupt the pulse stream. Also, remove any printer drivers, and run the port without interrupts.
If you look back a couple of months in the postings, or do a search, I posted info on an add-on parallel port that works with Mach. I have found it to be faster and a little more robust that the built in port on older machines. It is dirt cheap and would answer the question as to whether your existing port is the problem.
You will see a lot of posts showing calculations for max stepper speed based on the pulse rate Mach is running at. Remember to consider that the pule rate is the total and must service all of the drives. While it is unlikely all of the motors would run at full speed simultaneously, it is unwise to set a speed based on all pulse frequency being available for each motor. This is probably why your motors work fine individually and loose steps when both are running.
Make sure any wire shields are grounded on one end only, preferably at the control box.
Lastly, use the driver test program to see how fast your computer can generate the pulse stream, but do not assume you can run Mach that fast. Back off at least one speed setting slower than what the driver test determines.
-
You will see a lot of posts showing calculations for max stepper speed based on the pulse rate Mach is running at. Remember to consider that the pule rate is the total and must service all of the drives. While it is unlikely all of the motors would run at full speed simultaneously, it is unwise to set a speed based on all pulse frequency being available for each motor. This is probably why your motors work fine individually and loose steps when both are running.
Mach can output the kernel pulserate on all axis simultaneously.
Hood
-
Hood,
That may be true in theory, but in practice, I can reproduce an error every time if I max out the stepper speed on more than one motor.
With the exact same motor settings, increasing the pulse rate makes the problem go away. This is true for XY with two steppers or also with one stepper and one servo (I am in the process of changing over).
My conclusion is that Mach is unable to provide the full pulse rate to all axis simultaneously and needs some headroom for whatever else it is doing at the same time.
Your results may be different.
I don't know what the breakpoint is, but if someone wanted to do a bunch of testing, maybe some guideline could be formulated, for example not to exceed 90% of maximum theoretical speed, or something like that.
Another area where theory and practice collide is in the driver test vs actual program execution. My observation is that there is a correlation between CPU load and kernel speed when Mach is running. I can state with confidence that a successful driver test at a given speed is no guarantee that Mach can actually perform reliably at that speed, as theory may suggest.
-
Simpson 36...
Thanks for your information. This is all sort of new to me. I will look more closely at the applied voltage to the motors.
The Parallel port monitor is not TSR. It is not running when Mack3 is. I used it to set/clear data bits to see what effect they had.
I will find your post. and look into all matters you have outlined above.
-
Something I forgot to mention is to leave the task manager running on top of Mach to observe the CPU load. If the CPU pegs, you loose steps.
I just chased down a new problem where my X axis stepper (640 oz-in) was loosing .020 after every flat out run (135 in/min 10 accel). If I drop it back to 90 and 8.5 (where it is normally set), it stops loosing steps. I am switching to servos and decided to have a foot race between the stepper and servo while I had one one each axis. Needless to say, the servo won. I don't know where in that range the stepper started loosing steps and since the stepper is out of there in about a week, there is no incentive to dig into it. I have to measure each time to find if the stepper has been a bad boy as there is only an audible clue if it skips quite a few steps at a time. Onesy - twosy misses are acoustically invisible.
-
Hood,
That may be true in theory, but in practice, I can reproduce an error every time if I max out the stepper speed on more than one motor.
With the exact same motor settings, increasing the pulse rate makes the problem go away. This is true for XY with two steppers or also with one stepper and one servo (I am in the process of changing over).
My conclusion is that Mach is unable to provide the full pulse rate to all axis simultaneously and needs some headroom for whatever else it is doing at the same time.
Your results may be different.
Certainly had no issues on the lathe when I was maxing it out on the PP at 45KHz, because it was a new build I was monitoring the following error in the drives software and no problems there. Possibly this would point to computer rather than Mach but no intentions of starting arguments so suffice to say mine conformed to the theory.
Another area where theory and practice collide is in the driver test vs actual program execution. My observation is that there is a correlation between CPU load and kernel speed when Mach is running. I can state with confidence that a successful driver test at a given speed is no guarantee that Mach can actually perform reliably at that speed, as theory may suggest.
To my mind the driver test is there to weed out the really bad computers that have no chance of running Mach in their present state rather than saying it will definitely run Mach.
Hood
-
Hood,
Arguments are emotional and useless. Debates are factual and beneficial. Methinks we are both fact guys. You are far more experienced than I on these topics and I certainly would not question your observations. What it means to me is that both behaviors occur, and that becomes the fact that begs explanation.
You mention following error on a lathe at 45k. Three significant differences in my scenario (and I perhaps that of the OP) steppers, three axis moving simultaneously, high speed. Which actually brings up an interesting point; In the past, I had all steppers and therefor had no point of reference outside that realm. However, with my current odd combination of steppers and a servo on the same machine, only the stepper, which was maxed out speed wise in the motor tuning, lost position. The servo was spot on, although it was nowhere near maxed.
I make the assumption that Mach sends only pulses and no positional information, therefor it must have sent all of the pulses to the servo in order for it to maintain its positioning. One interpretation of this would be that these curious errors are confined to steppers. I have observed the lost steps with three different models of stepper motors on two different computers. All were being driven by Gecko203V, so perhaps that might be a suspect.
An interesting twist on this topic is that the Smooth Stepper in invariably described as having a 'high quality' pulse train. The logical implication being that Mach and the PP do not produce same. Although I do not recall ever reading any hard evidence to that effect, experienced an knowledgeable people use the SS, so I have to assign credibility to the concept. Taken together, all of the above evidence might seem to siggest that the Gecko 203V is particularly sensitive to a 'dirty' pulse stream, yet the OP is using differnet hardware, which defeats that theory. It takes time to learn which sources are reliable and which are not. The lesson in all of this is that in a situation where an unresolved problem is manefest, all possible known causes deserve mention.
-
To my mind the driver test is there to weed out the really bad computers that have no chance of running Mach in their present state rather than saying it will definitely run Mach.
Hood
You may view it as a pass/fail, but it also 'grades the papers' so to speak, and provides what appears to be a benchmark useable as a guide to setting us Mach. The logical assumption, and the one that I myself made, is that a passing grade at XX khz is a green light to set up for that speed. This is where the problem lies. Another area where the 'sin of ommission' is committed by the Mach documentation. I discovered that the driver test is not neccessarily indicative of how fast Mach can run on a given computer only after much wasted time and frustration :'(
Info worth sharing, methinks.
-
. . . . I was monitoring the following error in the drives software . . . . .
I have been jealose of this ever since you first mentioned it some months ago :-[
Soon I will have my hands on two new servo drives that have this capability. The tuning software for the CNCdrives products reportedly can both graph and provide textural data for real-time monitoring of the following error. ;D Should be fun to play with as well as very useful.
FYI, Rutex tuning software has real time textrual following error, but the tuning software cannot run while Mach is active. Leadshine tuning software graphs following error (as well as a host of other parameters) in real time also. I do not know if their software can run concurrently with Mach.
-
The big benefit of the SS is the pulserate, especially whe you are talking 2000 plus line encoders which is common on the industrial orientated servos, heck the spindle motors I have on the mill and lathe have intelligent encoders which can be set to 2 million counts per rev, even the SS wouldnt be much use there ;D
The pulse from the parallel port will vary from computer to computer, some will be very nice others not so nice. I have scoped most of the computers ports I have ever used with Mach but the SS is in a class of its own even compared to the best PP pulse. Thats not to say the PP is no use, it usually does a very good job.
Hood
-
. . . . I was monitoring the following error in the drives software . . . . .
I have been jealose of this ever since you first mentioned it some months ago :-[
Soon I will have my hands on two new servo drives that have this capability. The tuning software for the CNCdrives products reportedly can both graph and provide textural data for real-time monitoring of the following error. ;D Should be fun to play with as well as very useful.
FYI, Rutex tuning software has real time textrual following error, but the tuning software cannot run while Mach is active. Leadshine tuning software graphs following error (as well as a host of other parameters) in real time also. I do not know if their software can run concurrently with Mach.
The thing i really like is not only can I see graphically in the 4 channel scope window but also text of the realtime errors and it will store the max error both plus and minus. Heres a couple of screenshots of the software, no drives connected so no info there ;D
Hood
-
Hey Guys... I hate to say this, but I fear we are straying from my original question.
I think we might be digressing from the original subject a bit...
I went to the shop today and I stripped Windows XP of all of its extranious bits, as outlined in a Mach-3 document I have.
I tried running my profile, and for abiout three runs, the X-Hand-wheel returned to its starting point and the Z-Handwheel fell four steps short. I thouht things were getting better. Actually, for what I am turning, I could live with jogging "Z" back to zero a few steps and correcting the Z-DRO., but on the fourth try, "Z" fell short by 25 steps... Not acceptable! I tried slowing the motor's max speed settings down from 250 to 200. I don't like the way G0 sounds at 200 mm/Min and the end position never got better.
So, perhaps trimming Windows down is necessary and a good thing, but it hasn't helped getting my tool post to return to its original position.
Then I tried my experimental g-code to run "Z" back & forth several times, then run X back and forth several times. The program moves "Z" from 0 to 128, then back to zero three times at full speed using G0. Then it runs "X" fore & aft from 0 to 128 and back to zero, three times . I tried setting motor speeds at 200, 250, 300, & 350. With each increase in speed the motors sound smoother and get faster and every end run. Every time the script finishes both hand-wheels end back up at zero... every time! Then, I noticed something strange...
As the saddle moves from zero to 128 in either axis, I hear a "glitch" in the motorthat's operating... like it just skipped a step or two. Then on the return from 128 at about two-thirds of the way back (same as 1/3, the way out) , I hear the glitch again. This occurs on both the X-Axis and the "Z"-Axis... wierd! I put a mark beside the rails on each axis. It's like clockwork., When the saddle reaches either mark, from either direction, there is this little "bzzt". The faster the motors are running, the harder the "glitch" is to hear, but nonetheless, it is there.
I wonder if both these motors are skipping pulses every time and the lost pulses in one direction compensate for the same number of lost pulses in the opposite direction. That way, in this situation with one motor running at a time can always return the hand-wheel to zero??
Tell me I'm nuts!
Anybody know of a simple kit to build a TTL pulse-counter with about six digits of display?
-
Problems missing steps can be from a variety of things. It can be related to tuning, the PC not putting out a smooth pulse stream, the drivers expecting a long(er) pulse width, etc. A slight 'tick' sound when the motors are running is a good indication of the pulse stream not being smooth enough. The only way to tell is with an o-scope. Mach's Driver Test will test the interrupt latency of the PC but it cannot test the actual PC's hardware's capability of pulsing a parallel port that fast. Even new motherboards and add-on parallel port cards can fail to work properly.
A hardware based pulse generator like the SmoothStepper moves all the timing critical stuff off the PC and onto hardware that can very precisely generate a smooth and stable pulse stream faster than any of us would probably have a need for. I recommend a SS with all the systems I sell and I have yet to have sold a SmoothStepper and have an unhappy customer.
When you get a chance to look at the pulse stream with a scope it will give you a good idea of what the PC is able to do. Also look at data sheet for your drivers and see what the minimum pulse width is that they expect to see.
-
Thanks Jeff,
This skipping steps issue is the only thing keeping us from doing production.
I will tell my employer and suggest he consider springing for a smooth stepper.
I have an inexpensive digital scope on order. I hope to use it to see if I can see irregulatities in my pulse stream...
I'm also tempted to buy a digtal up/down counter kit and see if I can count the pulses and compare with the number of steps required to move a given distance.
-
The scope is what you need so you can measure the pulse width regularity. Compare the pulse width with what your drives expect to see. Feel free to contact me (follow link in signature) if you have any other questions, or just post them here :)
-
The big benefit of the SS is the pulserate, especially whe you are talking 2000 plus line encoders which is common on the industrial orientated servos, heck the spindle motors I have on the mill and lathe have intelligent encoders which can be set to 2 million counts per rev, even the SS wouldnt be much use there ;D
The drives I am looking at all have pulse multipliers, so pulse rate not an issue unless you don't have enough resolution for the accuracy you need.
What is important for the hi-res encoders is the bandwidth of the drives on the encoder side. The drives I am looking into at the moment can read 1mhz from the encoder. That's 6,000RPM with a 2500line encoider . . . plenty plenty for what I need. The step/dir pulse input is far less than that, but the multiplier saves the day. One order of magnitude is plenty for what I'm doing.
My only hesitation on the SS is the lack of swap axis, . . . deal breaker for me. I've been talking to some people about a completely different approach to that issue. May need programming assist from some of the big-dogs on that, but we shall see.
-
Hey Guys... I hate to say this, but I fear we are straying from my original question.
I think we might be digressing from the original subject a bit...
Didn't realize that when you leave the room, eveybody has to remain seated quietly until you return :-X
You have a host of suggestions, but you haven't done anything but try the same stuff over again that you already did, so there is really not much new to talk about on your issue.
If I remeber it correctly, the smooth stepper is less than $200 or there about. Your salary is a cost to your employer, yes? It will probably cost your employer more for you to dink around with scopes and building pulse counters than to buy the solution the gurus here have suggested . . meanwhile zero production.
Here is the real question: is this a critical path issue or not? If it is, and I were your boss, I'd be asking why you haven't already had Jeff-Birt overnight you a Smooth stepper. If it's not critical, by all means build yourself some new toys and play.
-
TurnerTom,
Please excuse me Tom, as i am attempting to help you. Now take a breath...........................
After reviewing all of your all postings I sincerily and honestly don't know where you are coming from
and will have a hard time answering you questions appropriately, but will try.
I have grouped your question's and remarks together . So please bear with me.
1A. Have problems with instability (losing steps) and am curious if anybody thinks adding a Smooth
Stepper card might help me?
I would suggest you not spend a penny on anything until you review all the components of your
system in light of how they relate to each other both, mechanically and electricaly, from the pc to the
stepper.
Yes, the SS can help if you have a PC with bad pulse stream out of the PP and have checked the
PC per all that has been recommned and know that it is the problem.
You can't clean up a crap signal very easily if at all.
No, the SS won't help if there are basic power supply, BOB, Driver or wiring problems.
1B. - the stepper motors (particularly the Z-Axis) fail to return exactly to their starting points
- tried numerous settings,A max speed of 300 mm/min ( say 12 inches per min ) would be nice
to obtain.
- the "Z" fell short by 25 steps,I tried slowing the motor's max speed settings down from 250 to
200. I don't like the way G0 sounds at 200 mm/Min and the end position never got better.
- The problem seems to show up when curring curves and both motors are running simultaneously
at varying rates...
Do you really realize how manny things can be a culprit for the above?
- Poor Pulses from the PC
- configuration - inproper steps / unit, backlash setting, pulse width etc.
- Bad wiring of the steppers
- power supply voltage and amperage
- BOB board influences
- Improper driver settings ( what kind of drivers are you using? Don't just say ******************, ,
post specific info on them, please, and maybe folks will have a look, but don't expect them to
to do the walking and looking for the information )
The sound of the steppers can tell a lot. If it sounds like your steppers have a bad bearing, being hit
with a little mallet in a reapeatble way, can 't get any speed, there are marbles rolling around in them,
then the PC really needs to be looked at.
2..Is there anything wrong with my choice of the Probotix RF isolator?
I don't know. Why do you need one?
3..We might be able to use a laptop at the lathe, we have a fairly new laptop running Windows Vista.
Is it worth trying an SS on that machine?
I don't use don't use Vista. There is a thread on using Vista, so you may want to read throught it and
see what they all have to say.
Some laptops work fine, some not so. There is a thread about using Laptops.
I treid the SS on three different laptops and all was fine with it.
4. If there is some glaring error in my present setup
I will assume your motor steps / unit are based on calculated values and you tested to see where they
started skipping and backed off from that point to leave some headroom. Even if it's 2 inches / min, but
here we circle back to the top of all the questions.
5.The motors are 3.5 volt so I assume the drive is outputting something close to 3.5VDC
Look at the power supply as whole to suite your drive and stepper requirements. Having info on the
motors helps.
6.This is all sort of new to me.
Walk before you run......you can't learn all this at once.
Read some on your owne, read the manuals, go to the gecko site and read for understanding.
Don't jump to conclusions and focus on one problem at a time. Ask for understanding.
7.I fear we are straying from my original question.
I hope ths gets you back on track.
8. I wonder if both these motors are skipping pulses every time and the lost pulses in one direction
compensate for the same number of lost pulses in the opposite direction. That way, in this situation
with one motor running at a time can always return the hand-wheel to zero??
I think reply #6 addresses this.
9. Anybody know of a simple kit to build a TTL pulse-counter with about six digits of display?
I don't know of any.
10. This skipping steps issue is the only thing keeping us from doing production.
I will tell my employer and suggest he consider springing for a smooth stepper.
Hmm .....But what will you say if the boss asks why?
-
I'm truly sorry if I have offended anybody. It was not my intention.
I think the best thing for me to do now, is to wait until the scope I ordered arrives.
Is it fair to assume the place to scope the pulse train is across the outputs on my Probotix RF Isolator? That is what is going into the drives...
I think I also need to tell you all, that this lathe was built and outfitted by a private individual. It was operated by a DOS computer, and the interface electronics between the parallel port and the Slo-Syn drive was also scratch-built. The profiles were written using Microsoft BASIC and are (for me) a nightmare to edit. We are hoping to move forward here, to use industry-standard software. This interface is disconnected for the time being, but I can, and have, reconnected it and the lathe still works properly from the old DOS machine.
Because this can still be made to work, I feel reasonaby confident that there is nothing wrong with the drive, the stepper motors or the wiring (at least from the output of the drive)...
-
Can't speak for others but I am not offendable in this venue. I'm like Seargent Friday on Dragnet . "Just the facts, maam". Personalities are pretty much irrelevant on a CNC forum. You might have a bunch of unpopular people buried under your house . . but that's not a CNC topic.
This stuff is really not complicated. Think of it like the old Vaudeville joke where the patient says "Doc, it hurts when I do this", Doc answers: "then don't do that !"
So, what is an RF isolator? . . answer; something you probably don't need. If nobody here knows what it is . . . . likely you aren't going to need it.
So here's the plan . . very simple: You need to connect the parallel port to a Break Out Board with an everyday all-pins-connected straight thru parallel cable, preferably short and shielded. That protects your computer and provides nice convenient pins to connect to. You then connect the breakout board pins to the step and dir pins on your stepper drives. Light gage solid wire is ideal for this. Use nice bright insulation colors, otherwise it won't work right . . nobody knows why ;) Your stepper drives should already be correctly connected to the motors, so that's a gimme.
How do you sculpt an elephant? Start with a big rock and chip off everything that does not look like an elephant. Anything you have left over that is not mentioned in the above plan . . chip off.
This is similar to one of my breakout boards. Note the LEDs on every pin. No guesswork here. The lights tell you what the port is 'saying' and on which pin. http://www.cnc4pc.com/Store/osc/product_info.php?cPath=33&products_id=47 (http://www.cnc4pc.com/Store/osc/product_info.php?cPath=33&products_id=47)
You are going to need a BOB no matter what, so it is not going to be wasted $$$ regardless of what your problem turns out to be. And my guess is that the problem may go away when you get rid of the strange leftover junk from the DOS setup. You are just digging yourself a deeper hole dinking around with a scope. All it is going to tell you is to get rid of something that you need to get rid of anyway.
If all of the above fails, then just put the whole thing in a box and send it to Hood. He'll fix it for ya :P
-
OK Simpson36, let's try this...
I will order the C1 - Parallel Port Interface Card as soon as I hear from you about the following relay board
I have a relay board that I bought with the RF Isolator. It has screw-terminal inputs & outputs. Any reason you think I should replace it with your C15 - Dual Relay Board?
The DOS machine and its interface is disconnected for now. I don't plan to reconnect that unless I can't get the Mach-3 system working.
Thanks for your advice...
-
So, what is an RF isolator? . . answer; something you probably don't need. If nobody here knows what it is . . . . likely you aren't going to need it.
That is kind of an odd attitude. Ignorance on your or other folks part should never be a determining factor. ::)
Tom - The RF isolation BOB you have looks interesting. I went and looked up the types of isolation IC's it uses and they seem like it would work well, it is just a newer style of isolation. It accomplishes the same things as an optically isolated BOB but used Radio Frequencies instead of light.
-
Jeff,
I'm the first person to admit, I am ignorant.
I'm an organbuilder, not a machinist, and I have no experience with CNC. I do have some programming skills and I have an X/Y plotter. The plotter was built by the guy who did the lathe. I have an older laptop that uses Windows-98SE and it has a parallel port. I can run the original BASIC programs in a DOS wiindow, and the plotter will trace the actual toolpaths on paper. The plotter is fitted with a digital counter that counts up or down. revealing the X/Y position in terms of the number of pulses sent.
I inserted break points at the end of each line or arc and ran the BASIC profiles. From that I wrote all the values displayed on the X & Y counters and put them into an Excel spread sheet. I was able to extract radiuses for the curves from the BASIC source. I then worked these spreadsheets to generate simple G-Gode scripts, which I saved as text files.
I also have my main PC... just loaded with software running XP. It too, still has a parallel port.
I added a second parallel input to the plotter and wired the output end to the appropriate points in the builder's interface with Mach3 requirements. The Point is, I can plug my laptop into the plotter and run plots of the profiles on paper... unplug the laptop and run my g-code profiles using the alternate socket on the plotter from my main PC through Mach-3, and trace right directly over the original BASIC plots, using my Win-XXP Desktop. It took some tweaking of the g-cod, but it works!
Based on this, I was convinced that I could convert the lathe to use Mach-3. I still believe this is possible, but I'm totally at a loss as to what to try next. I have to try something. I'm paying out of pocket for my hardware mistakes. It isn't fair to ask my employer to do this...
So yes, I am ignorant.
We turn drawknobs for our consoles (see attachment)...
-
Tom, first there is nothing wrong with being ignorant about something, we all started out at the same place in CNC. I was also not trying to be offensive to anyone. My comment about ignorance was directed to another persons comment (the quoted one) saying that if he or others didn't know what an 'RF isolator' was then you probably didn't need it. I was trying to say that the ignorance of others should not be a determining factor in your decisions. Get the facts, as you are trying to do, and make the best decisions you can.
-
Tom,
What are the steps / unit that your are using for the pc that you will ultimately use for MACH?
Can you show us how they were calculated?
I would suggest that you first confirm the signal quality generated from that PC ( at the PP outlet / pins ).
If you can have a knowledgable friend take a look at the pulse with the right equipment that would be benifical.
Then look at the signal with the cable hooked up. Then look at the signal with the BOB hooked up.
So you are just checking the quality of the signal from the source and on down the line until it feeds the drive.
I would also suggest that you start off with the Z axis, as that is the one that seems to be off.
RICH
-
To my mind the driver test is there to weed out the really bad computers that have no chance of running Mach in their present state rather than saying it will definitely run Mach.
Hood
You may view it as a pass/fail, but it also 'grades the papers' so to speak, and provides what appears to be a benchmark useable as a guide to setting us Mach. The logical assumption, and the one that I myself made, is that a passing grade at XX khz is a green light to set up for that speed. This is where the problem lies. Another area where the 'sin of ommission' is committed by the Mach documentation. I discovered that the driver test is not neccessarily indicative of how fast Mach can run on a given computer only after much wasted time and frustration :'(
Info worth sharing, methinks.
bit late to this game and very left of centre... but.. i did an install of a Minitech Mill with Mach3. SAme mill as mine, same software, and guess what... complete balls up. the drives were hooking and jumping or stalling.
I went through the usual suspects.. driver test said ok. cabling was good, no other spurious signals coming from anywhere as far as i could tell.. I was at my wits end.
I happened to do another drivertest and it was fine.. left it running while i was lloking for a way to tell this customer i thought the controller was a duck. drivertest suddenly went to fail then jumped back. noticed it happened as customer moved the mouse. it would work fine until the mouse was moved then would go to fail mode. completely bizarre. further testing revealed that ANY other activity on the pc would cause this behaviour even at 24K rung up my mate in the US and his first question was "is it an AMD?", it was, then next question was "is it a newer one, multicore?" yes again.
he suggested taking windows over to standard install from acpi, and lo and behold. worked perfect after that, even at 100K kernel.
soo..... never overlook even the most bizarre bit of hardware that can be interfering with mach3 signal. subsequent reinstalls onto different machines at my place i have follwed the mach optimization rules document. never a headache since. install - plug - cut.
simple.
-
Tom,
What are the steps / unit that your are using for the pc that you will ultimately use for MACH?
Can you show us how they were calculated?
OK...
The lead screws are metric... 4mm/turn
The motors are 200 pulses/revolution.
The drives are set for full-step.
Therefore, 200 pulses per revolution = 4mm, and 1mm=50 pulses
I'm using metric measure in Mach-3 and the Diamener Mode.
The diameters I am turning match exaclty with the G-Code, once I have accurately set the "Zero" point for the X-Axis.
I'm using Sherline 1/2 pulse mode.
I will post screen-shots of my motor setups
-
I run 40,000 steps per unit 4 axis at a time for 8 hrs at a time from the LPT port and no lost steps.
Just throwing more hardware at the problem rarely fixes anything. YOU have to find the problem as to WHY it is loosing steps.
Most lost steps problems can be related to misengineered machines. motors/drives/powersupplies TOO small for the stated application.
Next on the list is bad connections
Then on to Bad BOBS(slow optos on super fast pulse rates)
IF this is a commercial machine making money do yourself a favor and bring in a machine TECH that understands the process.
COmmon myths tell you stepper are not as good as servos. BULL, if properly engineered steppers can be MORE accurate and last longer than servos in most low speed applications. AND are much easier to setup and maintain. GOOD steppers are almost indestructable.
(;-) TP
-
So, what is an RF isolator? . . answer; something you probably don't need. If nobody here knows what it is . . . . likely you aren't going to need it.
That is kind of an odd attitude. Ignorance on your or other folks part should never be a determining factor. ::)
Tom - The RF isolation BOB you have looks interesting. I went and looked up the types of isolation IC's it uses and they seem like it would work well, it is just a newer style of isolation. It accomplishes the same things as an optically isolated BOB but used Radio Frequencies instead of light.
You have it backward. Ignorance is recommending something that you do not know works simply because you read a spec and *think* it will work.
SMART, is trusting the knowledgeable people on this forum who have may years of MACH specific experience and *know* what works.
Beating one's head against a walll for days on end with no result is usually the first step toward a CNC troubleshooting, but it would appear the OP has already accomplished that task and perhaps is ready to move on to step two; . . . making it actually work.
Lastly, the fellow with the problem has stated that this is for a business and not a hobby, and there is an objective here to diagnose a NON working system and get the thing going. Wasting time dinking around with oddball unknown components from an old system that are NOT working with MACH for the sake of saving a company $70 is just foolish . . . in my humble opinion.
Your mileage may vary . . . .
-
I will order the C1 - Parallel Port Interface Card as soon as I hear from you about the following relay board
I have a relay board that I bought with the RF Isolator. It has screw-terminal inputs & outputs. Any reason you think I should replace it with your C15 - Dual Relay Board?
If the relay is operating something now and the relay can be triggered with 5V, you should ne OK . . . . . unless it is RF isolated . . LOL! Be careful though that the relay doesn't draw more juice than the BOB can provide, or you can hurt the BOB. The chips are socketed, so it's easy to replace them (don't ask). Specs on the BOB are available on the site, you'd need to look up the specs on your relays.
You have some distance to go yet with this project and you are already getting plenty of opinions on this what-not and that doo-dad, so I am going to give you my best advice on the project menagement level. You stated that this is not a hobby project, so that changes the game. In my view, your decisions should boil down to how much your time is worth, and any losses associated with the machine not producing anything.These CNC parts are dirt cheap in comparison to what I might guess your wages are. You *might* be able to reuse an old piece and save $18, but if it takes you three hours and you need to buy or buld test equipment to figure that out, you have not saved anything for your company money AND, you still have an old part that nobody can give you advice on and now you also have a 'production' machine of unknown reliability. Would the relays shut down production if they fail?
As a matter of personal preference, I avoid 'multi function' stuff if possible. Problem is when one component goes bad, you need to replace the whole thing if you don't have room to add another board or if the failed feature can't be disabled, or if it took other stuff with it.
-
I happened to do another drivertest and it was fine.. left it running while i was lloking for a way to tell this customer i thought the controller was a duck. drivertest suddenly went to fail then jumped back. noticed it happened as customer moved the mouse. it would work fine until the mouse was moved then would go to fail mode. completely bizarre. further testing revealed that ANY other activity on the pc would cause this behaviour even at 24K rung up my mate in the US and his first question was "is it an AMD?", it was, then next question was "is it a newer one, multicore?" yes again.
he suggested taking windows over to standard install from acpi, and lo and behold. worked perfect after that, even at 100K kernel.
soo..... never overlook even the most bizarre bit of hardware that can be interfering with mach3 signal.
Great tip! Someone should collect all of these gems into a FAQ or tech tips document.
Just to add to the stew, my Jurassic dual P3 server died yesterday (finally) and I replaced it with my wife's old 1.8ghz Intel Pentium on a Gigabyte MB . . Windows XP home. Still really old machine, but plenty to run Mach.
Passes driver test at 60 and 65 with 'excellent' rating. Set Mach to 60, steppers get sbout 6" and lock up. Set to 45, all is well smooth as silk.
For $155 all of the PP issues can be sidestepped (yuck, yuck). Only hold up for me is swap axis. Once that is implimented . . or . . . I find another method to replace it, I'm going to go with the SS.
-
IF this is a commercial machine making money do yourself a favor and bring in a machine TECH that understands the process.
Exactly. While your solution is a bit on the extreme side, methinks there is not enough recognition that in a business environment, the priorities change and the costs are not the same as on a hobby level . . . possibly due to the large proportion of hobbyists on the forum . . . just a guess. Employees tend to think only in terms of purchases and don't consider thier wages to be a 'cost'.
COmmon myths tell you stepper are not as good as servos. BULL, if properly engineered steppers can be MORE accurate and last longer than servos in most low speed applications. AND are much easier to setup and maintain. GOOD steppers are almost indestructable.
A similar common myth is that a dump truck is not as good as a Ferrarri. What are these people thinking :P
OK seriously, how does one go about 'properly engineering a stepper'? . . and . . how does one identify a 'GOOD stepper'. Cost? Brand name? I'm not going to use any more steppers, but I think it would be valuable info to share.
-
I just skimmed through this, and it doesn't appear that anyone mentioned that resonance may be the culprit? Steppers are notorious for mid band resonance, especially when running in full steps. Does your drive have microstep settings, and have you tried them?
If not, a damper may help. It's helped a lot of people at CNC Zone double their speed and make their machines far more reliable.
http://www.cnczone.com/forums/showthread.php?p=256639#post256639
-
TurnerTom,
I noticed that you have the pulse width set to 5. Just curious if you have tried setting it at 2.
Also noticed someone posted a description of your 'RF isolator.' Seems like a bad strategy to me. What keeps the RF from one pin from bleeding over to receivers in other pins? That would certainly be one potential explanation as to why you only have problems only while two axis run simultaneously.
I have had a bad time with RF problems. For example, if I plug in a wireless network adapter (USB), the steppers immediately start stalling. This behavior is completely repeatable. Unplug adapter, problem goes away. Plug in adapter . . immediate stalling. I has similar problem with wireless Xbox controller.
Others will chime in and say that they have had no such problems with this or that wireless device.
This does not say that you will or will not have RF interference problems. What it does say is that the question of whether RF can interfere with the CNC operation is NOT a debatable issue. It is a fact.
Based on my own experience, I would not voluntarily introduce any RF devices unless there was no alternative.
-
I just skimmed through this, and it doesn't appear that anyone mentioned that resonance may be the culprit? Steppers are notorious for mid band resonance, especially when running in full steps. Does your drive have microstep settings, and have you tried them?
If not, a damper may help. It's helped a lot of people at CNC Zone double their speed and make their machines far more reliable.
http://www.cnczone.com/forums/showthread.php?p=256639#post256639
I've just discovered something to throw into this arena. I replaced a Gecko 203V with a used Centent 143 'anti resonance' stepper drive. This is an expensive commercial drive about three times the size of the Gecko, but available used for $45 to $100. There is also a 142 and a 146 model with sightly different specs.
Let me say that I love the Gecko203V for it's 'unkillableness' which saved my bacon more than once, and it does an excellent job with small(ish) steppers. But performance wise, the Centent drive is in a whole differnet category. There is NO resonance at ANY speed running a 900 oz-in NEMA 34 stepper parallel at 22x voltage. Although it never lost steps, this was my 'problem child' stepper on the Z axis from the standpoint of resonance. I've heard it said more that a few times that the commercial drives (stepper and servo) are in a different class altogether and that is certainly the case with this product.
-
My boss thinks he can borrow a scope from a friend of his and if so, it will be in the shop Monday or Tuesday.
If the scope is unavailable, I will all have to wait until the one I purchased for $55.00 arrives (assuming it works)
When this lathe ran under DOS (and the BASIC programs were outputting the pulse train directly at the correct speed), the stepper motors followed faithfully. If the conputer had to be replaced, then the timing loops that controlled the rate of data flow, had to be adjusted to deal with the new processor speed. This has always been sort of a nusiance, but it did work.
As soon as I can get some wave forms, I will get the display synched and try to shoot some images using slow shutter speed.
I'll submit the images here, if they don't make sense to me.
From what all I have read so far, this seems to me to be my best next step...
-
Simpson,
" if properly engineered steppers "
" How does one go about 'properly engineering a stepper? "
I believe what VMAX refers to, is the proper selection of a component to a system and not the actual design of a just the one component. Systems are "designed" at the professionsl level. True engineering of a system is beyond the understanding of most hobbiest's. In fact, the total engineering of a system requires the application of multiple engineering talents. In light of the above, one does not know how lacking they are until entering anothers discipline, or as Clint would say " a man needs to know his limitations".
With a defined basis of design, any level of engineering deisgn can be applied to support a selection. Good engineering judgment is always part of application of a component.
Discussions are great, but when entangled in replies to a thread, it goes off course real quick,
Rich
-
I believe Mariss from Gecko is the designer of the Centent drives. It would be interesting to hear what he thinks about your findings, as I was under the impression that Geckos are far more advanced than the Centent.
-
Tom,
Sounds good to me. When confident the signal is good you can then continue
to develop applicaton of Mach to your project. LOL
RICH
BTW: Suggest next time you start a topic, post in the context of what your after.
There is a difference between a software problem, an application of the software,
clarification for understanding, etc.
-
I believe Mariss from Gecko is the designer of the Centent drives. It would be interesting to hear what he thinks about your findings, as I was under the impression that Geckos are far more advanced than the Centent.
Are Mariss and Marcus the same person? I had email exchanges with Marcuss Freimanis (may have spelled that wrong), but I ocationally see 'Mariss' mentioned.
If the Gecko is far more advanced, it would be quite a bargain at half the price of the Centent. I found no discrenable difference between the two drives with small (NEMA23 400oz-in) motors.
-
I believe what VMAX refers to, is the proper selection of a component . . . .
That is my question. How does one go about selecting, identifing, determining a 'well engineered' component. A statment was made, I'm just asking it be expanded to a useful point. Exactly how does one go about matching these components together as you correctly point out must be accomplished. I see a recurring contradiction whereing stepper systems are consistently touted as being 'easy', 'uncomplicated', etc, yet there is a never ending stream of posts with 'missing steps' or something similar as part of the subject line and then a pile of shot-in-the-dark suggestions on how to diagnose the problem. Not unlike this very thread.
Therefor, the logical conclusion is that this is not so simple as it's made out to be.
It's a serious question. I've done a ton of research on these topics and I don't reacall finding any clear advice on exactly how to select a 'quality' component from the vast choices available in the world of steppers. i.e. is it all down to getign a consensus or recommendation from existing users, or is there some way to be proactive in interpreting a spec, graph, or other perfomance documentation?
I.e., starting with a clean sheet of paper, what are the caveats to avoid and what is a logical path to sucess in setting up a stepper system. I'm asking for the benefit of the OP and others, as I am not planning to use steppers again.
-
Brothers I believe.
Hood
-
YOUare right IF I just wanted to drive around town showing off I would buy the sports car those speeds shure are impressive.
BUT IF I wanted to actually do some work and actually MOVE the dirt I would BUY the dumptruck every time.
with 40 years of driving dumptrucks there comes experience(;-)
TUNING ??? (;-) THe dump truck requires you to ? do NOTHING no tuning required turn the key and fire it up
, THE sports car YIPES what is a ocilliscope anyway, PID what does that do??? Have to tune it with the machine in motion? why? WHy is it I turn one knob and it effects the other settings too HUH?
I have done both many times (;-)
(;-) TP
-
OK this says a lot:
When this lathe ran under DOS (and the BASIC programs were outputting the pulse train directly at the correct speed), the stepper motors followed faithfully. If the conputer had to be replaced, then the timing loops that controlled the rate of data flow, had to be adjusted to deal with the new processor speed. This has always been sort of a nusiance, but it did work.
Can you tell us WHAT has changed in the lathe setup. Same drives and motors? DID the DOS version run from the lpt port or a motion card that interfaced it to the drives? DID it run from a SERIAL port??
SOunds like you may just need to find the proper pulse width of the signal that makes the drives happy. UNLESS the BOB is distorting the signal?? OR the drive requires a really ODD signal to begin with.
IF you supect the BOB just as a test wire it straight to the port, eliminate the BOB to test. THen you will know for sure.
Obiviously this machine ran from a PC before . Now it is a matter of elimination to find the source of the problem.
Just a thought(;-) TP
-
OK,
I have a db-25, screw terminal board... no LEDs, no isolators. I could easily transfer the four stepper lines coming from the drive to this board. I can do this Monday. My parallel cable can then plug into this terminal board I just thought using a simple terminal board without any isolation was an absolute "no-no".
-
I just looked at the manual for these drives. THe inputs are opti isolated so NO further isolation is needed you can safely hook them direct to the port.
The MIN pulse width is 25 micro sec and the max pulse rate is 20Khz so that is good to go.
Depending on the exact drive (F or M) it will be either full/half step OR Micro step. ( VERIFY the switch settings)
ALSO it can take either step/dir OR CW/CCW signals Really need to verify this setting to be sure. ( CHeck the manual verses the switch settings)
As to the RELAYS I would simply get an opti isolator for those inputs OR double check the manual for the boards it may already have it(;-) NORMALLY you would NOT double up on isolation as the signal CAN degrade for each LOOP thought the isolator media.
THat should get you on your way to finding the problem(;-) GOOD luck and check back if you need help, We are always here(;-)
(;-) TP
-
OK,
The drive almost HAS to be the "F" Drive. It is the only drive which offers full-step. The lowest resolution for the "M" drive is half-step.
This brings up another question... Is there any advantage in setting this drive to half-step? I would think all I need to do with Mach-3, would then be to change my "Steps/mm" from 50 to 100. No other changes should be necessary. Hell, it might even eliminate my missing steps? It certainly couldn't hurt...
What thinkts thou?
-
Are Mariss and Marcus the same person? I had email exchanges with Marcuss Freimanis (may have spelled that wrong), but I ocationally see 'Mariss' mentioned.
Marcus is Mariss' son.
-
COULD BE (;-) I would set to the highest resolution for the drive. Then adjust MACH as you stated. FUll step mode is KNOWN for midband resonance in SOME applications(not all).
I woulds ALSO eliminate the double iso (;-)
(;-) TP
-
I'm, sorry, but I don't know what you mean... "double iso" (duh)
-
I'm, sorry, but I don't know what you mean... "double iso" (duh)
double isolation
-
Simpson,
1. The logical conclusion is that this is not so simple as it's made out to be.
Depends on where you are comming from.
If you start with a clean sheet of paper, then you define what you want the system to do
and go through all the itinerary steps to arrive at an acceptable design within all the
constraints that are placed on doing the task. This is done in a way which utilizes good
engineering principles, techiniques, judgement and includes past / present / researched experience.
The task utilizes all the different disciplines involved in a respected and valued manner. In the end a
rationalized design is provided in terms of cost, quality and time (CQT).
CQT is related and requires sacrifice to improve the other.
2. I don't reacall finding any clear advice on exactly how to select a 'quality' component.
The suppliers of material components cover the gamut of required items. There is a lot of choice
to be had and the evaluation of all of them would be mind boggling. User experience of tried items
is invaluable and manny recomendations / suggestions satisfy a general need in the hobbiest arena.Consensus may never be had. We are here to offer suggestions for consideration and not make decisions for someone. In the end is falls back to oneself to make a decision.
3. Is there some way to be proactive in interpreting a spec, graph, or other perfomance
documentation?
Yes, and when given, most folks don't full understand it, or, will shy away because it's too technical. For even a simple item a book can be written and probably already exists.
All too often data and application must be related to the system and intended use.
4.What are the caveats to avoid?
Putting it all together and then asking others to fix it for you and challenging their experience.
RICH
-
YOU are right IF I just wanted to drive around town showing off I would buy the sports car those speeds shure are impressive.
BUT IF I wanted to actually do some work and actually MOVE the dirt I would BUY the dumptruck every time.
Exactly. This is why the whole stepper vs servo argument is meaningless if undertaken out of context.
Certainly the debate is valid in applications where there is overlap . . . dump truck and Firarri will both take you and a friend to the local Blockbuster Video Store to rent a movie. Speed is identical while sitting in a traffic jam . . etc. i.e. applications wherein the advatages of either are not required nor are the disadvantages of either going to hinder the machine's function.
But as you correctly point out, the objective, winning a race or hauling dirt, should be the first item of discussion before any recommendations start being passing around. Unfortunately, more often that not, the fur starts flying before anyone knows what the application is. This was pretty confusing to me when I first entered the fray.
Incidentally, is 'VMAX' related to the motorcycle?
-
I'm, sorry, but I don't know what you mean... "double iso" (duh)
You should protect your PC from a short that could send motor voltage to fry the PC port or worse. Since your drives are already isolated, you do not need additional isolation between them and the PC. For example, Gecko drives are isolated, so the BOB I linked you to comes in a less expensive 'Gecko version', that has no isolation on the pins typically used to connect the drives.
My money is on your RF gizmo being the prime suspect, so since you do not need isolation for the drives and you have a no-frills BOB to try our, my vote is to go for it. Worth a try.
Full step vs half step vs micro-step *should* have no bearing of whether the setup functions correctly or not, although there are reports that checking off the 'sherline half step mode' box in Mach does solve certain mysterious lost step problems . . . . primarily on the Z axis for some reason . . . .even though it shouldn't have any effect.
Mostly I am interested in how the little scope works for you. When I had the Gecko servo drive, I was sniffing around available scopes to use in tuning and I selected this one: http://www.saelig.com/PSPCEL/PSPC019.htm (http://www.saelig.com/PSPCEL/PSPC019.htm) but never bought it as I got rid of the Gecko instaed.
-
I'm, sorry, but I don't know what you mean... "double iso" (duh)
double isolation
A common CNC term that describes what happens to husbands who spend too much on their toys ;D
-
I'll be sure to buy my wife a gift today!
-
I got it working!!!
All I did was to change my Slo-Syn spepper controller to half-Step... go into Mach-3 and under motor tuning change the "Units per" from 50 to 100.
I even took the max motor speeds to to 400 (There are some G0s in the code, so max speed IS executed)
I ran my most problematic profile over and over and EVERY TIME both hand-wheels returned to zero as expected.
I even turned the graphic display for the toolpath ON, and the steppers never skipped a step.
With twice as many pulses, each stepping half the distance, the motors sound better than they ever have. They just "purr".
I bet my curves are cutting at higher resolution. Less sanding afterwords... NICE!
My sincere thanks for everybody's help. You guys are the best!
-
GOOD JOB TOM, We expect to see some pictures soon of the machine and the parts you make,
(;-)TP
-
Howdy Tom, I have been fighting an issue very similar to yours.... for weeks now. discovered my "fix" tonight. I am using the keling controllers...but they say it requires the sherline half step mode (ie 40us pulses). I found a thread on here where someone finally explained how that actually works. You only get the 40us pulses at 25khz. The pulses are shorter as you select a higher kernal speed. I am cutting a 3d model now at 34k lines...looks perfect. Give that a try.