Ok, how many switches are there, what polarity (npn or pnp) and voltage? How are they currently wired?
Printable View
Ok, how many switches are there, what polarity (npn or pnp) and voltage? How are they currently wired?
There are 6 in total 3 home 3 limit all pnp nc each one has a common + and - up to a junction box on the gantry were they go off on there own each have there own separate input wire on the csmio . the main wire is a 12 core cy screened and earthed at the control box all running on 24v
many thanks
Just for completeness (I am not familiar with the csmio) as this thread is long is the 24V ps -ve connected to the -ve on the controller board.
I would be very surprised if it turns out to be the switches although I favour npn type.
As has been said I don't think the cable length should be an issue.
Have you tried a small capacitor at the machine end of the ps supply cable across the +ve and -ve.
I though Charlie had been advised which switches to use, but anyway for inductive proximity switches my advice would be 3 wire NPN, this is a useful thread Charlie; http://www.mycncuk.com/threads/7723-...0012#post60012
Unlike most Bob's the Csmio I/O's are not restricted on switch type or a set voltage/current so can use any Logic so PNP or NPN doesn't matter. NPN or PNP both do the same job they switch a circuit, just one sinks and one sources and the Csmio inputs accept either.Quote:
Originally Posted by EddyCurrent
For instance input one could use PNP @ 6V and Input two NPN @ 24v provided they both have there own voltage source.
This one reason why it's used so often on retro fits as it can use Odd ball logic and voltages of Bespoke controllers etc.!
So active low signal (i.e. break = limit/home hit), wired as per page 19 of the CSMIO/IP-M user manual?Quote:
Originally Posted by charlieuk
.
I have to say I doubt line capacitance is an issue at the CSMIO end. CY cable is 75-150pF/m so looking at 900 - 1800pF for 12m. So when switch opens that 1800pF has to discharge through opto-isolator/resistor. Assuming standard 10mA opto current, series resistor is 2200ohm, so fall time is approx 6uS worst case. One simple & quick test would be to put an additional 2200ohm 1/2W resistor to ground at the switch end of the cable (or even a couple in parallel). This will discharge the capacitance faster and will reduce the effective capacitive load on the switch. Try that first...
Watching the video more closely, when the plugin filter is reduced to 1 (video at about 2:49min in.) the switch works as expected but the auto reset won't work. So it looks to me like it's the input filter setting that is causing the problem with switch not triggering. The issue is then, why does it not reset when the limit override is selected.
From the manual.
"5. If the machine has entered the hardware LIMIT switch, you can leave it by switching on the Settings screen "OveRride Limits" button. It is also convenience to switch "Auto LimitOverRide" - this will cause that while the raid on SW limit switch the machine will stop , but you can click RESET and le ave the limit switches without any additional operations."
Is it saying that "Auto limitoverride" is for software limits (SW) ? and "Override Limits" button is for hardware limits ?
Charlie, set the input filter to 1 or zero and see what this "Override Limits" button is.
Whew you set it to low when you hit the limit and then hit ether of the over ride buttons it triggers another e stop straight away so it it making it to sensitive. Like I said though when you shorten the wire it's fine and but I'm sure at one point I did have it going as is. We have some new switches to try so will be giveing that a go later
Ok so we tried the new switch with not much improvement so tony put his oscilloscope on it and basically the large kinco driver for the x axis is putting out a huge amount of noise.
The input filter in the plugin is used to stop spurious signals induced in the limits wiring acting as though the limit has been triggered when it has not.
If you are not getting these false triggers then the value in the filter setting needs to be as low as possible with zero being the best. Any higher setting is going to ignore the limit signal for a brief period of time and all that is demonstrated in your video.
It may be that noise is a problem once the limit has triggered but the main issue seems to be with resetting the system using some kind of override for the limits, I can't be more specific because I don't have access to a csmio board or plugin.
I do know that if you set up the software limits they should stop the machine before reaching the hardware limits and you should use those all the time, but obviously the hardware limits MUST be working for safety.
Are you using the most recent plugin software ?
Just a quick update. We have managed to sort the problem which was noise being created by the x axis drive and the long cable. We first re did some of the earthing on various bits and put them in a nice brass terminal block which made quite a difference. We then swapped the cheap Chinese proximity switches for some made by Schneider which cost as much for one as 10 of the cheap Chinese ones (should have know better). The last step was adding a small resistor on the csmio end and it now all seams happy.we also got the PC water pump running and even on the long pipes it still pushes a nice flow of water. We have also fitted a relay so that it switches on with the spindle and dust extractor. Not quite to sure what is left bar finishing trying to set it up on mach3 and learning how to use mach3 and load g code ect. Again if there is anyone anywhere near Brighton that can help me learn how to use mach 3 that would be a massive help.
Sounds like a combination of things, pleased to hear it's now working.
We'll after a bit of trial and a lot of error I started to cut my first road runner. However the power socket trip out shortly after as I think I had to many things plugged in like a oil heater ect.
I now have two question, not that it worries me for this but would would you do if that happened on a big job?
And second the feed rate is really slow for cutting road runner in foam , I have increased it to max feed rate via the over ride on the main screen but it is still very slow so wondering were I increase it further? In the g code?
The problem is getting back to exactly where you were with regard to X,Y,Z zero positions. This is where Home switches come in, at least you can zero the machine coordinates with these, regarding work coordinates, when I've jogged the machine to where I want work zero X,Y,Z, I make a note of the corresponding machine coordinates from the Mach3 screen so if juice fails I can return to the same place. Maybe there's a better way but that's what I do because it's easy and understandable. You can then run the G code from the start or a line further on.Quote:
Originally Posted by charlieuk
Speed for rapids is set in Mach3 motor tunings, then feed speed is set with the F command in the G code, e.g. F800. You can override it while the G code is running using the slider on Mach3 screen.Quote:
Originally Posted by charlieuk
Cool will try adjusting it in the g code many thanks!
Remember you might loose steps if you go overboard with it.
Yes there is and it's really simple to do. Infact it's something you should get in the habbit of doing if long or important jobs just incase power or PC crashes.Quote:
Originally Posted by EddyCurrent
Save the work offset. On the Standard Mach 1024 screen set you'll find it OFFSET's Tab on lower right under SAVE WORK OFFSETS.
On Gerry's screen you'll fing it on the small Tab offset's Tab again Save OFFSETS.
If you notice when you close Mach you get the message box "Fixtures Changed" and "Save Fixture" option this is the same as Saving work offsets.
See pics.
Attachment 13811Attachment 13812
Charlie the Road runner file uses Imperial measurements and if using metric units then while Axis scaleing will increase the size it doesn't change the feed rates so you'll need to do this manually by editing G-code.
Cheers thanks for that I will try and figure that all out and get in the habit
ps dean I dropped you a pm
many thanks
Still playing here and trying to work out mach 3. I'm not sure if the problems I am coming across are me or something not working right. I have watched all of the tutorials multiple times and trying to read bits of the manual but I suffer from bad dyslexia so reading is painfully hard and slow.
The first thing that doesn't seam to have any effect is the offline button, with it ether flashing or not the machine still responds to all commands as far as I can tell. Is there something else that has to be done to turn that on not that it's that important.
the second is if I run the road runner program the hit stop button what is the process to start back up again? the reason I ask is if I hit stop then jog a axis to the side to take a look then hit start it sometimes does not return to the last point were I stopped the cut and it starts else were ? it also does not re start the spindle however I assume that is normal and it has to be started manually?
I allso have have a problem with loosing power to the x and y axis , I'm still trying to figure if it's in mach 3 that's stoping it or if it's a fault on the drive which I am waiting to see next time it happens.
Many thanks
I have been testing again and I'm wondering now if it's a setting to do with the arc when it moves between points however I still dong get why it's not starting back were it left off? The first pick I pressed stop jogged to the side and then hit go again. You can see the straight line were I did this and it did not return( centre of pic on his wing)The second pic I hit stop the pressed got to zero and then start and it did a similar thing?(going along his wing) Any ideas is it me or the machine?
Ps if there is anyone willing to give me a mach 3 lesson I would be happy to travel a way for it and pay if needs.
many thanks
The way I see it (and I am not an expert) is if you stop the machine and it was running on a curve it might not return to the same place. I would think you have to re start from a line number where it had been running a bit of G1 code .
I would try and experiment with some simple shapes like triangles and squares not circles.
and then try a feed hold and resume and then a stop then start ( I am not sure if pressing stop will lose steps)
If you have hit the E-stop button you would have to re home the machine ..Clive
Many thanks will keep on playing.
Any by idea why offline does not work? Does it normally ?
Cheers charlie
Charlie Offline is a (Mostly) parallel port driver related function and the Csmio plug-in doesn't use it.! . . . .To be honest I've never used it EVER so don't see the point and wouldn't worry about it.Quote:
Originally Posted by charlieuk
If you give me a ring then I'll walk you thru using Mach. To be honest Mach doesn't do much of anything other than move the machine and monitor inputs/outputs so when you have got it setup correctly and the basics covered you shouldn't have any trouble working it.
Think some of your trouble may be your not exactly setup right regards steps per and motor tuning. When this is correct then the rest will be easy with a little explanation.
I'm compleltely snowed under(Actually it's pissing down with rain not snow. . Lol) with work at minute otherwise would come down and spend some time with you.! Next time I'm down your way I'll let you know and call in for few hours.
Cheers I was wondering if that was the case again with it, it's just not knowing these things and being my first time ect.
I will give give you a call but I will just wait until this new PC is here I'm hoping today or Monday and it's set up on that.
Many thanks
Charlie just read your earlier post about not returning back to where it started from.? This could be the Step pulse edge is set on the wrong side in the drives so it drops a step with each direction change. This is easy checked and corrected.
To check just knock up some G-code for each axis that goes back and forth many times. Like Below.
G0 X75
x0
Copy and paste few dozen times into a file then Zero the Axis and make a mark to check if comes back to same spot then run the G-code.
Using Notepad then just do replace search and replace X with Y and repeat for each axis.
If it doesn't come back to the same spot then go into Ports and pins and toggle the Active Hi/low setting for the STEP to opposite state for which ever axis is wrong. This and the ENABLE setting are the only functions the ports n pins for motor outputs have on the Cslabs controller so ignore the rest.
EDIT: Other way to do it is thru the EM806 Drive software and changing the Active edge state but this is easier option.!
when i started with mach3 i also found this rather unusual , that the machine wouldnt return to the point where i stopped it
but when you start to read G code and start to understand what i means, then it all starts to make sense
for example , when cutting an arc ( as mentioned ) and you stop the machine lets say 1/4 of the way thru the arc , when you restart , it will not resume from that 1/4 point thru the arc ... it will either start at the beginning of the arc , or the next line of G-code
the same with height, z - axis ... if im doing a cut , stop mid way thru and then RAISE the machine ... when i restart , the machine doesnt know that it been raised , and will only change the height when a line of gcode tells it to , so you may do an air cut for a few lines of code depending on what you cutting ....
if i stop and want to continue , i rewind a couple lines of code ...
and yes , you need to start the spindle manually ( again g code tells spindle to start right at the beginning of the code )
keep at it and it all starts to make sense .... :)
If machine is moving and you hit STOP then you really should presume the machine as lost position and Re-home. If you want to pause the machine and resume then you should use FEEDHOLD.Quote:
Originally Posted by Blackrat
If machined is STOPPED and you want to resume from point in the G-code then use the RUN from HERE command. This will do a move back into position then pause with message asking for CYCLE START it will also give you option to start the spindle.
If using RUN from HERE then it's best done from the begining of Cycle or some position move in the code rather than half way thru canned drill cycle for instance.
RUN form HERE as saved my arse so many times I've lost count and use it all the time.
Cheers guys yes blackrat sounds like the same thing I will re read this and try and process it.Quote:
Originally Posted by JAZZCNC
are there any other learning tutorials that would help me learn this stuff? Cheers
Just cut my first board and im really happy with how it went. A few things need to be worked on like a vacuum system which I have found does not need much power atall to pick up the light weight foam and also a way to hold the foam down at the corners and also allow some sort of accurate registration so when you flip the block it lines up perfect to cut the other side
( im looking for ideas here) I am also wondering if the motor on the x axis is way over kill as it is spinning quite slow even when at full speed and gearing it more will just produce more torque than I would need.
gotta say that looks pretty good !
to locate the board ....
you have to have references cut into the foam block from the top side ... this is because the block is not accurate enough to flip over and carry on cutting
so instead of cutting the whole of the top of the block, id cut 6-8-10 (as many as you need to support the foam block) square patches into the block ... then when you flip the foam block over these machined patches locate onto the table by mean of MDF pieces that are the same size as the patch that was cut
hope it sorta makes sense :D
what software you designing the board in ?
That does make sence and is a good idea I have been thinking along similar lines but drilling some large say 1" dia holes in the table then using the cnc to drill some matching holes in the blank and using a large dowl to locate. Like you say it would be nice not to have to worry about the block being perfect. I guess I should try both and see
yup, thats exactly what im getting on at .... dowels in a blank that locate on the table sound good ... maybe machine round patches into the foam , and have the blank circular in shape
summing like that ?
Not sure I quite under stand ?
I was just thinking you need more surface area in the foam block .or the dowel will press into the foam .... So machining a bigger circular shape provides this larger surface area
I see what you mean I was thinking the dowels could be quite long and go almost all the way through the block, i think it would be stiff enuf then if I used large dowels. It doesn't need ultra high accuracy anything less than 1mm would be very nice. I'm kinda hoping the method I find i can do with out having to change cutters which is a huge 5"x 1 1/4 bullnose or spend too long having to machine things to set up. I think we are on the right lines though
I have been doing some more test cuts with the machine, yesterday I was doing a cut and a bolt on the z axis rail came a fraction out and cased a jam however I was pretty quick at stopping it and I don't think there was any damage.
However today I have had a error come up that I think I have had before in that the z axis is getting a Stall Detection error on it ie 5 red flashes on the driver ( it is a leadshine em806 hocked up to a nema23)
This does not happen straight away it has been maybe 40 min into a job and the z axis has totally stopped, luckily it has stoped at a point were it has not ruined anything yet however I cant find any mechanical problem that would have caused a stall and I can re set the driver and all is ok again but odiously that's not a cure.
so first I need to figure why it is triggering the stall detection? any ideas????
then second is there a way to hock up a signal to go to e stop if this happens for any other reason. ( I have yet to look into this)
so a little confused at the moment, could it be more noise?
many thanks charlie
I think this can happen if the acc is set too high you can set the driver to send out an estop so that the complete machine will stop if any driver goes into fault. I think the signal has to be set up in the software of the driver. I am sure Dean will pop up with the correct way to do it. ..Clive
Spoke to Charlie about other issue and why but you never mentioned the E-stop.!! . . . So Charlie it's easy enough to do. Tony will understand what to do.Quote:
Originally Posted by Clive S
The drives have fault output that you use to control a relay which is inline with the E-stop.
The fault signal Logic can be programmed in the drive so can be configured to suit your needs.
cheers many thanks for the advice dean just ordered a few more pulleys to gear it down and will strip it all down this eve and see if I can find what's wrong and were the binding issue is. sorry I forgot to mention the estop thing as I was more focused on fixing the problem. toney is down next weekend fot Christmas so will get him on to it as he has been after things to play with.
many thanks