Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: tmsmith on February 07, 2017, 06:01:25 PM
-
My pentium3 used to run mach3 using windows 2000. After battery problems due to a long period of no use it was repaired and was loaded with windows XP. This runs slowly compared with W2000 and I have not been able to drive my mill so far. It runs at 735Mhz. When my controller is switched on the spindle rotates at a reasonable speed but cannot be controlled. If I run driver test I get an error report 'Cannot find vectors' 'Pulsing too fast'. Pin settings do not seem to have any effect on the mill but do alter LEDs on the Diagnostics display.
Is this computer/XP not able to drive the mill or is it a setup problem?. Is a special parallel port requied.
T.M.Smith
-
Hi,
that this computer used to run Mach3 suggests that it should be capable of doing so again.
It sounds like its struggling to run XP smoothly and until it does so the running Mach3 is out of the question.
How much memory is installed?. I seem to recall having to add more RAM when I upgraded XP to one of the later service packs,
SP3 I think. You will probably need 2G to run XP SP3 smooth. Given the age of the machine finding compatible memory might
be an issue.
My XP machine became so unreliable I built a new computer based on a dual core Atom miniITX board. It has a built in PP and all
the usual connectors. Has SATA ports for HD, I used 64G mSata SSHD. You can buy a board like this for $75-$100 new. Add RAM and
an HD, reuse your power supply and monitor, probably for little more than fixing your old one.
Craig
-
Thanks for the reply joeaverage.
It is XP-service pack3 that is loaded but there is only 1gig of memory aboard. Just booting up is dreadfully slow compared with W2000 and I can only put this down to XP overloading the system.
There is a minimum of software on the machine.
I have tried Driver test again. Without Mach3 being loaded the system tested as OK. I then loaded Mach3 and ran again and this time it gave the error Pulse Rate too fast, vectors not found. The display then disappeared. Reran and it gave a satisfactory result. This with the mill disconnected. I am not aware what parts of the system are tested.
The computer is an HP Invent e-Vectra if that means anything to anybody. It seems it is not possible to get hold of W2000 anymore hence using XP.
Malcolm Smith
-
Hi Malcolm,
there was at a time a list of things that you could do to your PC which made it better for Mach3, haven't seen the list for a while but maybe
someone else will chime in.
I think I ended up with 2 Gig RAM with XP SP3. Have also seen comments on the forum that SP2 is adequate and avoids some of the CPU
loading which occurs with SP3.
Craig
-
Something is causing issues, and it's not XP.
I run Mach3 on a 1Ghz PIII with 256Mb of RAM.
With just Mach3 running, the PC only uses about 60Mb of RAM.
If I run driver test I get an error report 'Cannot find vectors' 'Pulsing too fast'.
You need to resolve this issue.
First thing is find the Mach3 optimization guide, and go through the steps one by one.
Actually, the first thing I would do is change the PC mode from ACPI to Standard. This may require re-installing Windows. I have a PC that would not run Mach3 in ACPI mode, but works fine in Standard mode.
Personally, I would consider 1Ghz to be the absolute minimum, based on my experience. While it works fine, there's very little margin for error.
-
Thanks for the replies.
I have found one guide suggesting changes to make using msconfig..
I have printed it out and will see what I can make of it.
Malcolm
-
Well I have made the changes suggested by the Mach3 Optimisation guide. I think XP boots a little faster and the Driver test says everything is OK.
But the drive motors are still not responding.
I think I need to check that the Parallel port is live.
I am using config. P&P setting I had written down when everything was OK using W2000.
Very frustrating.
Malcolm Smith
-
Hi,
what does the driver test actually say...I know that it gives a general 'OK' but there are a lot of details.
Can you post a screen shot?
Craig
-
Hi Craig,
It actually says system excellent, with the green bar at about 50%.
I have checked voltages on the P port. pin8, 0-3.2v with Z+/Z- using the jog keys.
pin 10 similar with Y+/Y- buttons.
pin 12 0-3.3v with X+/X-
Obviously this is the port at the back of the computer and is not the pins I had expected.
Not so easy to take a screenshot unless I use a camera; there is only the bare minimum of software on the machine.
One other thing that seems odd is the pulse rate shown on the diagnostic screen which varies from 1 to ~ 300
Malcolm
-
Hi,
don't know how your numbering the pins but that's wrong.
The output pins are 2-9,1,14,16 and 17. Further your machine should be producing 5V logic high. Might expect 3.3V with a modern machine
or laptop or late model parallel card.
Re-read the installation manual for Mach3. Pay close attention to the PP pin numbering and report back the results for the DIR pins for
X,Y,Z.
-
You have a 3.3V PC, and your electronics may require 5V signals.
It actually says system excellent, with the green bar at about 50%.
...............................................
One other thing that seems odd is the pulse rate shown on the diagnostic screen which varies from 1 to ~ 300
One big issue with the driver test is that it either fails, or says it's "excellent".
Varying by 0-300 is not excellent. Mine varies by 0 to about 4.
-
Yes the pin numbering is reversed when compared to Mach3 manual, so the actual numbers are 2,4,6 and they are measuring at 0/3.3 volts depending on which (+/-) key is pressed. No idea what voltage the pins were at under W2000, I will have to check with computer shop as to whether XP or their repair might be responsible. It certainly ran reliably under W2000. I have not determined which pin controls spindle speed.
Malcolm- who is going out for the evening.
-
Hi,
well your PC has 3.3V logic outputs, you will have to use a breakout board that can accommodate that.
Think we still need some details about the Driver Test, variation of 300 just doesn't sound right.
A screen shot or write it ALL down and post. If the driver test is not right then EVERYTHING you do subsequently is a waste of time.
Craig
-
Craig,
I cannot see how XP will change the pin outputs from 5v to 3.3v, is this possible. I have no idea what the pin voltages were under W2000 but the mill part of the electronics has not changed (unless it has blown a fuse somewhere). The spindle runs OK on manual. I have taken a photo of the Driver Test screen on my tablet but so far failed to get it onto the PC. The pulse rate variation I gave was taken from the Diagnostics screen not the driver test display.
Malcolm
-
I cannot see how XP will change the pin outputs from 5v to 3.3v, is this possible.
No, this is dictated by the motherboard of the PC. If it's the same PC, then it was 3.3V before.
Are you measuring the step, or direction pins? You can't really measure the step pins, as they are quick pulses. Measure the direction pins for an accurate voltage reading.
It sounds like Mach3 is working, so if your drives are not responding, it may be a wiring issue.
-
Today I checked with the computer shop who repaired my HP Invent and installed the XP system. I was told nothing they have done would alter the PP line voltages.
The battery had leaked but I was not told of any damage to the mother board.
Could of course be my setting up of port and pins, will check that again.
Malcolm
-
Well, After altering so many pin setting I am totally confused the z-axis will drive downwards with G0 Z-2 but will not drive back up??.
The X and Y axis do not respond. It now looks as though it is not the voltages on the parallel port that are the problem but probably the pin settings, and yes, I have read the manual.
Malcolm
-
More playing about with ports and pins and I now have X/Y/Z axis driving *one way* using the jogging keys and code in MPI screen.
Unfortunately the direction cannot be reversed using the same methods.
Can anyone tell me what is incorrectly set?
At least I have proved it is nothing to do with WindowsXP nor port voltages so it must be my setting up!
Malcolm
-
Hi,
firstly I'm not sure about the assertion that the voltages are correct. To drive a stepper in one direction the step pin be logic
low, for most BOBs that means 1v or less. To drive in the other direction requires a logic high, in your case 3V or so.
If your BOB is expecting 5V inputs then 3V may be insufficient to signal the intention to drive in that direction. This fits with
what you've described. Whether that is the case or not is impossible to determine unless you post some information.
Take some screen shots or at least write what settings you have applied and post those. I don't have ESP, unless you give me
some information to work with....
Craig
-
Craig, thanks for your interest, you are my only ray of hope at the moment.
Performance today gave same results as previously posted. Checked continuity on all pins of the P/Port cable and could find no fault.
Can you specify what data you would like to see and I will do my best to obtain it.
I am hoping to get a copy of windows 2000 but I am unsure whether that would provide a solution.
Are there special drivers that should be loaded for XP?
Malcolm
-
Hi,
don't go changing OS, I don't think that's the problem.
I need screenshots of your ports and pins pages, at the least motors,inputs and outputs tabs. Screenshot of general config page
would help.
Need a copy of the manual for your BOB and/or drivers.
Craig
-
Craig.
I will need a little time to pull this together, I built the outfit in 2007 so not exactly fresh in my memory.
I don't recall getting a manual the BOB/driver board which came ready assembled.
The only identification on the board or case is a 'made in the USA' sticker and 'version4' on the CB.
The unit was bought from a firm in Glasgow, Scotland. I will look to see if I can find any paperwork.
Meanwhile I have photp'd the board and will see if I can send a copy.
The spindle rotates but with no control but not concerned with that at the moment.
I will see what I can do re ports & pins.
Malcolm
-
Try again
-
and again
-
You can't use Port1 Pins 2, 4, 6 for Output Signals when they are already used for Motor Outputs. ;)
Tweakie.
-
Thanks Tweakie. Will try different settings tomorrow.
Malcolm
-
Tweakie,
First thing to say is I changed my CNC computer for a spare I have which was also running XP. No difference in control so I conclude the problem lies elsewhere.
Using your figures/instructions I set up the spindle pins which had not been my priority earlier. Obtained immediate control of the spindle speed and direction, great!!.
Noted that spindle speed does need setting up. Secondly no speed displayed in the DRO. Played about for s few minutes then Crash?? spindle stopped dead.
Opening up the covers revealed that the index plate had come loose and working up the shaft had jammed the shaft. Re-secured it but every thing dead, further examination
revealed a blown 5amp fuse. Replaced, and we had lift off with the spindle.
However neither X,Y, nor Z axis would move at all, not even in a singe direction as I had yesterday. Another blown fuse??. There is power to the driver control box, (red light ON).
Where from here??
Malcolm
-
Have you turned the Jog ON/OFF button to OFF ?
Tweakie.
-
I have not touched the jog button. Will check this tomorrow.
Thanks tweakie
-
Found time to look at the jog button and I try my jogging tests with it active.
That does not appear to be the problem or the solution.
I find I can jog in one direction with each axis, say X+,Y+ Z+ using the buttons on the MPG Mode 'fly out'.
The minus direction buttons produce no movement. These two states can be reversed by checking active low, but no how can I get any
axis to drive + and - using the buttons.
I feel sure it is a port and pins problem but what needs altering or should work with the X,Y,Z axis's.
Malcolm
-
Hi Malcolm,
For an axis to drive in either direction it requires the Direction signal to change state between High and Low.
As was mentioned in an earlier posting some CNC controllers require an LPT output with 5 volt TTL signals and they will not work reliably with 3.3 volt TTL signals.
Because they are cheap to buy, perhaps consider the purchase of a PCI Parallel Port card which has 5 volt TTL outputs (these may be purchased from a CNC supplier such as CNC4PC, etc.) and use that LPT port to drive your CNC controller).
If you attach your <profile>.xml file from your Mach3 folder (copy and rename it to Malcolm.xml befor attaching) I will check your Ports and Pins setup etc. and see if I can find any obvious problems.
Tweakie.
-
Tweakie, thanks for you sticking with me. I will post my xml file tomorrow when I go into the shop. I have changed the pin settings so often I cannot remember where I started.
Today I checked the P/port output pin voltages again while pressing the axis direction keys on the jogging 'fly-out'.
Pin 2 - Pressing X- gave 2.37v, X+ gave 3.32v
Pin 4 - " Y- " 3.07v Y+ " 3.32v
Pin 6 - " Z- " 3.06v Z+ " 3.32v
These readings reversed if motor output set to Dir low Inactive but the voltages remained the same.
Two things strike me; very little change in the voltages and how did it work before under W2000.
Could I have an electrical problem on the mother board. By the way at start-up it states the driver is initialised.
Malcolm
-
Hi Malcolm,
That does not look like good news :'(
I suspect that either your port an pin assignments are incorrect or the parallel port on your PC is defective.
Check that your X axis does in fact have port 1 pin 2 for the Direction signal and that you are measuring the voltage between pin 2 and pin 18-25. That being the case perhaps invest in a new parallel port card ?
Tweakie.
-
Yes tweakie I cannot make any sense of the voltages either and I am sure the pin allocation is correct. Incidentally pin 1 gave 4.1v, the highest reading on the port.
I think it is time to take the lid off and have a look. Oddly my spare XP machine gives similar readings; any chance of the port driver being corrupt.
I have now got a copy of W2000 and may try that on my spare M/c.
I am attaching my xml as requested.
Malcolm
-
Hi Malcolm,
Something strange here...
Can you attach the .xml file that you are currently using (<profile>.xml) - the one you have attached has no ports and pins configuration.
Tweakie.
-
I have just copied this file from the computer having first verified that it does drive the three
axis's in one direction only.
I have named the file Thomas.xml after posting problems
Malcolm
-
Hi Malcolm,
Assuming you are indeed using the parallel port to drive your CNC machine then you will need to configure the Ports and Pins in Motor Outputs and insure there are no pin number conflicts in Output Signals etc.
This is a screen shot from the last .xml file you posted.
Tweakie.
-
Well Tweakie I cannot understand this.
If I look at ports and pins on the computer that is not what I see. The entries under step and direction are 3,5,7 and 2,4,6 and pin 17 for the spindle???
Malcolm
-
Hi Malcolm,
You need to post the .xml file that you are using. It is in your Mach3 folder and it is called <your profile>.xml
Tweakie.
-
Tweakie,
Things not quite as easy as expected. I have two profiles and hence two 'xml' files in Mach3 directory. However I found by checking dates that the newest files had no extension
which threw me somewhat, the old ones had the xml extension. On copying the new files to a USB stick they acquired the missing extension?!.
Checked mill action using the new file and it was just as before; X,Y,Z axes moving in one direction only.
-
Hi Malcolm,
That .xml does have ports and pins configured, however, as mentioned in my earlier post (#24) You can't use Port1 Pins 2, 4, 6 for Output Signals when they are already being used for Motor Outputs.
If you disable Output#1, Output#2, Output#3 then your axes will most likely drive in either direction.
Tweakie.
-
Tweakie,
Dashed straight out and disabled the outputs and got no movement at all!!!.
Re-established the outputs and was back to being able to drive axises in one direction only.
Thoughts:- re-install Mach3. Look inside computer for damaged mother board which could have resulted from flat battery. Anything else?
Malcolm
-
Hi Malcolm,
Something else to try...
Disable the Outputs# (as previously mentioned).
Change Step Low Active (to a tick) for the X, Y, Z axes in Motor Outputs.
Change Step Pulse & Dir Pulse from 5uS to 2uS in Motor Tuning.
Tweakie.
-
Tweakie.
Before reading your latest mail I found a record of how I had set up mach3 originally in 2007 and noted some differences from my current settings.
I put in the 2007 settings and everything worked!.
The main difference was that the motor output step and direction pins were in the wrong columns ( if I make myself clear).
spindle was selected - step pin = 1, dir. = 17, tick, tick 1,1
input signal used was - timing enabled - port 1, pin 15, tick, X, 0
output signals used - enable#1 - tick, 1, 14, tick
output#1 - tick, 1, 16, tick
Testing the system the spindle reacted to M3,M4,M5 as it should but the spindle started to turn only when a speed of 3600 or higher was requested.
So I messed about and succeeded in stopping the spindle rotating with M3!!!, still OK with M4. Do not know what I altered but the red LED on spindle electronics permanently ON.
With M4 the red LED flashes in tune with speed of rotation.
Further work required to solve this one. NOTE record all changes made in future.
At least it is not a faulty computer, just an embarrassed operator.
Malcolm
-
Hi Malcolm,
Excellent result :)
Tweakie.
-
Hi Malcolm,
computers are especially good at making people look like plonkers, and just to think its only going to get worse. Good luck.
Craig
-
Things are now looking reasonable having having calibrated the axeses and got control of the spindle.
Used 'write' to engrave some text. It looked fine on screen but when cut was backwards and upside down.
It will save me time if anyone can say what I have done wrong.
Malcolm
-
Hi,
I thought you were supposed to be 'arse about face and on your head'...lol
Probably the most common configuration is XY 0,0 is with the tool at the extreme left hand side of the table and the
table at its furtherest away from you, ie the 'lower left hand corner'. Note this may not be the machine co-ordinate 0,0
but your work co-ords. As the table moves left X increases from 0 and as the table moves toward you Y increases from 0.
I remember it as the same orientation and direction as an X Y Cartesian graph that we were all taught at school.
The active point is the tool tip.
You can reverse directions by checking/unchecking active low on your direction pins. It pays however to keep in mind
the 'standard' orientation if you ever get given some code to cut or if you wish to post your own code.
Craig
-
To complement Craig's description, the Right Hand Rule drawing shows the tool movement in relation to the work-piece.
Tweakie.
-
Everything seems to be sorted now. Engraving text OK, ie I don't have to stand on my head anymore.
Thanks for all your help Tweakie.
Note :- Must now make a note of my setting for when the computer goes belly up again.
Malcolm