Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: fusion on June 09, 2018, 01:32:50 PM
-
Hello,
I have built a lathe and I'm having trouble with threading.
My system is very basic. XP Mach3 .062, CNC4PC C10 BOB, C3 Index board and optical sensor, C6 spindle Control board. My disc has a slot that is 1-1/2" wide. I have tried different widths from 1/2" up to 1-1/2" with no real improvement. I use NEMA 34 stepper motors with Leadshine Drives.
The RPM is steady to +/- 1 rpm. I have a 3/4 Hp motor driven by a VFD.
The Index signal is working as far as I can see with the diagnostics LEDs. I've tried active Hi and Low. It's on Pin 15.
I have slowed down the motion to 50 IPM with 12 acceleration. I'm not loosing steps. I usually thread at 400 RPM or there about.
I use the simple thread wizard for the code. I have been trying to make a 3/8-16 thread.
I have tried Index Debounce settings from 0-1000.
When I select Use Spindle Feedback in Sync Mode, it makes random motions prior to each pass. Strange ZigZag motions. I'll try to attach a screenshot.
When I do not select Use Spindle Feedback in Sync Mode, the strange ZigZag motions go away, but it makes a bad thread because it's not starting from the same startpoint.
What I don't understand it why is Mach sending the ZigZag commands.
Thank you for your help!
Mike
-
hi
what are you using to get the signal for threading ?
change the pin you are using from 15 to another pin it could be a bad pin on the parrallel port ?
i had the same problem as you about 18 months ago & it took 12 months to sort & it was a silly problem that could have been sorted very easily , i was using mach 3 062 version on a windows xp system & all i did was to buy a windows 7 32 bit system and loaded all my old xml files , tool files and all other files needed & it started working as it had done before
this is the type file i use for threading for 1/8" bsp external thread
G95 G64
M6 T0505 (externnal threading tool 55 deg)
M3 S500
G0 X9.75 Z15.00
G76 X8.55 Z-10.00 Q1 P0.090 J0.090 L45 H0.080 I27 C0.350
G93
G0 X40.00 Z40.00
-
Hi ministeam,
I use a C3 board and optical sensor from CNC4PC. With that I use a disc about the same size as a DVD, with a 1-1/2" slot.
I'll try another pin and see if that changes anything.
Mike
-
I tried a different pin, 13. No change.
I removed the C3 and optical sensor from the system, and installed a Hall effect sensor and magnet. I tried many different Debounce settings. No joy.
I tried ECP mode on the PP. No change. Set it back to EPP.
The RPM reading was a bit more stable using the Hall effect sensor.
When I run the program while the diagnostics page is up, I can see the wait for trigger flash with each weird motion.
I also tried version 066. It was the same.
-
hi
something i forgot to mention last night was i used the c3 boards at one time but found them very unreliable , i have 4 off on my bench that will not work . i have been using a round proximity switch from cnc4you for the last 12 months & it has worked very well , the only problem is it will not work using a plate with a slot i had to use a piece of ali about 3 /4" wide x 1 1/4"long to cover the slot & overhang the edge of the plate , i will look at my xml to see what settings i had
all the best
robert
-
Hi,
I tried something new. I use a switching power supply for my steppers. I thought maybe that was causing electrical noise and interfering with the index signal. I disconnected the power supply to eliminate any noise from the power supply or stepper drives. It still has erratic movements. They are not as bad as before.
It seems to be a noise issue. I'm just not sure how to isolate the cause or eliminate it.
Do I need to ground the BOB?
Mike
Sent from my SM-G892A using Tapatalk
-
In a effort to simplify the troubleshooting, I disconnected the C3 index board and the C6 spindle control. So the only active component is the C10 breakout board. The spindle was turned on at the VFD and held a rock steady 400 rpm. No stepper drives or power supply was active either. Even with it as bare bones as this, Mach was still making strange moves.
The C10 BOB is new. Could it be bad? It seems to function just fine with the expecption of threading.
I'm about the throw in the towel and switch to the Centroid Acorn.
Thanks for any help.
Mike
Sent from my Acer Chromebook R11 (CB5-132T / C738T) using Tapatalk
-
Fusion,
- different widths from 1/2" up to 1-1/2" with no real improvement
The turn manual provides for calculating the slotwidth. The slot width can be wider than necessary.
If too narrow the RPM may not show. Since you changed over to a Hall effect mute point on the
width of the slot. Just manualy turn the spindle by hand and you should see the index led turn on and off
in diagnostics. I cover testing of the index in Threading on the Lathe writeup.
- I have tried Index Debounce settings from 0-1000
Index debounce is simply how many times Mach must see the index before it accepts the input.
I personaly use a value of 100, others can get away with 10. If a high index value is used you may
have a long waite time before the index signal is accepted and the threading begins ( changes to G95).
- motion to 50 IPM with 12 acceleration
I also use the Simple Threading Wizard . You should always click the "calc number of passes"
and it will inform you if the accel / feedrate is not adequate based on you motor tuning for the
inputed Spindle rpm.
- I also tried version 066
Use version 062 and don't use 066 as it is problematic.
- It seems to be a noise issue
VFD's can cause noise problems. Not sure noise is your problem.....
- You have backlash enabled and thus there is motion which takes place to implement the backlash.
Suggest you turn backlash off and see if the strange motion goes away. As far a using backlash relative to
to threading................GET RID / minimize BACKLASH if you want accurate threading. PERIOD!
- In a effort to simplify the troubleshooting, I disconnected the C3 index board and the C6 spindle
control. So the only active component is the C10 breakout board. The spindle was turned on at the
VFD and held a rock steady 400 rpm. No stepper drives or power supply was active either.
Even with it as bare bones as this, Mach was still making strange moves.
The C10 BOB is new. Could it be bad? It seems to function just fine with the expecption of threading.
HMM........ the signal for RPM to Mach for RPM readout is coming from the break out board with the
VFD as the source of the signal. There can only be one signal for rpm readout when threading.
I thought that you changed from using the C3 index board to a Hall affect? So the hall affect should be
used as the single source and you also click the spindle button on to get the RPM readout.
Manualy turn on and manualy set rpm of the VFD.
You don't want two signals fighting each other for the basis of the RPM used in threading!
Just some thoughts,
RICH
-
- You have backlash enabled and thus there is motion which takes place to implement the backlash.
Suggest you turn backlash off and see if the strange motion goes away. As far a using backlash relative to
to threading................GET RID / minimize BACKLASH if you want accurate threading. PERIOD!
I've done all I can to eliminate backlash. The problem I am seeing is there with backlash enabled or disabled.
HMM........ the signal for RPM to Mach for RPM readout is coming from the break out board with the
VFD as the source of the signal. There can only be one signal for rpm readout when threading.
I thought that you changed from using the C3 index board to a Hall affect? So the hall affect should be
used as the single source and you also click the spindle button on to get the RPM readout.
Manualy turn on and manualy set rpm of the VFD.
I did this too. I had the spindle manually controlled at the VFD set to 400 RPM. The stepper drives and PS were disconnected. The C3 board was disconnected. The C6 board was disconnected.
You don't want two signals fighting each other for the basis of the RPM used in threading!
Just some thoughts,
RICH
I do wonder if I may have a bad BOB. I hate to keep throwing money down a hole.
I uploaded a video to you-tube. This run is fairly clean, but there are some strange motions being sent by Mach3.
https://youtu.be/RAAVSQNnDMk (https://youtu.be/RAAVSQNnDMk)
The attached image shows some really strange motions.
-
hi mike
does the bob you are using do threading , in the past i have seen bob that are only designed for milling
all the best
robert
-
I assume it's OK to use this BOB for threading. I have read other posts where people successfully used it on lathes with threading. When I exchanged emails with Arturo at CNC4PC he didn't say it was a problem.
Sent from my SM-G892A using Tapatalk
-
Manualy turn on and manualy set rpm of the VFD.
I did this too. I had the spindle manually controlled at the VFD set to 400 RPM. The stepper drives and PS were disconnected. The C3 board was disconnected. The C6 board was disconnected.
So you keep it basic. Just use the VFD manually for a test. The hall effect index is used to provide input to Mach for the RPM.
So.....load Mach3 , turn on and set the vfd ( you should not see any rpm indication until the you click spindle button, then you should see rpm via the input from the index ),load the threading program and do an air test of the threading. better yet would be to just do a scribe test of the threading for a test.
I personally also prefer to have the wizard provide g32 threading code since every axis move is delineated in the code.
So maybe change the macro ( how to change the macro is in the Threading On the lathe Writeup).
RICH
-
Thanks Rich. What you described is exactly what I did. The video I linked to shows the weird movement Mach is calling. Those strange ZigZag moves.
I will have to try the G32 setting.
Another weird thing it does sometimes is slowing the feed to a crawl. In threading it has to move fast in the Z to match the pitch, but sometimes maybe even mid-pass, it slows to a snail's pace.
Arturo sent me a video that shows how to check for continuity between the power supply negative terminals and the chassis and ground. My machine passed that test.
Sigh...
-
Another weird thing it does sometimes is slowing the feed to a crawl. In threading it has to move fast in the Z to match the pitch, but sometimes maybe even mid-pass, it slows to a snail's pace.
That is a motor power problem as it relates to the depth of cut you are taking. Note that the VFD has no feedback to maintain the rpm
under motor load.
I strongly suggest you read Threading on The Lathe writeup I did as it makes no sense to repeat what has already been noted in the writeup.
RICH
-
There's no cut being made. The stepper drives are not powered. Mach only thinks they are moving. The motor rpm is steady within 2 rpm.
Sent from my SM-G892A using Tapatalk
-
I did read the file you suggested.
Sent from my SM-G892A using Tapatalk
-
Update!
After removing everything in the system except the index pulse and the breakout board, I was still having this problem.
I decided to reinstall all the system components back to my original setup. I still had the same problem.
I then reformatted the computer drive and installed XP, the drivers, and Mach3. Then threading worked perfectly. I have no idea what was causing the issue, but it's gone now. WOOT!!!!
-
Fusion,
Glad it's working.
I didn't forget this thread just dealing with so many things and didn't have the time to get into a detailed look.
I did find that, when simulating, that there were additional moves shown in the graphic screen and didn't matter if G76 or G32 was used for threading.
Additionally , depending on how one selects a the start point of the threading the graphics would change, and BTW that was using you XML file.
To get a "cleaner" graphics display of the gcode produced you select a start, referenced / tool change location, with a 3x the lead away from physical
location of the material and I didn't get goofy display. Note that it's a screen issue and you should not actually have an actual axis move as shown on the screen. I know the above is not clear but, again, just don't have time to write it all up / explain in detail.
That's why I said to actually do a thread!
Anyway, it's working for you and life is good......
RICH