Good to see your machine cutting something like that Joe - looks like you got over the final hurdles! I'm especially saying that as my machine is going to be pretty similar to yours, well the top half of it anyway!
Printable View
Good to see your machine cutting something like that Joe - looks like you got over the final hurdles! I'm especially saying that as my machine is going to be pretty similar to yours, well the top half of it anyway!
Ok here you go. 90Deg cuts about 6mm deep and 60deg 10mm. It's set in MM's and used Arc's post processor so shouldn't be too jerky. My advice to speed up the job is lower the velocity and turn up the acceleration because the moves are very short, it also bounces around all over the place so don't be surprised. EnjoyQuote:
Originally Posted by Kitwn
Edit: X Y zero is bottom left corner of 400mm square.
That's brilliant, thank you very much! I'll get onto that tomorrow (It's 7:45pm in WA and I'm checking the forum on a night-shift tea break at the moment.) :moon:
Hi all, I can confirm it is a lot of fun having a working machine!
Having enough bits of material to cut is now becoming the limiting factor!
I have been trying to master the use of a round over bit to create smooth curves, which has highlighted another issue which I have now ironed out - tramming the spindle. This video is very helpful for anyone trying to do this:
https://youtu.be/D6pfIoyhgTc
I am still getting the odd problem however:
> I have had occasional random stalls on the x axis. I have slowed the velocity in motor tuning to (I think and will check later) 3500mm/min from 5000 to see if that helps.
> I have had a couple of ePID errors on MACH3 after a tool change, which led me to this:
https://www.machsupport.com/forum/in...?topic=32334.0
And this:
http://www.mycncuk.com/archive/index.php/t-10244.html
Which made me think I should look at changing my version of MACH3 to 062 as there seems to be issues when using a motion controller (mine is the ipm) with version 066 for some.
This video is also useful for those trying to square up x and y. Simple and effective method if you have limited measuring equipment like me!
https://youtu.be/lU0iZKajpoo
Hi Joe,
About the stalling, have you tuned the motor current PIDs on the motor drivers? I have EM806 rather than AM882, but I think the process is essentially the same, connecting up the RJ11 port to the COM port on your PC and running the "protuner" software.
This will make the drives change how they apply current to your motors to match them better, and allows you to set the motor current more accurately than the dip switches. Improved the speed I could get from my machine quite a bit.
Andy
Can't remember if I had ePid errors on my Mach3/IPM setup, but I certainly did see issues when I was running .066. Biggest single issue that I can remember (from a couple of years back) was losing work coordinates and similar things during an M6 tool change. I have a set of macros that I use based on, basically, code I nicked from other sources and combined with the "official" CS-Labs M31 homing instruction (rather than using native Mach3 homing). One is a plain Z height setting off a touch plate, one sets Z height off the touch plate then goes off to measure tool height against a fixed plate set in the bed, and the third just uses the fixed touch plate (for occasions when you have machined away the reference surface, for example). These last two were virtually unusable on .066, and currently I'm running .028 which is much better. It did seem to get a bit confused when I added in my wireless MPG but at least I've found workarounds for those problems.
Can't remember if I tried .062 or just went right back to .028. My recollection is that most of the changes in these last couple of versions were more to do with lathes and (I think, possibly) rigid tapping - nothing that I needed for a router.
Thanks Andy,Quote:
Originally Posted by AndyUK
I’ll have a read about how to do it, but sounds good if it will help with the stalls. Only seems to happen on x which is a big NEMA 34 turning two 2005’s via a belt. If I was buying screws now (rather than 6-7years ago) I’d up the pitch.
Hi Neale, thanks - I’ll try downgrading to v 062 or earlier and see how it goes I think. It all seems a bit random, although the ePID errors seem to come after a tool change, which is consistent with the forum posts I linked to earlier. I wonder if this also helps to explain the error I mentioned a few posts back, where (generally) the Z axis, (although a couple of times the Y too) will not move, but the DROs keep moving on Mach3, as if it is. I thought I had solved this by reinstalling Mach3, but it has happened a couple of times again since.
This evening I was cutting a test for a hangboard (a climbing training tool I want to make) - I had cut some pockets and a profile fine, changed the tool from a 1/2” router bit to a roundover bit, all fine, next changed to a v bit for some engraving, again worked ok, finally I changed the tool to a drill for the last operation on the job. It spins up fine and moves X and Y but does not move Z at all. On Mach everything is as it should be; it looks from the DROs like Z is moving according to the G code - it completes what it thinks is the job and stops the spindle. I tried jogging; X and Y are fine but I have the same issue with Z. I try ‘ref all home’. Again, Z doesn’t move but the DRO just keeps on and on until I press stop. So I open up the electronics cabinet to take a look at the drives, and the Z drive’s red LED is flashing 5 times every 5 seconds. According to the Leadshine manual (http://www.leadshine.com/UploadFile/Down/AM882m.pdf) under “12. Protection Functions” on page 13, this indicates “Sensorless Stall Protection - The AM882 can detect the motor stall status using if motor shaft speed is above 300RPM. When the detection is active, RED LED will blink five times within each periodic time (5s).” But the motor shaft never even turned! It only ever moved on the DRO on Mach, not in reality. So it doesn’t seem to fit the error code on the drive?! When I turned the cabinet off and on again to reset the drives, Z was back up and running and would jog about fine. Seems to be a similar issue to the ePID error though, which I think is also to do with Mach asking the motors to spin too fast.
This forum thread (https://www.machsupport.com/forum/in...?topic=29032.0) suggests going back to version 03.043.022 of Mach3 and not to use your old .XML file, which sounds a bit of pain TBO. It suggests the issue is connected to the CSLabs controller in relation to later versions of Mach3. Also mentions noise issues connected with usb keyboard which Jazz has mentioned before too. The CSLabs is-m manual also recommend using a PS2 keyboard (which I don’t have these days and I’m not sure the pc I’m using has the port for either!)
I’m going to try installing an earlier version of Mach on an old laptop I think, and try that... if it works I’ll set up the machine PC again with an older version of Mach and a new keyboard! Unless anyone has had a similar experience and can shed some more light..?
Those of you running an IP-M, what version of Mach are you using??
And me!Quote:
Originally Posted by JoeHarris
I think most people recommend you mark a square rather than a triangle and check if the diagonals are equal in length. I recently did this by using a 22.5 degree engraving bit to mark the corners with 1mm deep holes on the spoil board and then improvised a large pair of compasses (dividers?) by using a small pair clamped into one end of a piece of wood with a panel pin nailed through it at the other. I don't have a picture at present but can add one later if anybody's interested. I recon this will show up smaller errors than the CNCnutz method and is less prone to setup errors since you are comparing the similarity of two distances rather than trying to make absolute measurements of three.
Re the other CNCnutz video you referenced: It's a pity Peter trammed his spindle before having the chance to see my 'taut wires' thread a couple of weeks ago. His machine could use that method as the rails are above the base.