Congrats!
Awesome!!
Akk,
I’m “really close” to being able to cut the strut plates, but I’ve found a problem with the dimensions/steps settings.
I can move the machine in both X and Y. both show a distance moved less than the commanded distance. They seem to be about the same in each dimension: commanding 1000mm measures at about 997 or 998 distance on the workspace.
Where/how should I start tweaking?
Mike
Start by checking the grub screws on your pulleys.
Ryan,
I’ll check, but they were all installed with permanent thread lock, and they come back to zero without seeming to lose position.
Mike
That is a very uncommon number. Are your belts so loose they are sagging?
No play at all on the pulleys. It’s all very tight, I can’t wiggle the drive steppers by hand after I home them. The pulleys seem to be well locked to the stepper shaft.
What’s next?
MB
No sagging belts.
Is 2000mm twice the error or the same?
Did you buy everything from me or get anything elsewhere?
How are you measuring it?
The issue is nothing on the drive train is variable. So asking for 1000 and getting 500 is one thing but we don’t get subtle errors.
All the stuff is from you, I reused a couple steppers from a previous MMP3D build. I also tried 400mm and got about 398.
I used a large scale on the big measurement, but a much finer one for the 400. I can do a measurement with digital calipers, but the maximum distance is 150.
I’m using 1x2x3 block against the stepper motor side, using a Very sharp drafting pencil and a machinist’s I mark the position, them moved 150mm (per the controller) and measured 149.35.
MB
Normally 2mm and under is the grub screw. It staying the same, 2mm, for all errors means it is mechanical not electronic.
Check all three axis from minimum to absolute maximum, It will probably require a tape measure but it can reveal some issues.
here are some videos of the belt tension(s).
I’ll do that. One problem with checking maximum dimensions is that don’t have a surface as large as my potential work area. The side rails are adjustable and are currently homing to the forward edge of the table (unlike the picture above). I check the dimensions using the available table.
Mike
Is each axis off 2mm or just one? Is both sides of the Y off 2 mm?
Y2 is 39.343 measured, against 39.370 expected.
Y1 is also 39.343. so I’m off 1.55mm on both sides of the machine
That’s 0.155% so I guess I am close enough
I’ll check the X again and let you know.
MB
X measured 31.468 against an expected (800mm) of 31.469. So I guess I just wasted your time (along with my frustration). Getting a sharper pencil must have helped
My brother-in-law says the I’m expecting too much (due to my German heritage, precision, always precision) “we’re not making a Heritage Armoire!!”
Thanks, Now I can cut the strut plates without worrying that the cuts will be off by more than 0.155%.
Mike
You really need to test much larger distances. 30mm is nothing compared to 3000.
Did you run the squaring tests, that is far more important than a tenth of a mm. Set up sheets of paper or cardboard or anything and use the machine to pop holes in them or mark with a pen anything at the full extent of the machine.
There is lots to learn, testing like you are doing, is a huge learning step. It is not wasted time for anyone.
I think there is a lot of miscommunication happening.
I think this is a mix of inches (31.5) and mm.
Ryan is saying that the grub screw may be riding back and forth across the flat of the motor shaft.
The mechanical issue is probably backlash. Meaning it has a little bit of rotation where the grub screw is moving across the flat of the motor instead of moving. If you get 2mm of error at 800mm and 2mm of error at 400mm, it is probably because there is 2mm of backlash, and you lose that much distance every time you change direction.
The motors have exactly 200 steps per rotation. The pulley has exactly 16 teeth, and the GT2 belt has fiberglass reinforcement and should have 2mm distance between teeth. The belt can be stretched further under a lot of tension, but that would give you 802mm, not 798mm. The fact that you are a little under is another weird symptom.
I have to agree that it is a mechanical issue, and I think either the grub is rocking across the flat, or something is cracked.
Jamie has a tool that can make a little ruler draw in both directions. It is great at showing backlash: G-Code Test Pattern Generator
2mm over 800mm is usually fine. But if it is 2mm backlash, it will be 2mm over 10mm or over 100mm, or in whichever way the load is pushing the bit.