Profile

Cover photo
Ben Delarre
85 followers|192,225 views
AboutPostsPhotosYouTube

Stream

Ben Delarre

Discussion  - 
 
Ok guys I need some more help again.

Got the machine back together and what I thought was moving correctly. Got my e3d extruding nicely and my bondtech extruder working perfectly.

So I decided to run the full burn-in gcode again now that I'd got everything settled in.

Works great for about half an hour, then all of a sudden something sticks. Last time it was the z axis sticking on an upwards move, previous time it was xy. So I quickly kill the machine when the steppers start making that godawful noise, power it all back up again after a reset and everything is moving fine.

Since this was the second time it happened I figured, oh the steppers must be overheating, so I checked their temps before killing the machine and they weren't that hot at all, nowhere near where I would expect them to be having thermal issues anyway. Warm, but not hot.

To recap, I've oiled everything, the XY shafts and Z smooth rod with sewing machine oil and the Z leadscrews with synthetic grease.

Each time this happens I've tried reseating the xy bearing block holders and jiggling my z clamps around a bit and I think I've got it all working, I'll get halfway through the burn-in and then it'll do this again.

I'm at a bit of a loss as to what to do now. Any suggestions? Is it time I just gave up on my plastic parts and asked one of you kind folks to print me some new ones?
1
Daniel fielding's profile photoOliver Seiler's profile photoVic Catalasan's profile photoBen Delarre's profile photo
15 comments
 
Not a bad idea, I'll try that after a complete rebuild if the problems are still occuring.
Add a comment...

Ben Delarre

Discussion  - 
 
Hey all,

I'm setting up a new workshop and just realised I need a PC out there to run Chilipeppr on to use my Shapeoko.

Question: Has anyone used a Raspberry Pi 2 as a viewer for Chilipeppr?
1
Kurt Schuppert's profile photoAndres Santos's profile photoBen Delarre's profile photoJohn Lauer's profile photo
13 comments
 
Well, with the laptop, why I like it, is that you take it into the shop to jog around with the machine and get the Gcode all dragged/dropped from your CAM app. Get the machine zeroed out correctly, i.e. probed, then send the file. Then bring the laptop with you as you leave the shop so it doesn't get dusty. I tend to monitor my jobs thru the webcam app as I'm upstairs away from the machine.
Add a comment...

Ben Delarre

Discussion  - 
 
A bunch more photos, sadly no progress.

The smooth rods now fit into the clamps perfectly, no problem there at all.

However....if i place the bottom z mounts in the correct measured positions then I can barely turn the leadscrews at all. I have to loosen one side which pops back a little further and then the screws move fine.

However, I still get binding when trying to drive from the motor so I'm guessing its still too tight.

At this point I'm at a loss, spent the whole day on this today and no progress, pulled it apart and back together again more times than I can count.

Anyone who has built the V2 z-axis got any tips?

Otherwise I think it might be time to give up and see if new plastic parts would help. Anyone up for printing me a set of Z axis parts for the Eustathios V2, including the motor mount in ABS?

I'm happy to pay whatever is necessary. I'm in Los Angeles, so someone in California would be best as I'd rather not wait for postage.
2
Isaac Arciaga's profile photoJason Smith's profile photoBen Delarre's profile photo
3 comments
 
+Jason Smith yes it does. its almost definitely bad tolerances on my printed parts. I managed to improve it substantially by sanding down one of the top clamps and it now runs ok-ish. There's some rubbing noises and an odd periodic donking sound that I think is the leadscrews rubbing on the bearings. The leadscrews also show a pronounced wobble at the top which I guess means they are not properly lined up at all. I don't think I can sand them all down enough to get that right so its new part time!

+Isaac Arciaga I'm in Redondo Beach, so not far at all. When do you think you would be able to print them all off? Happy to pay you for your time, effort, filament and wear and tear, I just want my printer working! :-)
Add a comment...

Ben Delarre

Discussion  - 
 
Hey all,

I need a little help figuring out how to assemble the z-axis.

I tried attaching the z-axis bed supports to the bed extrusions, screwing the z axis lead screws into the bronze nuts, and then inserting them into the z axis supports. This sort of worked, but i soon discovered that the entire axis is somewhat skewed.

Its clearly leaning to one side, I managed to force the smooth rods into the clamps but I think the whole mechanism is rather tight now.

Any suggestions as to why the z is skewed?
1
Isaac Arciaga's profile photoErik Scott's profile photoBen Delarre's profile photoEric Lien's profile photo
12 comments
 
+Ben Delarre if it is just over extruded you could just use a rat tail file and test fitting. Just take a measurement of the rods in the mounts (no bed installed), then compare that to the spacing of the bearings on the bed to see if there is a difference. That will tell you if there is an interference... And how much must be removed.
Add a comment...

Ben Delarre

Show off your work!  - 
 
While we didn't use FastLED for this, there were some really helpful tips here on driving all these APA102 leds.

In the end we had 1768 leds being driven as 17 separate strips by the LEDscape engine running on a Beaglebone Black. These were all output through one of RGB123's excellent breakout boards.

The entire array is running off of a single 600W 5v supply. We actually turned it up to 6v to account for the voltage drop when the LEDs go full white. We then have a second internal regulator to ensure a stable 5v supply for the beaglebone black, while the breakout board level shifts the 3.3v signals up to whatever the LED voltage is at the time.

This all seems to work great. These animations are actually being generated in JavaScript using NodeJS via our FLED engine (https://github.com/SupplyFrame/fled). Once the pixel data is calculated it is streamed over wifi at 30fps to the beaglebone black, which then renders it out to the leds via LEDscape.

Thanks all for the help, enjoy!
7
1
Ryan Clough's profile photoBen Delarre's profile photo
11 comments
 
Awesome! more stuff I didn't know.

And yeah, I think it makes sense to have LEDscape as it is, it makes an excellent driver for externally generated graphics. If you want to do that with teensy it starts to get a bit convoluted cause then you have to worry about all the interrupts or sending data over serial, etc.....at that point its time to move to something like BBB.
Add a comment...

Ben Delarre

Discussion  - 
 
First moves in xy. Y moves seem great. Not even run the burn in gcode completely yet. X moves are a little noisy. Not yet sure why. I moved one of the pulleys away from the bearing block a little that seemed to help, so perhaps they are rubbing on the bearings?

Z axis pulleys arrive tomorrow then hopefully I can finish that stage. Printing will have to wait till after I move house and get the shapeoko up and running again to cut the alu plate.

For those with a careful eye you might notice something odd about my eustathios, can you guess what? Bear in mind you are looking at it from the left. 
7
1
Ben Delarre's profile photoEric Lien's profile photoSalahEddine Redjeb's profile photoOliver Seiler's profile photo
15 comments
 
Thank you very much +Eric Lien !
Add a comment...
Have him in circles
85 people
Cheryl Slinkard's profile photo
James Hollingsworth's profile photo
Yooch Wan's profile photo
Miroslav Bajtoš's profile photo
Fin Bot's profile photo
Mike Miller's profile photo
Chungwei Hsiung's profile photo
Shannon Wolfe's profile photo
Riley Porter (ril3y)'s profile photo

Ben Delarre

Shared publicly  - 
 
New Mawkin music video! Awesome stuff lads.
1

Ben Delarre

Discussion  - 
 
OK here's some photos. Not sure it helps make anything clearer though. You can see in the photos that the left side smooth rod fits into the clamps perfectly, while the right side is a good 2mm too far over.

Only way I can get the fight side in is to move the top clamp back so the whole axis is twisted which is obviously wrong.

It feels like the bed is too wide but I have checked it against the solidworks model and it is spot on and cut by misumi anyway.

I was also wondering if perhaps the lm10uu bearing in either or both sides is not square resulting in the whole axis being tilted to the right.

I don't know. Its close but it's just not coming together right.

Any suggestions? 
1
Isaac Arciaga's profile photoBen Delarre's profile photo
18 comments
 
Got the xy up now. But it's caused a donking noise again in the z. A little more tuning I guess. Still.... Almost there! 
Add a comment...

Ben Delarre

Shared publicly  - 
 
 
While we didn't use FastLED for this, there were some really helpful tips here on driving all these APA102 leds.

In the end we had 1768 leds being driven as 17 separate strips by the LEDscape engine running on a Beaglebone Black. These were all output through one of RGB123's excellent breakout boards.

The entire array is running off of a single 600W 5v supply. We actually turned it up to 6v to account for the voltage drop when the LEDs go full white. We then have a second internal regulator to ensure a stable 5v supply for the beaglebone black, while the breakout board level shifts the 3.3v signals up to whatever the LED voltage is at the time.

This all seems to work great. These animations are actually being generated in JavaScript using NodeJS via our FLED engine (https://github.com/SupplyFrame/fled). Once the pixel data is calculated it is streamed over wifi at 30fps to the beaglebone black, which then renders it out to the leds via LEDscape.

Thanks all for the help, enjoy!
11 comments on original post
1

Ben Delarre

Discussion  - 
 
For those building the Eustathios V2. Be warned, 36 tooth 8mm bore pulleys do not fit in the Z-axis blocks as they currently stand. Not a big deal, but 36 tooth are easier to find than 32 tooth.

Have ordered some 32 tooth replacements from ebay...z-axis on hold!
1
Dat Chu's profile photoGus Montoya's profile photoBen Delarre's profile photo
4 comments
 
Yeah. Well at least the ones I got from inventables don't. The rim of the pulley rubs on the extrusion. There's not much in it a mm or two smaller might fit fine.
Add a comment...

Ben Delarre

Discussion  - 
 
Success!

I added the shims and that killed most of the noise, there was still a nasty high pitched squeal when making diagonal moves.

So I repeated my earlier alignment procedure. Moved the carriage to the middle, loosed both the bearing blocks in one corner, made sure they could move up and down and then slid the carriage into the corner and locked up the bearing blocks. Repeated for all corners.

After that things started running perfectly, no grinding noises no high pitched squeals, just stepper motor whine and belt noise.

Got to say I love the noise this thing makes when changing directions. :-)
7
James Rivera's profile photoEric Lien's profile photoJason Perkes's profile photoBen Delarre's profile photo
7 comments
 
+Jason Perkes yeah, the full XY gantry is far more hypnotic than the X gantry with moving bed model on the office Prusa. Can't wait to get this one finished!
Add a comment...

Ben Delarre

Discussion  - 
 
I need some help figuring out how to get my xy axis smooth.

Before putting on the belts the axis was reasonable. I could move in x or y without too much effort as long as I moved both ends of the axis together to avoid racking.

I have hooked up the belts on the axis and the motors. I loosened the pulley set screws on each axis and tightened the belts then shifted the axis to one end so it was square and then locked the pulleys.

Now it takes a lot of force to move the axis. I can do it with one hand holding the central carriage but it's not exactly easy. Takes a fair force to move it, once moving it feels reasonably smooth but still requires some force. Is this normal? Obviously there's a lot of friction from the belts and the motors, I'm just not sure if it's too much or not.

I have not yet hooked up the motors to my controller, that's a job for tomorrow night so I guess I will see if it moves smoothly then.

Any advice on getting this all optimal? Not added any oil yet, is there a best type? I have some silicone based stuff and some sewing machine oil. 
2
Whosa whatsis's profile photoDerek Schuetz's profile photoBen Delarre's profile photoEric Lien's profile photo
14 comments
 
+Ben Delarre glad to hear it worked. It's kind of finicky at first... But when it is aligned it is evident. Can't wait to see video of first moves, and your first print. The break-in code helps smooth things out even further.
Add a comment...
People
Have him in circles
85 people
Cheryl Slinkard's profile photo
James Hollingsworth's profile photo
Yooch Wan's profile photo
Miroslav Bajtoš's profile photo
Fin Bot's profile photo
Mike Miller's profile photo
Chungwei Hsiung's profile photo
Shannon Wolfe's profile photo
Riley Porter (ril3y)'s profile photo
Links
Basic Information
Gender
Male