Welcome to the SeeMeCNC Community

Hello All.

According to what I have heard. I am supposed to be able to calibrate and press print without entering a custom Z-offset. I have cleaned my nozzle and recalibrated a couple of times and then tried printed but I get the same result. See below.
I have cleaned and re-calibrated three times and this print looks the same each time.

Are you calibrating cold?

TwoTone wrote:Are you calibrating cold?

Yes. When I run the script with the extruder and bed on. it crashes for some reason.

I have checked the PSU voltage which is 12.22V

Here is a video of my printing doing its thing. Hopefully this helps to find whatever the problem is.

[youtube]http://www.youtube.com/watch?v=ZtBiFqB9IFg[/youtube]
http://www.youtube.com/watch?v=ZtBiFqB9IFg

It looks like you've got a lot of tension on those clips. I put a #6 washer under them on mine so I could snug them down without too much tension which could cause some warping of the glass. If you don't have the washers, you need the screws to be loose so the bed will "float".

The too-tight clips might be bending the glass a little. You've also got them on there backwards. The wider end should be holding the glass - the thin part is a handle.

g.

Also, you clearly have a leveling issue in the Y direction (along the Z axis tower). To me, that hints at a loose end stop switch on the Z tower or perhaps it is not triggering cleanly. Anytime I see an issue directly in line with a tower like this, I immediately investigate mechanical issues in that area.

Noircogi wrote:It looks like you've got a lot of tension on those clips. I put a #6 washer under them on mine so I could snug them down without too much tension which could cause some warping of the glass. If you don't have the washers, you need the screws to be loose so the bed will "float".

I can easily find washers but do you mind showing me what you did. So I have a good understanding of what needs to be done?

Or you could just use every other hold down so only 3 are touching the glass. That way you won't get any warping since you have not over constrained clamping the bed down. This is how I run my V3, there is absolutely no reason to have 6 clamps and if not properly adjusted could actually cause problems.

I found one mistake I made with the supports under the bed. They were all not screwed on the same direction. The clips have a long and short side. Now they are all on the long side and the glass sits flush with the top. The plastic tabs are now loosened and flipped the right way too.

I also checked the Z Axis tower. The limit switch is still solid. I loosened the screws on all the towers to make sure they are all sitting nicely on the plastic tabs in the injection molded parts.

I then tried printing again and the circle is the same. I do have one other issue I need to look at and its the Y tower. the belt is maybe too tight as it makes a rubbing noise when moving. I see little bits of rubber dust on the melamine which means the belt is rubbing against the wood.

Any other ideas guys?

Still building mine, but I put the circles/dots up on my bed supports. But the glass is not flush, I can twist the blue holders over the glass but I have to loosen them a bit. Any recommendations?

morgandc wrote:Still building mine, but I put the circles/dots up on my bed supports. But the glass is not flush, I can twist the blue holders over the glass but I have to loosen them a bit. Any recommendations?

It should be flush. Your doing something wrong.

morgandc wrote:Still building mine, but I put the circles/dots up on my bed supports. But the glass is not flush, I can twist the blue holders over the glass but I have to loosen them a bit. Any recommendations?

Heres a few pictures to help you. As you can see, the black tabs that go under the bed have a long and a short side. the short side goes down, long side up.

Made an interesting discovery. When I print the same circle with MatterControl I get the following result.
Which looks much better than the S3D profile I was trying out.

What do you guys think? I think at this point the machine is fine.. the profile is screwy.

javiercordero20 wrote:

morgandc wrote:Still building mine, but I put the circles/dots up on my bed supports. But the glass is not flush, I can twist the blue holders over the glass but I have to loosen them a bit. Any recommendations?

It should be flush. Your doing something wrong.

NO!

The glass is going to be proud of the surface by about 1/16"! Javier has it assembled correctly. If your glass is even with the melamine plate, you've installed the bed support parts upside down.

g.

Javier, you might want to try setting your first layer height to 190% - that's what I'm doing with mine.

g.

geneb wrote:Javier, you might want to try setting your first layer height to 190% - that's what I'm doing with mine.

g.

I am looking over S3D's settings and noticing some odd differences.

First is the printer size:
In MatterControl its: While in S3D using the settings provided by selecting the printer in the dropdown. I get the following:
Side note Gene. so your saying the glass and the melamine should not be flush? That sounds odd considering it sits perfectly and the plastic tabs hold it nicely.

The guy that designed the machine said that's how it goes, so that's how it goes.

The S3D settings for the build diameter are off, but that's not going to affect how a print goes - it'll just yell at you if the model you've loaded would exceed that diameter.

g.

S3D specifies the size of the largest rectangle which will fit. That number of 197 above would correspond to a (197*sqrt(2))=278.6mm diameter round build area. I cranked mine up a bit.

Noircogi wrote:S3D specifies the size of the largest rectangle which will fit. That number of 197 above would correspond to a (197*sqrt(2))=278.6mm diameter round build area. I cranked mine up a bit.

I use 220 x 220 with 110 as the origin offset. You'll have to experiment with the "flip build table" to verify that it's printing in the exact orientation that's expected.

I am doing a test cube print using MatterControl right now. No custom Z-Offset. If this works then I need to figure out what is wrong with S3D.

I prefer S3D as my slicer. I'll show you all the print soon.

MatterControl prints slow.. a 20mm with 10% infill cube take 1hr . The speed is def very off.

I can't imagine how any slicer could result in the first photo you show with the thin extrusion near the Z tower. The fact that "Matter Control" printed a good looking part is purely good fortune and points even more strongly at some mechanical issue IMHO, possibly homing related on the Z tower.

Now the first layer thickness attribute in S3D could have a visual effect if it is putting down a thinner first layer, but you should be able to measure the thickness around the outline for the MC print and see if there is any variation.

mhackney wrote:I can't imagine how any slicer could result in the first photo you show with the thin extrusion near the Z tower. The fact that "Matter Control" printed a good looking part is purely good fortune and points even more strongly at some mechanical issue IMHO, possibly homing related on the Z tower.

Now the first layer thickness attribute in S3D could have a visual effect if it is putting down a thinner first layer, but you should be able to measure the thickness around the outline for the MC print and see if there is any variation.

I understand what you mean but I am trying to make sense of the results I see. This is the test cube: bottom of the cube:

Did you print a cube sliced in S3D? I may have missed something but it seems that:

1) you printed a calibration circle sliced in s3d and it shows thinness near the Z tower
2) you printed the same calibration circle sliced with MC and it looks pretty darned good
3) you printed a small calibration cube in MC and it looks pretty darned good
4) you are drawing conclusions that S3D will not print the small cube as well as MC based on observation 1

Is this correct? What you are seeing in the first photo you posted is very typical for delta printer issues. If you printed nested circles down to about the size of the cube I be you'd find that the smaller the circle, the better it looks and circles near the middle will look perfectly acceptable. The calibration errors typically present near the perimeter.

Please hold.. I made a discovery. A mistake during assembly. More soon.

Fixing it first then I'll post.