Trouble with bed level on my Ormerod 2

Hi,

ever since I bought my Ormerod 2 I have never been able to print reliably... I had all sorts of problems (missed steps, Duet replaced, very often first layer adhesion is faulty, etc.). All in all I have spent a lot more time and filament trying to make the printer work than actually printing things...

Now after forgetting about the printer during quite some months, I'm back on it. I made sure the bed was near perfectly level and I happily printed a rather larger piece at 0.1 resolution, which came out almost perfectly (apart for some oozing around the corners). I was surprised, happy and life was good, I was even planning on trying ABS to make stronger replacement parts for the printer as some are showing cracks.

After my almost-perfect-4h-print I tried printing Z gears, and again adhesion problems (never could print that one). I checked the bed level, and one corner (X200,Y200) is low by more than one mm. I'd like to correct this, only I don't know how I can set correctly this particular angle as there is only one screw in the middle of the Y axis and the other angle is set OK... Does this means the bed is bent? What can I do about it?

And by the way don't take the beginning of the message as a bare complaint, I learned much and I'm happy with it, I chose an open source printer for a reason, I deal with it. It is just really frustrating having a 3D printer for 2 years now and never having been able to really use it...

Quote
mat_fr
...I checked the bed level, and one corner (X200,Y200) is low by more than one mm. I'd like to correct this, only I don't know how I can set correctly this particular angle as there is only one screw in the middle of the Y axis and the other angle is set OK... Does this means the bed is bent? What can I do about it?..

If your smooth rods is twisted you will always end up with one corner low or high (if your glass is perfectly flat that is) - the fix is to adjust your acrylic end parts, here's an old posting about it

While we wait on (for) a functional auto-probing...:
[forums.reprap.org]

But you should still be able to print small parts like the z-gear with the bed out of wack if you zero the bed in the center of the printing position

Erik

Agreed - it is your Y ground rods that are not parallel which is the culprit, so the front of the bed goes up & down relative to the back as you move in Y. This probably happened as you moved the printer into and out of storage because the acrylic end plates are not fixed all that solidly. In fact it is possible to "adjust" it by simply rotating the acrylic plates by hand - careful not to apply so much pressure that they crack. In your case the front of the right leg needs to be twisted up slightly and/or the left leg down (looking at the printer with the Z post at the back).

I had been planning to replace the acrylic plates with laser-cut aluminium plates & more solid fixings to the Y extrusion which also allow the front ground rod to be adjusted (one end would be able to be adjusted in Z, and the other end in X), but I have decided to build a new printer of my own design from scratch instead (when I get a round tooit).

Dave

Thanks for your answer Erik.

I have indeed changed the Z rod which began to have hard points leading in missed steps. But it is not entirely straight. I didn't think of that reason why the bed would bend... I'll try to fix that...
Thanks for the lead, I'll read it and give it a try if I can.
Meanwhile I'm printing the gears (iamburnys version) at 0.2 resolution, I cheated and tell slic3r that the bed is 100x100 so it centered the print closer to the origin. Not a great solution but so far so good. Hope the dimensions will come out OK (another problem I had since the beginning: I never succeeded in printing articulated parts because dimensions always came out a little too tight).

It's not the Z rod, it's the Y rods - and it is not that the rods themselves are bent, just that they are not exactly parallel to each other. They need adjusting, not changing ...

Dave

Quote
ormerod168
If your smooth rods is twisted you will always end up with one corner low or high (if your glass is perfectly flat that is) - the fix is to adjust your acrylic end parts, here's an old posting about it

While we wait on (for) a functional auto-probing...:

Erik, what problems are you having with the auto bed compensation? The 4-level probing scheme can compensate for twist in the rods (and it does on my Ormerod) - although it is better to eliminate the twist as much as you can.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Ho, sorry Dave I hadn't seen your answer.
And reading it I realize I hadn't the right rod in mind as I was talking about the Z axis

I have replaced the X axis arm by an aluminum version and I suppose it would be a good idea to have the end plates as well. But honnestly I don't know how to find these (for the arm I ordered it from this forum actually).

Quote
mat_fr
Ho, sorry Dave I hadn't seen your answer.
And reading it I realize I hadn't the right rod in mind as I was talking about the Z axis

I have replaced the X axis arm by an aluminum version and I suppose it would be a good idea to have the end plates as well. But honnestly I don't know how to find these (for the arm I ordered it from this forum actually).

Check out [www.3dprintandstuff.co.uk]

Edited 1 time(s). Last edit at 09/06/2016 09:08AM by dougal1957.

Thanks Dougal, I'll think about buying one of these

I could uplift the left plate, both are not level on the desk but the plate seems to be, so that's good news.

Thank you all for your help, I suppose I'll go into more trouble soon but this forum is great to get solutions when I think there are none left

By the way I could eventually print the Z axis gears, which is great, but even at 102% scale it's still too small (in every dimensions, for what I can tell). I'll try with 103%. Is it a common problem? When I run calibration tests motor movements seem OK. I know there can be many reasons why holes come out too small, but here it is all dimensions (between 2% and 8% from what I quickly measured). Is it a common solution to just upscale the print in the slicer for mecanical parts? OK I'm getting off subject for this thread, sorry

Edit : Bed equation fits points [50.0, 10.0, -0.042] [60.0, 180.0, -0.077] [210.0, 180.0, 0.010] [210.0, 10.0, 0.031]
Seems nice to me

Edited 2 time(s). Last edit at 09/06/2016 12:24PM by mat_fr.

I had the same problem with the Y end plates on mine so I designed a replacement end plate that allowed me to raise and lower the front Y rod to get the bed level at the front. Have been using this for quite a while now, and print quite a few large items that are 20cm across without bed compensation. Link to my design below.

[www.thingiverse.com]

I'll sure check it out.
If I can (at last) get good prints I suppose I will first print improvement or replacement parts for the printer itself so hopefully I have less details to worry about in the future

Quote
dc42

Quote
ormerod168
If your smooth rods is twisted you will always end up with one corner low or high (if your glass is perfectly flat that is) - the fix is to adjust your acrylic end parts, here's an old posting about it

While we wait on (for) a functional auto-probing...:

Erik, what problems are you having with the auto bed compensation? The 4-level probing scheme can compensate for twist in the rods (and it does on my Ormerod)

Hi David - No problems at all but I don't use it much, BTW thanks much for all your work on the firmware, everything is running very smooth!

Quote

- although it is better to eliminate the twist as much as you can.

That's what I did - all of it - I only use comp. to take out a small heat-bulge at the center of the glass and only if printing extreme low layer height at high bed temp.

Erik

Apart from small circles, the dimensions are just about spot-on with my printer, so not sure what your issue might be. Maybe try printing this cube (use 10% infill) and measure it after printing - it should be exactly 20mm per side.

Dave

Edited 1 time(s). Last edit at 09/07/2016 09:44AM by dmould.

I wonder just how accurate is the IR sensor when dealing with bed autolevel.

I mean I used to recalibrate the Z axis each time when turning the machine ON. it's allways off by 0.1 to 0.3mm from one startup to the next, and depending on the lighting condition. I can live with that, but actually I hated having to calibrate the IR sensor itself instead of just manually zeroing the z axis at startup, because by default each print generated by slic3r homed all axis, so if the sensor was not spot on it was useless.

I assume the bed leveling works by difference between readings on each corner, but I wonder if the lighting, the kapton cleanliness, the paper position, etc. can adversely affect the reading and hence the leveling?

Which IR sensor:

1. The original Ormerod 1 3-wire sensor => very sensitive to ambient IR.
2. The Ormerod 2 4-wire sensor => 15 times less sensitive to ambient IR than the Ormerod 1 sensor.
3. My differential IR sensor => insensitive to ambient IR, unless it is so strong that it saturates the phototransistor.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

To get accurate readings for bed levelling the sensitivity to ambient light is not too important because that should remain constant between readings and you are only looking for the differences in heights, not the absolute heights. However significant differences in the texture or colour of the bed at the various probe points will affect the accuracy no matter what sensor you use.

Dave

I have the Ormerod 2 4-wire sensor.
Just the other night I didn't think about it and run a Z homing with my desk 20W hallogen light on. I really thought something was wrong with the printer for a few seconds.
The printer is on the side of my desk, a windows is just 50cm away so I wondered if sunlight (ambiant, not direct, but still could matter) could light some corners more than the others.

Anyway I'm just being curious, the bed is pretty much level now and the IR sensor seems relevant (and autolevelling doesn't seem that much important with the level I have right now).

Thanks

Matthieu

Quote
dmould
To get accurate readings for bed levelling the sensitivity to ambient light is not too important because that should remain constant between readings and you are only looking for the differences in heights, not the absolute heights. However significant differences in the texture or colour of the bed at the various probe points will affect the accuracy no matter what sensor you use.

Dave

True, but I infer from the question that the issue he has is not so much with bed compensation as with the Z=0 height.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
mat_fr
I have the Ormerod 2 4-wire sensor.
Just the other night I didn't think about it and run a Z homing with my desk 20W hallogen light on. I really thought something was wrong with the printer for a few seconds.
The printer is on the side of my desk, a windows is just 50cm away so I wondered if sunlight (ambiant, not direct, but still could matter) could light some corners more than the others.

Anyway I'm just being curious, the bed is pretty much level now and the IR sensor seems relevant (and autolevelling doesn't seem that much important with the level I have right now).

Thanks

Matthieu

Halogen light is bad for the sensor, so is sunlight coming directly through the window when there are no clouds. Fluorescent and LED light are OK.

Quite a lot of Ormerod users have upgraded to my differential mode IR sensor.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks I'll have a look at the mod.

Quote
mat_fr
... so I wondered if sunlight (ambiant, not direct, but still could matter) could light some corners more than the others.

It can in X, but not in Y (because all Y positions are at the same place relative to the light - it is the bed that moves in Y, not the probe)

Dave