PRINTER_RADIUS, EEPROM bug, insanity

[theverant]

Yeah, the official current firmware does have that comment, it's different in SMCNC's. I'll try it it, either way in the morning and see if it works.

[Gimbal]

My apologies. I wrote "firmware" when I should have written "EEPROM". There's no way to adjust PRINTER_RADIUS from the EEPROM table. (But I certainly wish there was!)

The spelling and other errors should be corrected in the 1.10 version of the manual.

I could understand a possible EEPROM bug if PRINTER_RADIUS was something you could manipulate via the EEPROM table, but it's not. It's a compile-time adjustment and AFAIK, cannot be adjusted outside of that.

I'm not quite sure what you've got going on, but I'd recommend starting from scratch and fresh firmware. Tweak only _one_ thing at a time.

g.

Hi G take a look at this:

https://github.com/Extent421/Repetier-Firmware

Anders

[barnett]

Nathan,

Reading though this post, I'm wondering:
- is this troubleshooting something that used to work, or first time setup?
- Stock arms/effector or custom?
- Are there other firmware changes you're making - like z height for example - that work for you normally?

Barnett

[theverant]

Nathan,

Reading though this post, I'm wondering:
- is this troubleshooting something that used to work, or first time setup?
- Stock arms/effector or custom?
- Are there other firmware changes you're making - like z height for example - that work for you normally?

Barnett

Thanks for chiming in Barnett, I appreciate all the angles I can get.

I was previously using Polygonhell's firmware, and had everything in an almost good place. I never did nail it down, but I never had problems with doming like this. Pretty sure they were extruder related - my single wall calibration cubes were coming out beautifully, but nothing else would print.

Things I've done since then:

- Disassembled, sanded, and put lithium grease on the stock arms and effector (they were a little stiff)
- Moved printer to work so it could live in a climate controlled environment and I could rule out environmental issues (there is a different printer there doing beautifully)
- Took off the top, adjusted the towers (they weren't square), put the top back on.
- Messed with belt tension/position (two of the towers have belts that keep sliding over, so I've been trying to fix that)
- experimented with the bearing covers posted previously, but they seemed to be causing jerky, uneven tension spots as I moved the cheapskate manually. Most people seem to do fine without them, and I didn't want them causing issues, so I removed them.
- Added cork spacers to the motors

Things *seem* to be good mechanically - things were very stiff previously, now everything is moving very easily, but not flopping around loosely.



Will try to reset the eeprom as geneb and polygonhell suggested.

Cheers,
N

[barnett]

Yeah seems like it might be the firmware then. I've only used polygonhell's firmware so I can't be too much help with the other flavors. I would just make sure your firmware uploads are working by making a temporary change to z height or inverting a motor direction (just be ready to kill it when you run G28).

I'm using the Xnaron magarms and a different effector, but I don't think I changed any of these numbers:
END_EFFECTOR_HORIZONTAL_OFFSET = 33
CARRIAGE_HORIZONTAL_OFFSET = 35
PRINTER_RADIUS = 198.25

I have adjusted these:
DELTA_RADIUS adjustment is +3.2.
DELTA_DIAGONAL_ROD = 271.0 (more for size issues than for dome problem)

Towers square means no gap at top and bottom?

[theverant]

Towers square means no gap at top and bottom?

Towers square means square to both the top of the printer (where the build surface mounts), and the build surface - perpendicular (front/back and side/side) when checking with an engineers square. For me this means equal gaps around the top, about 2mm each. I thought I needed to pull them in tight, but after checking with an engineers square I realized this was not the case and it was just causing the towers to lean off axis. I suspect that was part of my problem when printing, but I still think most of the problem was with the extruder/hotend (didn't seem to be putting out plastic for larger prints properly).

If this is not correct procedure, how do you do it?

I should have stuck with Polygonhell's firmware, maybe I'll try it again. Might be good to compare. Rule out SMCNC's as a problem.

[barnett]

My gaps are 1mm or less at the top and bottom. That's all I checked and I haven't had it apart since I first put it together.

Sounds like you're fine there. I'd keep chasing the firmware I guess.

[theverant]

So it seems like I reset the eeprom, or at least part of it. Now the height is behaving as when I turned the firmware off - the print head is zeroing about 75mm above the print bed.

I think these are correct:
#define STEPS_PER_ROTATION 200
#define MICRO_STEPS 16


Should:
XAXIS_STEPS_PER_MM be 80?

Anyone have ideas about what might cause this behaviour? Not much closer to printing, but it does feel like progress...

[Polygonhell]

If you have a recent machine, all the values should be correct, the only one you need to check is pulley teeth. The steps are calculated from the belt pitch, the pulley teeth and the two values you quote.

If you look in the EEPROM the correct value for steps/mm with plastic pulleys and a 1.1 RAMBO board should be 3200/40 or as you have it 80.

What is the Z home position set to?

[theverant]

My machine is not a particularly new one. I got it around February. I know for sure that it is Rambo 1.1 though. All the bits and pieces (pulleys, bearing mounts, etc) are metal, though.

[Polygonhell]

Then you need to set the pulleys in configuration.h to 15 teeth and the correct steps per in the EEPROM should be
3200/30 or 106.67
If those were not correct it would pretty much be the source of all of your issues.

[lordbinky]

I counted my aluminum pulleys I was set as replacements for the split plastic ones I received. I counted the teeth and they are 20 tooth. Make sure you count the pulley teeth to be sure.

[theverant]

Then you need to set the pulleys in configuration.h to 15 teeth and the correct steps per in the EEPROM should be
3200/30 or 106.67
If those were not correct it would pretty much be the source of all of your issues.

Okay, so that seems to have fixed the height problem. Massive doming also seems to be fixed. Proceeding with calibration as normal

THANKS ALL!

[geneb]

\o/

[m4r1n5]

I am having a similar issue with mine, where the center is slightly high but the edges are not. The center is only .33mm high, but I can't bring it down any lower. Changing the firmware hasn't worked so im assuming Repetier or th e Rostock is not recognizing the Printer_Radius changes i tried to make.

After reading through the discussion, I'm still confused as to how to effectively lower the center.

[Polygonhell]

I am having a similar issue with mine, where the center is slightly high but the edges are not. The center is only .33mm high, but I can't bring it down any lower. Changing the firmware hasn't worked so im assuming Repetier or th e Rostock is not recognizing the Printer_Radius changes i tried to make.

After reading through the discussion, I'm still confused as to how to effectively lower the center.

Whe you change the printer radius value it will affect the height of the outside points, not the center.
You have to measure at least one outside point and the center after each change.

[geneb]

"Each time you change PRINTER_RADIUS, you'll need to save your changes and upload the
new firmware to the RAMBo. Note that the Arduino IDE won't be able to perform the upload if you're
still connected to the Rostock MAX with Repetier-Host. Click the “Disconnect” icon in the upper left
corner of the Repetier-Host window before you try to upload your changed firmware.

Once you've uploaded the new firmware, you'll have to re-calibrate the base of each tower as
you did in the previous steps using Scripts 1 through 4."

(pp 157, 158)



g.

[m4r1n5]

Ive changed the radius in arduino, but its never affected my printer. how do I ensure changes I make work other than just "upload" and wait for arduino to show "upload done"?

[lordbinky]

There's a line in there to turn off EEProm (set it to 0). That helps for confidence sake, as well as a power cycle after your firmware upload.

[geneb]

If all you do is do the firmware change and never re-run the basic level procedure, you're not going to see any changes.

g.

[m4r1n5]

If all you do is do the firmware change and never re-run the basic level procedure, you're not going to see any changes.

g.

So youre saying once i change the radius from 198.25 to 198.75, I go back and recalibrate the axis? I would assume my axis would be off also but theyre the same. Does recalibrating allow repetier to adhere the new radius? AND if this is the case, does anyone know the rough center height change per .5mm change in radius (i.e expand the radius by .5mm and the extruder roughly drops by .02mm). I would think enough people whove done this would know. this would definately help all of us get a quicker result than .5mm at a time.

[geneb]

When you change the PRINTER_RADIUS value, it does one of two things depending on which way you change it.

It mathmatically(sp) dips the center of the circle while raising the outer edge of the circle, or it raises the center of the circle while lowering the outer edge.

You must make a mechanical change to account for the change in the outer edge of the circle. This is why the instructions are written the way they are.

Keep in mind that the person that developed this calibration technique (that's proven to work) has probably forgotten more about delta configuration printers than either one of us knows.

What REALLY pisses me off is I just discovered I had not credited the original author with the calibration process I'm using. I'll get that fixed ASAP. Here's the source material I used: http://minow.blogspot.com/index.html#49 ... 9571907051

Sections #2 and #3 are the relevant portions for the process we're discussing.

g.

[m4r1n5]

Turning off the EEPROM, loading the new Printer Radius value, THEN turning it back on through Arduino definately does the trick!

As to my other question on Hot end travel height per 1mm Radius change, I'm seeing roughly .3mm travel up on the hotend for every 1mm I shrink the Radius. It might not be consistant with everyone, but its agreat round start if you're 3-4mm off and dont want to go .5mm increments.

I do think you should add the EEPROM issue in a future manual update. VERY important. would have saved me about 4 days.

[geneb]

That tip was added to the Rostock calibration blog a few weeks ago. I'll get it added to the docs.

g.

[JohnStack]

Thanks Geneb. Too many problems/much confusion about this right now - and your process is _very clear_.