Orion not reliable

[hevilp]

I got hypercube evo, so bed is moving.

Acc and jerk I will tell later when Im at Home.

What should to untighten the screws should do? Just for my Info..

[Moriquendi]

It has helped in the past, it reduces the preload on the Orion PCB.

Are you using the original Orion parts or a specific HEVO mount.

Idris

[hevilp]

Okay, I just made my own carriage, where your top is integrated, so it is the same base.
The clamp is your shipped original.

Is it normal that the duet only shows 1000 when it got touched? No other value (except 0)

I will make a Video if this helps

[hevilp]

M566 X900 Y900 Z12 E120 ; Set maximum instantaneous speed changes (mm/min)
M203 X9000 Y9000 Z600 E6000 ; Set maximum speeds (mm/min)
M201 X800 Y800 Z250 E800 ; Set accelerations (mm/s^2)

[Moriquendi]

The speeds and accelerations look reasonable. It's perfectly normal that the Duet shows either 0 or 1000, it's a digital system, either on or off.

Let me know whether things improve with the adjustment to the four assembly screws.

Idris

[hevilp]

Good news, It seems this was the problem, now I got understable results which are reliable and repeatable.

Thank you very much.

I went down with the moving speed of the probing to 240.

It is possible that I got better results with analog mode.

Please your help into the document thread and maybe in the manual.

I used the PCB 1.2 with the 90 degree mount and got the same problems and I think to reduce the base pressure would have helped, too.

I think I will tune a bit but Im happy now with good results

[hevilp]

@Moriquendi After it runs good for a while, then I need to change my carriage.

I did not change something other and got again the problems.
Why I have the problems again? I followed all tipps and cannot reproduce good values in G29.
Only the Z probe point is in G29 the same, rest of the values are bad and the mesh says to high. I did not tighten it so much and did not change the sensitive with potentiometer.
If I move the nozzel manually with Z=0 above the bed, the distance is the same always, so the orion values are bad again.

What can I do more?
Isn't there better option for debug?

https://youtu.be/rKfbzwR_Mv0

https://youtu.be/AFNk8YpOqZo

The z Probe is always good and reliable. The Mesh grid measure is again not reliable so not useable

[whosrdaddy]

I would suggest you do multiple probes.
I set my printer to probe (max) 5 times to ensure I get consistent results for G29.

Play with A and S parameters of M558. I use A5 S0.01 (try 5 times max, 10µ tolerance)

[hevilp]

I will do so, but I think it is something with the PCB and mechanical pressure on the Orion, It worked for good before I reinstalled it

[hevilp]

I'll got now often
"19:14:51 G29 Error: Z probe readings not consistent"

[dc42]

10um is a very tight tolerance, especially if the bed is a soft material such as PEI. Try S0.02 to double it to 20um.

[hevilp]

I tested many hours and my conclusion:
To get a working, reliable, reproduceable Orion is just luck. Any potentio position, bed speed (240-400), any pre pressure on the Orion itself, and I cannot get values like before.

I get values like the pic, unbelieveable

If I do with A5 S0.02 , it always aborts with "Error: Z probe readings not consistent" at unreproduceable positions on the bed.

[hevilp]

And here

Just unbelieveable bad and just not possible results with glass on heatbed!

[ajcraig99]

I stopped using my orion after having simular problems. During tests it worked very well and was accurate and repeatable but during printing there was too much play in the mount causing wobble of the hotend.

Increasing preload on the peizo made the hotend sort of rigid enough but made the peizo far less responsive and reliable. It seems you need the perfect balance of preload on the screws but after 3 weeks of fiddling, loctite and reassembling I could not get satisfactory results from it.

I may give it another chance in the future but I think it will need some epdm or foam grommets on each screw to reduce wobble and a slight redesign of the mount.

[hevilp]

Maybe the developer can take care again of this thread?

@Moriquendi maybe?

[Moriquendi]

Something was mentioned to me by another customer that I hadn't thought to include in the manual. This customer was having issues with unreliable results in certain parts of the bed, he solved the problem by zip-tying the wires from the hotend (heater and thermistor) and especially the Orion wires to the carriage with a bit of slack so that they could not apply changing forces to the Orion.

I also notice that you have the fans on during probing. Try turning them off and see whether it helps, the vibration from the fan can cause spurious triggering.

Idris

[hevilp]

@moriquendi said in Orion not reliable:

Something was mentioned to me by another customer that I hadn't thought to include in the manual. This customer was having issues with unreliable results in certain parts of the bed, he solved the problem by zip-tying the wires from the hotend (heater and thermistor) and especially the Orion wires to the carriage with a bit of slack so that they could not apply changing forces to the Orion.

The fan is always off.

Maybe I need to tighten them more? I have them fixed with zip ties

As I can see it with my eyes, the probing process is fine.

I will do another test later this day:
G29 with single probe.
Find out the 'worst' value.
Use this spot for G28 Z
and do it there many times and see what I get.

[Moriquendi]

I thought your part cooling fan was on in the video?

When you changed the carriage and the problems came back, did you change from using the groovemoutt Orion top supplied with the kit to using the grey integrated carriage in the picture?

Idris

[hevilp]

@Moriquendi
The carriage is basicly always the same. Before I got a direct mount with BMG and pancake stepper motor. The last values were reliable and good. Even with fan running. Since one week the fan is always of to eleminate the bad affects

Then I changed the design to bowden. I removed the mount parts for direct drive and decreased the height. I did not change the rest.

I never used the groovemount you shipped. I always use my integrated, where I got bad and the very good values, too.

Here is the picture of the "old" mount.

[hevilp]

More hours of testing and still not reliable values. I removed all cables and even the filament tube.

I changed to another carriage, which is like mine but I had the hope that is better, but not.

Sometimes I can I got differences up to 1 mm, which I could see with my eyes.

The z Probe and the first mesh value got a difference like 0.006 mm, fine, Rest not