Acclerometer on a Delta printer

CaiJonas

Hi Guys,

I've just installed an accelerometer to my Deltaprinter but I'm not shure how to deal with the result. Maybe you can help me out.

I've used this code for measuring:

G28
G1 Z100 F20000
G1 X-50 G4 S2 M956 P0 S1000 A0 G4 P10 G1 X50 F20000

and here is the result:

Best Regards,

Cai

zapta

@caijonas, as a layman, I would think that your main resonance is around 65Hz. If you want to use to to configure some input shaper you need to see what kind of configuration it expects.

My 2c.

dc42

@caijonas it would be interesting to see what the results are for a Y move and for a diagonal XY move too. Also, it can help to see the accelerometer trace before doing the frequency analysis. I agree that there is a resonance at about 65Hz, but there may be another at about 28Hz.

CaiJonas

@zapta Yes I guess youn are right. What do you mean with kind of configuration?

@dc42 How can I make the accelerometer trace visible, not shure what you mean.

Here are the results:

Y Move:

G28
G1 Z100 F20000
G1 Y-50 G4 S2 M956 P0 S1000 A0 G4 P10 G1 Y50 F20000

XY Move:

G28
G1 Z100 F20000
G1 X- 50 Y-50 G4 S2 M956 P0 S1000 A0 G4 P10 G1 X50 Y50 F20000

So I guess you are right with 28Hz.

Best Regards

dc42

@caijonas you have posted the frequency analysis traces from the DWC Accelerometer plugin. It would be useful to see the raw data traces too.

CaiJonas

@dc42 sorry, not my day

I've done again in all directions because of some congfig changes I've made regarding acceleration and yerk. So here are the latest data:

x-direction:

G28
G1 Z100 F20000
G1 X-50 G4 S2 M956 P0 S2000 A0 G4 P10 G1 X50 F18000

y-direction:

G28
G1 Z100 F20000
G1 Y-50 G4 S2 M956 P0 S2000 A0 G4 P10 G1 Y50 F18000

xy-direction:

x_raw.csv
y_raw.csv
xy_raw.csv

Hope that now it's more usefull.

For me it looks like there is resonance at 67 Hz

BTW I'm using Adafruit LIS3DH with this config:

M955 P0 C"spi.cs3+spi.cs4" S4800 Q400 R10 I20

Thank you very much for your support,

Berst Regards,
Cai

dc42

@caijonas I agree, you have resonances at around 67Hz. It's particularly interesting because it shows a double peak, which is characteristic of two resonances at a similar frequency with coupling between them. This makes sense, because I think the elasticity of the belts+motors will give rise to three modes of oscillation, however XY motion will only excite two of them strongly.

Sometimes the raw data trace provides additional information, such as showing ringing after motion has stopped. In this case, it didn't.

CaiJonas

@dc42 thank you very much for all this information.

I have some questions:

In the M956 command, to wich value should I set the F parameter. To the speed of the outer wall in the slicer settings?

Unfortunately I do not see any improvement by using M593.

How should I set this up?

Is it possible to eleminate more than one frequency?

Does it make sense to mount the acceleration sensor to one axis carrier?

Best Regards,
Cai

CaiJonas

@dc42 just want to let you know that I had good results by eleminating 30Hz. By using M593 P1 F30.

Is it possible to eleminate more than one frequenzy? E.g. M593 P1 F30 F60?

Best Regards.

uk25

@caijonas said in Acclerometer on a Delta printer:

Is it possible to eleminate more than one frequenzy? E.g. M593 P1 F30 F60?

Also interested in that... is it even possible, and will it be in 3.4 firmware?