Problems with HE280 installation on Max V2

[AlanZ]

I added an HE280 to my Max V2, and am having an issue with the accelerometer
This video shows how a G29 moves the hot end to the X tower, crashes into (and slides off of) the bed, then repeats.

https://www.youtube.com/watch?v=8LUDLukC1_k

I had previously manually set the Z zero position, and did a test print, so I should not be that far out of whack for Z height.

I THINK that I have it wired correctly, so before I open the machine back up again, I want to see if there is something that needs to be changed in the configuration.

I am using the Dev build from Feb 13 2017 with the config items listed below.
I spliced most of the new wire whip (premade by seemecnc) into the wires on my previous wire whip, and ran the few new wires to the base.
The old whip is coiled up at the to of the printer, with nothing connected.
I have hooked up a second extruder, but did not set it in the configuration yet.

It's possible/likely that I made an error in the wiring, but want to eliminate the software side of it before going to the hardware.

Anyone see this before, and have an antidote?
Your thoughts are appreciated on where to look next.

[AlanZ]

/*
Printer Model List as used throughout this firmware
Orion = 1
Rostock Max V2 = 2
ERIS = 3
Rostock MAX v3 = 5
Hacker H2 = 6
*/
// ### Define your Printer Model here! ###
#define PRINTER 2

// SeeMeCNC Bowden w/PEEK barrel = 1
// HE240 on ERIS w/accel probe = 2
// HE280 w/accel probe = 3
#define HOTEND 3

// ### Number of active extruders
// 1 is standard, 2 is with the Y coupler for dual filament input
#define NUM_EXTRUDER 1

// ### CLONE settings ###
// 0 is none, for just a standard single extruder/nozzle setup
// 1 is dual input filament setups - this is for the pins.h file to configure the second input reading
// 2 NOT USED YET, but will be for dual independent hotends/nozzles/extruders
#define CLONE 0

// ### Define your motherboard here! ###
// 301 = RAMBo 302 = MINI RAMBo
#define MOTHERBOARD 301

// ##### Older Orions w/ATX had Y inverted and NEW PSU on orions needs opposite ###
// 1 = ATX on older machines 2 = Rail style PSU on newer machines ############################
#define POWER_SUPPLY 2

[Dale Eason]

That looks like the printer is going too far toward the X tower and goes off the edge of the build plate. Where probes should go is set by the EEProm values:

Z-Probe X1,X2,X3 and Z-Probe Y1,Y2,Y3.

They are specified in the configuration.h. But if you did not clear the EEPROM before uploading the new firmware they may not be set right. Use matterControl to read the EEPROM to see what the values are. On My upgraded V2 X1 is 100.565 and Y1 is -71.340.

If it had not been going too far off the bed I would have said you had one of two other problems I have seen people have. Z Max length is too big compared to your actual value or the Accelerometer is not working as it should. There have been discussions of each of these issues in both trouble shooting, V2 and V3 threads. But if you are still stuck and are sure your x and y EEPROM values are good then I can help further.

Dale

[AlanZ]

I did clear the EEPROM prior to uploading the new build, but I will try it again.

It actually hits the build plate, gouges into the tape, and tilts so it slides off the build plate.

[Dale Eason]

Ok, then it is most likely the accelerometer may not be working right.

You can use the g-code terminal and M261 command to diagnose it.
Look for the int pin message and see if change from high to low. First use the m261 command and look at the value of int pin. Then thump the hot end with your finger and issue another m261 command it should not change value. The thrid m261 command without a thump should send it back to the initial value.

If it does not change check your wiring of the Int pin signal. It is shared with the hot end thermistor. Make sure you have the polarity of the thermistor wiring correct.

[AlanZ]

Dale,

My current EEPROM reads
x1 -123.565 (note the negative vs positive in x2... is this indicative of reversed wires?)
x2 123.565
x3 0
y1 -71.34
y2 -71.34
y3 142.68

[AlanZ]

M261 results. The INT Pin value changed from 1 to 0 then back to 1
pre-thump
<-INT PIN: 1
<-read reg 49 value: 21
<-read reg 53 value: 0
<-read reg 57 value: 0
<-read reg 45 value: 5

thump
send m261
post thump
<-INT PIN: 0
<-read reg 49 value: 101
<-read reg 53 value: 0
<-read reg 57 value: 0
<-read reg 45 value: 6

resend m261
post thump
<-INT PIN: 1
<-read reg 49 value: 21
<-read reg 53 value: 0
<-read reg 57 value: 0
<-read reg 45 value: 5

[Dale Eason]

Dale,

My current EEPROM reads
x1 -123.565 (note the negative vs positive in x2... is this indicative of reversed wires?)

No that is the correct value. The left edge of the build plate is in the negative direction. The back edge of the plate by the Z tower is positive Y direction.

So that all looks good. Hmm.

Dale

[Dale Eason]

What are your eeprom values for Max Z height?
Dale

[Dale Eason]

So it looks like the accererometer is working so it probably is not a problem. Other than Max Z height I running out of the usual causes. Default Max Z height on a normal V2 is 360 I think.

Dale

[AlanZ]

z max length is 360.906

So with the numbers so far, does it look like I've got it wired properly? Or could there still be a polarity issue?

[AlanZ]

Is it worth trying an earlier Dev build?

[Dale Eason]

z max length is 360.906

So with the numbers so far, does it look like I've got it wired properly? Or could there still be a polarity issue?

Those number look good. Wiring should be good too.

Most of the things I have seen to cause that problem are ok on your machine. Maybe you can post your EEPROM settings. You can get them from Mattercontrol.

You can also try send G28 (home) then G0 Z10 to move the head to 10 mm above the bed. Check that it is close to that value above the bed. Next send G30 X0 Y0 that will probe the bed at X0 and y0. You could try various values of X and Y to see where it works or not. But always make sure the G30 is done from 10 to 20 mm height.

Dale

[Dale Eason]

z max length is 360.906

So with the numbers so far, does it look like I've got it wired properly? Or could there still be a polarity issue?

Wait you said "360.906" well the default value when the EEPROM is cleared and firmware uploaded is 350. It is set in the configuration.h file. At least that is the value in the December 09 version of software I use. I have not looked at the latest releases.

Dale

[AlanZ]

There is an old saying "No good deed goes unpunished"

After clearing the eeprom again, and uploading the firmware again, I still had the same problem.

However, I made one more change that solved the problem.
I replaced the GeckoTech build surface (that is held by magnets on a GeckoTech heat spreader) with the stock glass build surface (clipped to the heat spreader)... and now it calibrates properly.
If I swap build surfaces, the problem returns.

I don't know if there is something about the GeckoTech surface (though I had tape on it), or its resiliancy, or interplay with the magnets that's causing the issue... but switching to the glass plate lets the autocalibration proceed without incident.

After some testing, I will likely switch back to the GeckoTech for printing... but for now, with this new hot end, and dual extruders, I will try to minimize the variables.

At least it's nice to know that my wiring was ok.

[Dale Eason]

That's great. Glad you finally figured it out. I don't know why except that the accelerometer must not get the desired acceleration from that combination of build plate. That value is configurable in the configuration.h file I think. You could play around with it. It might also be in the EEPROM settings. I can not recall.

Dale

[AlanZ]

Am about 2 hours into a 5 hour print of a two color tree frog. Though it needs some tweaking, it's looking good.