HE280 Calibration on v2

[Turpinator]

So after my original hot end going out/having issues, i upgraded to the he280 fancy stuff... Got it all installed and fuse replaced (blew one during disassembly, whoops). And it works. It prints things like normal, but the calibration does not seem to work.

When attempting to run the calibration gocde file, it homes, then goes down very fast, about 4" from the bed, it quickly reverses direction, homes itself, and ends the print job.
https://www.youtube.com/watch?v=hdAk_eIKuks


And if i manually run the first gcode command (G69 S2), it homes, goes ~2" from the bed, then moves directly to the right (on the X axis) near the edge of the bed. It then slooooowly moves down towards the bed until it touches it, and then keeps going. Eventually it stops, homes, then retries, but not before skipping teeth on the belt (or steps on the motors).
Here i stop it before it keeps forcing itself into the bed.
https://www.youtube.com/watch?v=4UE-ldSoofw


Thoughts? I2C plug is plugged in, firmware updated (tried again just to be sure), and ive got no idea.

[geneb]

Did you upload the new firmware as well?

g.

[Turpinator]

Yup. (noted right at the end)

I did set the printer to '2' (v2), but went though and updated any values for the new hotend from v3 that I saw (mainly temps). Should I really just set this to 5/v3? Id have to tweak the current and axis inversions, but that's all I can think of.

I have yet to try running the cal from the display though, maybe ill try that this afternoon. Also, maybe ill try setting the firmware to printer 5/v3 and trying again? But... if something wasn't set up correctly (ie accelerometer not enabled in the firmware), the calibration menu shouldn't have shown up, right?

[Turpinator]

However, I did not clear the eeprom or anything like that. Ill also try that.

[geneb]

Ah, there's the issue. If you're using the HE280 hot end, you've got to tell the firmware it's running on a V3. You'll need to invert one of the motor axes as well - Y I think, but I'm not sure. Load the firmware up, hit home and be ready to turn off the power. Which ever one moves the wrong way is the one you need to invert.

g.

[Turpinator]

Awesome. Thanks. Ill change it and report back.

Thoughts on why running the gcode file vs sending gcode commands yielded different results/actions? (fixed the second video link)

[Turpinator]

Ah, there's the issue. If you're using the HE280 hot end, you've got to tell the firmware it's running on a V3. You'll need to invert one of the motor axes as well - Y I think, but I'm not sure. Load the firmware up, hit home and be ready to turn off the power. Which ever one moves the wrong way is the one you need to invert.

g.

No dice.
It does the same/similar thing. When using the gcode file, it goes down fast, and then about 1" from the plate stops, homes, and then finishes the print.

Reuploaded firmware as v3, same thing.
Cleared eeprom and then reuploaded firmware as v3, same thing.

Also copies of the console/whatever after running the gcode print file.
http://www.hastebin.com/urerimexiz.hs
http://www.hastebin.com/kiqekejodo.hs



OOOOOOOO! I figured it out. Well. I know why its happening, just not whats causing that to happen...

Code: Select all

<- start
<-Transformation matrix: 1.000000 0.000000 0.000000 0.000000 1.000000 0.000000 0.000000 0.000000 1.000000
<-Free RAM:876
<-Autoretract:0
<-X:0.00 Y:0.00 Z:0.000 E:0.0000
<-SpeedMultiply:100
<-iis2dh accelerometer initializing...
<-accelerometer i2c recv error.
<-read reg 49 value: 0
<-read reg 32 value: 7
.
.
.

So something with the accelerometer not connecting/something. Its plugged in. I checked. A few times. Havent checked my wiring to be certain, but i was pretty meticulous when I did all of that.

[Turpinator]

On further inspection it looks like the temp reading is rather far off. I could barely extrude pla at 200C. Played with it and got it okay at 230, enough to make a test print. Didnt get a chance to actually check it, tc is mia.

[Turpinator]

Alrightly, so anyone have any idea whats going on/how to fix it? I upgraded only to downgrade.

Checked the hotend with my tc, reported ~135c when the display said 230. Of course, i cant get it in the exact same spot as the thermistor, because... well... the thermistor is there. Tc was in contact with the heating block.
Suuuuure. I can replace the thermistor, but then ive got to take it mostly apart, all for an arbitrary value. The thermister currently works at room temp, reporting pretty much the correct temp.

As for the actual calibration part... Theres an issue somewhere with the I2C port, board, or the 5 ft of wire that the I2C wires are...
Cant really shorten the wires. Dont have a spare board to check if its the accel board. And dont have a spare rambo laying around.

[rurwin]

There is a huge thermal resistance between two objects just touching each other. You need to appy some pressure and preferably some insulation around them.

Is there a good reason that nobody suggests IR gun-type thermometers? Maybe the active region is too large or the surface has the wrong emissivity. But otherwise they would be much better than a contact thermocouple.

[Turpinator]

Oh yeah. Measuring temp of the hotend with a Tc (or thermistor) just by pressing it to the surface isn't all that accurate, but it can at least give you a reference temperature. As for the thermal resistance... In a perfect world, where the Tc isn't giving off any heat, the resistance just slows the rate it hits temp. Hmm... I could... Get a mug of oil, submerge the heating block in there, let it stabilize temp, and then read the oil's temp. My Tc is an immersion probe, after all. My first thought was to use water, but.. that's just going to boil off.

As for the IR gun thermometers, most cheap ones have fixed emissivity settings and rely on the fact that /most/ things have a very high emissivity. The aluminum heater block, on the other hand, has a very low emissivity, probably in the range of 0.1


Either way, im not worried about the thermistor. If I know how wrong it is, I can correct for it. However, the accelerometer is substantially harder to correct for.