Posts made by droftarts

@magnets99 to calculate accurate timings out of the slicer, it will need to know the speeds and accelerations the machine is capable of. Some slicers then confusingly override the machine settings by emitting these in the exported Gcode. You should be able to turn this off, and use the slicer machine settings just to calculate times. It gets inaccurate anyway as soon as you start setting pressure advance and input shaping.

I think you’re running a Volcano hot end, and because of the larger melt zone, they do tend to string and blob more, particularly on small/slow prints and features.

Ian

@magnets99 said in The Plastic Sturgeon (Watercooled franken printer):

I'm using orca slicer

Yes, I've pretty much switched from Prusaslicer to OrcaSlicer. OrcaSlicer also has some built-in calibration tests, but I haven't tried them myself. I also recommend the 'Califlower' test by Vector3D, though unfortunately it'd gone up in price recently with a new version.

a bit frustrating that i have a box of ADXL345's but it doesn't support that one.

The accelerometers supported by RRF are the LIS2DW12, LIS3DH and LIS3DSH. They have a fast readout mode, which allows us to read the acceleration data without getting the artefacts in it that we saw with the ADXL345 in testing (Klipper also suffers from this).

Ian

@Simonjb yes, the reason for this bug appeared after 3.5.4 was released. It’s been fixed, but there hasn’t been a release since. See the last couple of updates here https://github.com/Duet3D/DuetWebControl/activity

Ian

@jhanda I’m on my phone, so it’s difficult to spot errors in your config and macros. I can have a better look in the New Year. However, generally, it sounds like the Z offset of the BLTouch and the SZP are referencing the same Z0. You’re not the first to struggle with this, and I’m going to review the SZP instructions in the New Year: https://docs.duet3d.com/en/User_manual/Tuning/scanning_z_probe_calibration

Initially, you need to set Z0 with the BLTouch (assuming this sets the nozzle height correctly), then use this to calibrate the SZP Z offset. Note that:

• make sure that no heightmap is applied when homing Z with the BLTouch or calibrating the SZP. Add M561 and/or G29 S2 before the SZP calibration to make sure any mesh is disabled.
• SZP calibration changes with temperature. You will need to recalibrate it before scanning if the temperature changes. As you have a BLTouch, you can do this in a macro, and can be done before every scan.

Ian

@magnets99 I’d think you could do the same thing with a BLTouch for 3D shapes, as the sensitivity and repeatability is usually around 0.01mm. However, if it’s a 2D shape that you’re trying to get the outline of, like a PCB, I’d think keeping it simple and using a scanner, or taking a photo of it, would be simpler and faster.

Ian

@mrehorstdmd another thread talks about connection weirdness when using VPN, maybe it’s related to that?
https://forum.duet3d.com/topic/37190/duet-2-wifi-don-t-connect-to-android

Ian

@Superbrain8 Okay, sorry about that!

While there are a lot of changes between 3.5 and 3.6, the configs are generally unchanged. I can’t see anything obvious in you config.g, but I’m on my phone at the moment, so easy to miss things. Best to revert for now and I’ll test in the New Year.

Ian

@darylprice sending M997 from DWC should be sufficient to update the 6HC to the same firmware version as DSF. I think DSF will only auto update minor versions (ie 3.5.3 to 3.5.4) but not major versions (ie 3.4 to 3.5), but M997 should fix that. See https://docs.duet3d.com/User_manual/Reference/Gcodes#m997-perform-in-application-firmware-update

Ian

@magnets99 that’s pretty slow, but as ever, it’s more dependent on your slicer settings.

Printing speed is generally related to how fast your hotend can melt filament, which is expressed in cubic mm a second (mm^3). You can calculate this from layer height, wall width and speed, eg printing at 0.3mm layer height at 0.6mm wall width (ie nozzle width) at 100mm/s is 0.3 x 0.6 x 100 =18mm^3/s. A normal E3D V6 nozzle can extrude at about 10mm^3/s, a Volcano should be able to do 25mm^3/s. It’s also dependent on what material you’re printing, and the set temperature. Most hot end manufacturers have a rough figure for the throughput of their hot ends, and there are online reviews too. Some slicers also have built in flow testing you can run. RRF has also recently introduced heater temperature feed forward, which increases hot end temperature as speed increases.

Non-printing speeds (eg travel moves) can generally be faster, and are only limited by the speed the motors can cope with. Tuning speeds, accelerations and jerk will show the maximum that can be achieved.

That’s a very brief outline to the many happy hours of tuning you have ahead of you!

Ian

@Superbrain8 we’ll have to test this to work out if it’s a bug or ‘feature’, but unfortunately we’re unlikely to get to testing until after the New Year. It’s possible this has been fixed in 3.5.4; if you can upgrade and test with that it would be useful, and possibly with the 3.6.0-beta.2 if you’re up for it.

Interestingly, I asked @dc42 a while ago if we could scan with the SZP in the Y direction, rather than X. Scanning in X showed up a fault with my X axis, with a lump every 62mm, which turned out to be a flat spot on one of the wheels of the X axis.

It would be interesting to do the same in Y. Sounds like you might have found a workaround to do that!

Ian

@magnets99 Nice! Looks like an Ultimaker 2 or 3 extended frame, but the XY motors have been moved outside the chamber. I’d guess the intention was to do higher temperature materials, except with the BLTouch in there you’re limited to 80C ambient. I think the hot end is an E3D Volcano or similar, so it’s probably meant to print fast. However, the Zesty may be the limiting factor, as the cable drive can wind up and fail if there’s too much extrusion pressure. Swapping the extruder for one of the modern, light, direct drive options would probably make it reliable.

Ian

@SPAX Please try RRF 3.6.0-beta.2: https://github.com/Duet3D/RepRapFirmware/releases/tag/3.6.0-beta.2

Ian

@jensus11 you can define points in the mesh for the BLTouch to avoid, see the notes on G29 S4 here: https://docs.duet3d.com/User_manual/Reference/Gcodes#notes-8

Ian

@yoshimitsuspeed said in Automated and 3D printed ski and snowboard Ptex/HDPE repair:

@droftarts

Welp can't post links so here it my response with them removed.

You need a reputation of at least two to post links. It stops the spam posts.

That's BS. The drip method is very far from ideal in many ways.

I never said it was the best way, just how I remember doing it as a DiYer. I said a glue gun method would be better. I don’t know what repair shops currently use.

You can generally get quite a few grinds before it's just too thin and or compromised. But at some point any board or skis reaches the end of it's practical lifespan.

I still don’t like them! If I still had my own skis, I’d want them to last as long as possible. When hiring skis, you’re more likely to damage the ski if the edge is thinner, which causes wasted time and usually an insurance claim.

Ian

@Maurits it’s a bug, has been reported and fixed in 3.6 betas, but has not been back ported to 3.5. See https://github.com/Duet3D/RepRapFirmware/issues/1058
Use 3.6-beta.2 https://github.com/Duet3D/RepRapFirmware/releases/tag/3.6.0-beta.2

Ian

droftarts created this issue in Duet3D/RepRapFirmware

closed [bug]: RRF 3.5.3 - Extruder drives on mainboard-as-expansion-board not moving #1058

@Chriss For an explanation of 'HRP' (Head Reference Point) see https://docs.duet3d.com/en/User_manual/Tuning/Defining_tool_and_Z_probe_offsets

Ian

@dogboy said in 5 Phase steppers (0.72 Deg.) using a SCARA system and Duet 6XD.:

My next question, is how can I use the M92 command to set the steps/degree if one degree of rotation on the joint does not give a whole number of steps?

Steps per mm don't need to be integers, you can use decimal points. See the example in https://docs.duet3d.com/User_manual/Machine_configuration/Configuration_SCARA#example

M92 X28.6738351 Y28.6738351 Z2560

Ian

@RodrigoRMaraujo That one won't work with the Maestro, as it has addressable RGB LEDs, which the Maestro doesn't support (well, it does in 3.5, but there's a timing issue that hasn't been fixed). See https://docs.duet3d.com/en/User_manual/Connecting_hardware/Display_12864 for supported displays. Basically, you need a display with a ST7920 controller chip, like this https://www.aliexpress.com/item/1874961796.html

Ian

@ComedianTF2 The M569 command shouldn't enable the drivers, only configure them. Only sending M17, or sending a motor move, should enable the driver. You should be able to send M569 commands as often as you like after running config.g, and the driver should not enable. At least, that's what happens on the Mini 5+ I have here. Unfortunately I don't have a 6XD to test. You can send M569 P0 to check how the each driver output on the 6XD is configured.

How have you wired the stepper driver? I assume you have wired the EN+ from the driver to +5V on the 6XD, and EN- to D0 EN- on the 6XD, as shown here https://docs.duet3d.com/Duet3D_hardware/Duet_3_family/Duet_3_Mainboard_6XD_Hardware_Overview#connecting-external-motor-drivers. How is the 'Driver Enable Polarity' (En_Pol) jumper set?

In the table, it says

Drive initial time 2 - - s

The 'Drive initial time' of 2 seconds might be the reason that enable appears to be delayed. Try sending M17, wait 2 seconds, then a move. The manual also says:

The 4-28V signal level is used for protection of pulse, direction and enable.

It's possible that it's not reliably picking up the enable signal, which is 5V from the 6XD. Check your wiring. If you disconnect the enable wires, the driver should be enable at power on, and be enabled the whole time, which might be easier.

Ian

@Egon-Net I've checked with @dc42, and yes, there is a new version of PanelDue firmware coming, but it hasn't been released yet. We are also not recommending anyone to use beta2+4 any more. Unless you have a specific issue that is resolved with one of the beta.2+ releases, it's probably best to revert to an earlier release.

Ian