I'm looking to change the mounting position of the IR sensor. Currently, it isn't in line with the nozzle, and it is mounted on the x-carriage rather than the hotend. This has two adverse consequences: 1. Any change in the amount of head sag affects the relative heights of the nozzle and IR sensor. 2. Any movement between the hotend and the x-carriage (e.g. due to thermal expansion in the nozzlby dc42 - Ormerod
Here are some other possible reasons for Z homing and bed compensation not working well: 1. Running Z homing and bed compensation when cold, then heating the bed. Best to heat the bed, then home Z and run compensation. 2. Using the original alu tape. Tiny differences in the y-position when the sensor probes can make a big difference to the IR. Use white tape or white paper instead of alu tape.by dc42 - Ormerod
The problem with using a home-made non-switching PSU is that you will need a very large smoothing capacitor. With both heaters going, you would need about 25000uF to keep the ripple down to 6V. The steppers should be OK with this ripple, but the fan probably wouldn't be. A better way would be to run the two heaters from unsmoothed DC straight from the bridge rectifier, then use a diode and a smalby dc42 - Ormerod
QuoteIngeneo70 Using the USB cable to do printing atm. The quality is good, the only problem I have is that the 1st layer is sometimes a bit "stringy". I will make another post about this if this continues after I have tried using the SD card to print. If you are using the latest firmware then SD print and USB print should give the same results. I found it helpful to change the slic3r settings tby dc42 - Ormerod
The original IR sensor only works well when there is no significant sunlight or incandescent light falling on the white tapes. The modulated IR sensor fixes that. RRP will hopefully have them available within the next few weeks, or if you can do SMD soldering then you can modify the existing board as described in this thread . Before I had the modulated sensor, I was doing much the same thing asby dc42 - Ormerod
I can't see how a pure overhang would work, and the only times I have printed pieces with overhangs, they have come out stringy. So when I design parts, I stick to the 45 degrees rule. Bridges are a different matter, they can work OK as long as they are not too long.by dc42 - Ormerod
I've now updated my firmware build to 57f, which is available at . This version achieves a file upload speed of 1.96Mb/sec, almost as good as the USB upload speed. To get the best upload speed, disable debug (see 3 posts earlier) and if you are running a local web server, replace reprap.js by the one at . I have also attempted a fix for the issue that Bob reported in this one.by dc42 - Ormerod
Bob, it does indeed appear from the log that the upload code is sending a line about 1024 bytes long, when the buffer size reported by the firmware is only 800 bytes. The buffer size variable should also get set by the poll messages that are sent every second, so I think you should only have a problem if you try to print or upload within a very short time of connecting.. Nevertheless, it would beby dc42 - Ormerod
My latest binary is now 57e-dc42, available at . With this binary and some adjustments to timings in the .js file, I'm getting a file upload speed of 0.56Mb/min. This is still quite a lot slower than the 2Mb/min achieved over USB, so I'll take a further look at improving it when I have time. For the fastest file uploads, debugging must be off. Therefore, if you have M111 S1 in your config.g fileby dc42 - Ormerod
Quotebobtidey These updates are very welcome thanks. Making it easier to use all the time. On the file upload speed, I thought I saw an even bigger buffer on DC-42's firmware potentially allowing even more lines to be grouped together but it now seems to be reverted to the original and the web side allows up to 5. Was there no speed benefit above that or I guess there may have been some compatiby dc42 - Ormerod
Likewise, you need to replace COMx by the actual COM port that the Duet appears as after you have pressed the Erase button and then the Reset button. Use Device Manager to see which COM port number it is.by dc42 - Ormerod
QuoteGregL @dc42 I have picked up your ini files from 22nd jan 'Here are my standard slic3r files. The 429b one is filament, 2kb one is print, 930b is printer.' I updated the simple.ini and the slic3r.ini with the RepRap ones as the instructions suggested It is not clear to me which of the Slic3r files your new ones replace. Help! Greg Edit to add - I think I have found the locations to post tby dc42 - Ormerod
It sounds as though the PID parameters need adjusting. The firmware doesn't yet implement the M130 to M133 commands that allow them to be changed, but I could add them if you would be interested in trying different PID parameters.by dc42 - Ormerod
I'm not clear what you mean by "z=0.2" in the x=100 values. Do you mean that the bed is 0.2mm higher than it would be if it were flat? If so, then I too suggest you look at johneato's comments. I have a similar issue: clamping the glass in the corners gives it a slightly domed shape. I can't measure this accurately yet, but I estimate that for me the centre of the bed is about 0.1mm higher than tby dc42 - Ormerod
Thanks, Matt! May I suggest a few additions: 1. A "Home all" button. This is faster than homing all axes in sequence. 2. A "Run setbed.g" button. 3. Buttons to move Z in steps of 0.05mm without having to compose g-commands. For example, make the z-buttons 0.05, 0.5, 5 and 50mm instead of the current ones. This could be made a configuration option.by dc42 - Ormerod
My close-up eyesight isn't too good either, and I too use a strip of paper. If there is any PLA on the end of the nozzle, I heat it to 150C first. Still, it would be nice to put an LED or three on the hotend assembly, powered from the fan supply.by dc42 - Ormerod
I've implemented this in the firmware build at . I also made doing a G92 command set the appropriate "axis has been homed" flags. I'm not certain that enforcing Z=0 as the minimum height is a good idea, because it means that after homing Z, if you put the head over the bed at a point where the bed is lower, you can't drop the head any more unless you do e.g. a G92 Z1 first. Maybe I should allowby dc42 - Ormerod
QuoteTMD_RS436 QuotePaulHam Thanks John, that's cleared it up for me! Perhaps ill give it a go in the future! I see that you have had to bend the pin for the ir sensor too, I have a feeling mine will snap off ! Yeah I thought the same, I am so tempted in buying a 50pin IDC plug and connect that wire to pin 33 (I.e. the bent pin) Paul I put an extra connector at the bottom of the board for thiby dc42 - Ormerod
Thanks Ray, I just got the spring kit from Maplins.by dc42 - Ormerod
This needs a specification to start from. I'm thinking along the following lines: 1. Remember which axes have been homed. At power up or reset, none of them have been homed. 2. If z-homing is being done with an IR sensor, then if a command is given to home Z but either X or Y has not been homed, do not execute the command but report an error. Likewise, refuse a G31 command if X and Y have notby dc42 - Ormerod
Quotetracyb I have no problem with the homing buttons - but after the machine is homed the firmware should not allow you to perform a movement that would drive the motor out of limits. For example if your x is at 80mm a button press to decrease x by 100 should take it back to the zero and no further. It could be added into the firmware and configured for the bed size of your machine. I'll take aby dc42 - Ormerod
I've not tried changing the print speed, but I have increased the extrusion max speed and acceleration, using these commands: M201 X800 Y800 Z8 E350 ; set accelerations, reduce Z, increase E M203 X3000 Y3000 Z180 E2700 ; increase max extruder speed I did try increasing the max X and Y speeds, as high as 20000 at one time. Higher speeds seem OK but cause more vibration.by dc42 - Ormerod
I want one! How will you be attaching the front bearing to the bed frame?by dc42 - Ormerod
I get that when I disconnect the USB port without exiting the serial monitor first. I suspect a bug in the device driver. Exit Arduino IDE, reset the Ormerod, then load Arduino IDE and try connecting again.by dc42 - Ormerod
I use 4mm retraction, 198C for the first layer and 193C for the remaining layers. 205C and 200C were about right using the old firmware, which was overreading by about 7C @ around 200C. That was corrected at revision 0.53. For the bed, I am currently using 75C for the first layer and 65C after that, but I am printing direct on glass. For printing on Kapton, I was using 60C.by dc42 - Ormerod
QuoteGregL Try cannibalising a retractable ballpont pen I think it needs stronger springs that those.by dc42 - Ormerod
Quotejohneato The glass does deform and can be measured. The bed is mounted on springs but doesn't help because the glass is clamped to the aluminium/ PCB by the edge clamps which do exert quite a force. Maybe it's time to try 4mm glass then? I've had some cut and I may get a chance to try it today. I also have a selection of different Swiss clips, so I'll see which ones clamp with the least forby dc42 - Ormerod
I doubt that the aluminium by itself is stiff enough to bend the glass much. I suspect it is the alu and pcb sandwich together that is providing the stiffness, aided by the corner screws. Some people have used springs to support the weight of the bed in order to make levelling easier, and this might have the beneficial effect of clamping the alu less tightly to the pcb.by dc42 - Ormerod
Quotedmould With X axis at the physical end of its travel (end of the arm), my IR sensor is still 12 mm in from the edge of the glass - it cannot physically move any closer to the edge because the LU bearing is hitting the plastic part that holds the runner. So what? The difference between X=53 and X=230 (which are the X-points I use for bed calibration) is plenty to establish the tilt of the beby dc42 - Ormerod
I've fixed this in my latest build at . I'll submit a pull request when the other changes I have made in this commit (which are to allow Matt to improve the performance and functionality of his web interface) are tested and found to work properly.by dc42 - Ormerod