ReplicatorG + TeaCup FW + MK7 + custom machine

I'm having some problems getting a custom machine up and running. Using the ReplicatorG control panel, I can accurately move the extruder around in 3D space, control the heater temperature and extrude. However, I am having a bunch of problems calibrating the MK7.

I thought it would be easy to find STEPS_PER_M_E, MAXIMUM_FEEDRATE_E, and E_STARTSTOP_STEPS for a MK7 and I'd be printing. However, I haven't been able to find a definitive STEPS_PER_M_E for a MK7 (and the microstepping that goes along with it). Right now I am using:

#define STEPS_PER_M_E (50235/4) //!< MK7 defaults. (with quarter stepping)

I have my ReplicatorG machine.xml to match
<!-- stepspermm is incoming filament length, 127 is ca. 4 RPM, 1600 ca. 50 RPM -->


When I generate Gcode using ReplicatorG 0034 beta (need to use the beta or my temp sensor doesn't work), I get a bunch of warnings.
"You're moving too fast! G1 X0.39 Y0.4 Z0.17 F3300.0 Turns at least one axis faster than it's max speed." I don't understand what needs to be changed to get Gcode with slower feedrates???

If I throw caution to the wind and proceed, I get either get too much or virtually no PLA coming out depending on what I print.
When I print the cube, I need to use the real time control to print about 10% feedrate. When I try to print the snake, the extruder moves around but I don't get any PLA out.


Any suggestions on the proper MK7 settings and how to generate Gcode without moving too fast would be appreciated.

Quote

"You're moving too fast! G1 X0.39 Y0.4 Z0.17 F3300.0 Turns at least one axis faster than it's max speed."

Nice! This ReplicatorG thinks it knows what the machine can do. Apparently you have to disable this limit somehow in this application. Teacup won't spit an error, it will simply reduce the speed to what the machine can do and go ahead.

That said, speeds and limits in firmware vs. sending host are independent of each other. Teacup doesn't know which application sends the G-code and ReplicatorG doesn't know about firmware settings.The only connection between them is the sent G-code, so to find out what's going on, you should look at this G-code, first. Advanced RepRappers calibrate by sending G-code manually.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Thanks for the reply. I think I understand how Teacup simply takes gcode, but I'm not clear on how information in ReplicatorG machine.xml changes the generated gcode. (And the requirements on machine.xml matching the firmware.) After a bunch of frustration, I think what I need is to edit a skeinforge profile to match my machine. My XYZ already look good, so I think it's just a matter of getting the extruder feed rate and maybe some limits correct. Since I am using a MK7, I thought the settings would be readily available.

ReplicatorG seems a bit buggy? (or maybe it's just the skeinforge profiles that come bundled with various versions?) When I updated form 0029 to 0034, I lost my extruder temperatures in the control panel. My extruder temperature does work with 0034 beta. However when I edit the Skeinforge settings from ReplicatorG, I have to use the task manager to exit ReplicatorG since it hangs.

What's the normal procedure advanced reprappers use? Do you run Skeinforge on it's own (create a custom profile) and just use ReplicatorG to send the gcode?

This morning, I noticed there is a new YouTube video on setting the Skeinforge extruder settings. I will measure the filament and adjust my firmware steps and give it another try tonight.

BTW, nice job on the Teacup firmware. I didn't have any trouble setting up two basic machines with 328 and mega boards. The majority of what I have done so far is play with the ReplicatorG control panel.

Looking forward to the day when my 12 year old daughter can take one of her Google sketchup designs, save the .stl, and print using Replicator.

Try using Slic3r to slice the files and Pronterface to send them to the printer. Eliminate RepG and Skeinforge completely and you will be so much happier.

P.s. RepG's xml is to set things in the simulation not the actual printer, all of the settings are done in Skeinforge if you still want to go that way. Also most people use Pronterface to send the gcodes even if they are using RepG to slice them.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Thanks for the advice, Pronterface and Slic3r seem to make more sense and I'm making progress. The first few layers come out ok but I must be extruding too much PLA and the nozzle starts plowing through. The default slic3r config was set up for 3mm filament and I'm using 1.8mm. So it looks like I need to tackle more of the settings.

I have the Printer G-code flavor set to Teacup, but I still have to edit out the G92 E0 ; reset extrusion distance to get it to print??? Should I be using relative E distances?

Yes you need to use relative extrusion. That would be why you have to much plastic being extruded.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Quote

I have the Printer G-code flavor set to Teacup, but I still have to edit out the G92 E0

Watch out for E_ABSOLUTE in your config.h. There's a description of what it does. You might also want to try the Gen7 branch, which supports M82/M83 already (and comes with a new description of E_ABSOLUTE in config.default.h). The Gen7 branch is no longer Gen7 specific and works on all sorts of electronics.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Thanks, getting absolute/relative extrusion correct straightened out the excess extrusion. Just a note, the config.h has:

/** \def E_ABSOLUTE
Some G-Code creators produce relative length commands for the extruder, others absolute ones. G-Code using absolute lengths can be recognized when there are G92 E0 commands from time to time. If you have G92 E0 in your G-Code, define this flag.
*/
#define E_ABSOLUTE

Using slicr3d, I always get a G92 E0 command in th3 first few lines of G code. You may want to change the comments to looking at the E values to determine relative vs absolute.

My printing progress is now limited by a translation and a slight rotation that occurs during the infill??? My skirt and perimeters look great (10mm square .stl) but when the infill at 45 degrees starts it has a slight translation (2mm) and rotation that carries over into the next layer. My XY movement seems correct when I test absolute movement but either the build table is shifting (spring mounted) or I'm dropping steps during this motion. I tried slowing down the speed last night and it didn't seem to help.

Tonight, I will measure origin before and after build to see if the table or steppers are what's shifting. My machine has an extruder mounted on an XY system with a build table that moves in Z. So if my build table shifts it could explain the problem (but after a few layers the shift is so large I don't see how the table could shift this much.)

Uhm, what is "straightened", then? Absolute or relative? So far I thought slic3r does use absolute E commands, so the config.h comment would be correct.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

slic3r 0.7.1 has a "Use relative E distances" setting. What I meant was when this box is checked, slic4r still generates the G92 E0. From the comment in config.h I thought I wouldn't see the G92 command with absolute extrusion distances. I guess with absolute you will only see this once while with relative you will see many???

I'm currently using relative (unchecked box in slic3r and commented out E_ABSOLUTE in config.h). It also worked with box checked and E_ABSOLUTE defined. I was just suggesting that the comment in config.h wasn't clear (at least to a rookie like myself.).

So my extrusion is calibrated and working fine. Now I have an accumulating error in my xy and probably my z. My machine is constructed from a salvaged Xray copy machine from the 1980's. I replaced the cabinet of drive circuitry with modern stepper drivers. Old school construction with 200 step Superior Electric steppers and 3/4" rods with 28" ball screws. 1"/rev linear travel XY, 0.2" linear travel/rev Z. I know it's way overkill for printing, but my first project was to print some plastic for my kids before mounting a router.

Is it normal for error to accumulate with english travel when printing metric dimensions? I see approx 1mm of error after 20 back and forth diagonal moves:
G1 X-10.0 Y-10.0
G1 X+10.0 Y+10.0

I tried slowing down the speed and changed acceleration from 1000 to 10 and speed/acceleration doesn't seem to matter. I have plenty of current and my voltage is stable so I don't think I am missing steps. I think it's just numerical roundoff accumulating. This also makes sense that straight lines like the skirt and profiles would look ok but infill would accumulate errors.

; generated by Slic3r 0.7.1 on 2012-04-18 at 12:38:21
G92 E0 ; reset extrusion distance
G1 X-13.940 Y-14.120 F540.000 E0.01257
G1 X-13.560 Y-14.460 E0.02518
G1 X-12.940 Y-14.900 E0.03754

; Absolute box checked. generated by Slic3r 0.7.1 on 2012-04-18 at 12:40:20
G92 E0 ; reset extrusion distance
G1 X-13.940 Y-14.120 F540.000 E2.01257
G1 X-13.560 Y-14.460 E2.03775
G1 X-12.940 Y-14.900 E2.07529
G1 X-12.720 Y-15.030 E2.08791

G92 E0 is pointless in relative mode. Relative means, the distance is reset for every single move. I'll ask the slic3r developer to remove this to reduce confusion.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

See [github.com]

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY