LPC port of RepRapFirmware

Quote
dc42

Quote
PB
Sinned,
thank you for helping to verify that everything were connected right.
Found that at F100000-50000 some part of image appeared and was becoming increasingly in sync
M918 P1 F30000 helped completely (but screen redraw is slow)
Is there some way to get same at faster redraw or it is not worth efforts?

I believe the problem is that the SBase drives the 12864 display with 3.3V signals, but the minimum input high voltage specified by the datasheet for the ST7920 chip on the display when it is powered from 5V is 3.5V. That's why it only works at low data rates. They should have included level shifters on that board, like we did on the Duet Maestro.

Seems Sbase do have level shifters i.e. SN74LVC1T45DCK at 0.15, 0.16, 0.18
I can not guess why it has problem with RRD displays.

Quote
PB
Seems Sbase do have level shifters i.e. SN74LVC1T45DCK at 0.15, 0.16, 0.18 [attachment 112060 Untitled.jpg]
I can not guess why it has problem with RRD displays.

May it be incompatibility like inverted SCK levels (wrong CPOL) or time shifted MOSI to SCK (wrong CPHA)?

Quote
PB
heat.tempSensePins = {0.24, 0.23, 0.25};
heat.heatOnPins = {2.5, 2.7, 2.6};
fan.pins = {2.4, 2.6};

M307 H0 B0 S1.00
M305 P0 T100000 B4981 C1.632834e-7 R4700
M143 H0 S160
M301 H0 P16.2427 I0.6705 D98.3699 B30

M307 H1 B0 S1.00
M305 P1 T100000 B4981 C1.632834e-7 R4700
M143 H1 S280
M301 H1 P240.7739 I18.5211 D782.5151 B30

M307 H2 B1 A-1 C-1 D-1 ;By my idea it should release 2.6 from heatOnPins for use it as PSU fan
M305 P2 T10000 B3435 R4700 ;PSU thermistor
M143 H2 S70

M106 P0 S0 I0 F500 H-1 ;hotend fan
M106 P1 S0.5 I0 F500 H2 T40 ;PSU fan

Is there nice way to turn on PSU fan when thermistor is over 40C and halt system if it is over 70C?

If you don't want to use 2.6 as a heater, then just remove it from the heatOnPins array in board.txt.

As for setting up the PSU thermostatic fan, just guessing here (never tried that before), but maybe its not yet monitoring heater 2 temps? So have you tried creating another "tool" with Heater2 (but without any drives etc attached)? Perhaps something like M563 P2 H2 might work. I also think that command M106 will only turn on at 50% when triggered?, if you want it running full speed i think you need S1.

Quote
sinned
it looks like its using different pins for SPI functions on EXP2 header. You might have to jumper the SPI pins for SCK, MISO and MOSI: 0.15, 0.17, 0.18. Even thought it worked on Smoothiware, the LPC port only uses the hardware SPI pins, and the others are SW driven. sdavi might be able to help here.

Looking at that pinout, seems they have wired up the external SDCard to use the same interface as the internal card (SSP1). Currently in the port, i've only allocated 2 cards and assumed that the external SDCard will be on the SSP0, so the external SDCard on that board won't be working with that pin assignment at the moment. I might make that a configurable option in a future version.

Quote
PB

Quote
PB
Seems Sbase do have level shifters i.e. SN74LVC1T45DCK at 0.15, 0.16, 0.18 [attachment 112060 Untitled.jpg]
I can not guess why it has problem with RRD displays.

May it be incompatibility like inverted SCK levels (wrong CPOL) or time shifted MOSI to SCK (wrong CPHA)?

BTW Sbase 1.2 schematics available [github.com]
The only thing I can confirm about Sbase 1.3 that level shifters IC at same places are present in 1.3 board.

Quote
PB

Quote
PB

Quote
PB
Seems Sbase do have level shifters i.e. SN74LVC1T45DCK at 0.15, 0.16, 0.18 [attachment 112060 Untitled.jpg]
I can not guess why it has problem with RRD displays.

May it be incompatibility like inverted SCK levels (wrong CPOL) or time shifted MOSI to SCK (wrong CPHA)?

BTW Sbase 1.2 schematics available [github.com]
The only thing I can confirm about Sbase 1.3 that level shifters IC at same places are present in 1.3 board.

Schematics for MKS_SBASE_v1.3_002

Quote
PB
BTW Sbase 1.2 schematics available [github.com]

Which version are you using?

The pinouts for EXP 1 & 2 appear different between versions.

Quote
sdavi
If you don't want to use 2.6 as a heater, then just remove it from the heatOnPins array in board.txt.

As for setting up the PSU thermostatic fan, just guessing here (never tried that before), but maybe its not yet monitoring heater 2 temps? So have you tried creating another "tool" with Heater2 (but without any drives etc attached)? Perhaps something like M563 P2 H2 might work. I also think that command M106 will only turn on at 50% when triggered?, if you want it running full speed i think you need S1.

Thank you. Following configuration is working for PSU cooling fan thermistor control:

heat.heatOnPins = {2.5, 2.7}
...
M307 H2 B1
M305 P2 T10000 B3435 R4700
M143 H2 S70
M106 P1 S0.5 I0 F500 H2 T40
M563 P1 F1 H2
G10 P1 R0 S0

And especially helped that I stopped specifying A-1 C-1 D-1 in M307 which broke everything.

BTW seemingly working RRD display at F100000 and less somehow broke ethernet and made web connection intermittent. I decided not turning on display and it immediately completely helped with network. I consider it strange thing but as of now decided not to go deep and work without display.

Glad you got the themo controlled fans going!

Which menus are you using? There is not enough memory to utilize both networking and a full set of menus.

On my ReArm, I have networking and one minimal menu for Off/On and temp control & display.

Otherwise, its a full set of menus, no networking and using a host like Octoprint.

With the Sbase, just use a TFT32 display, cheap and less hassle.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Quote
MKSA
With the Sbase, just use a TFT32 display, cheap and less hassle.

How does that connect to the mainboard?

When I try auto calibration G32 everything seems to proceed fine, homing and probing all points without any error messages but resulting M665 M666 parameters are completely same like before calibration.

config.g
M208 Z0 S1
M574 X2 Y2 Z2 S0
M558 P5 R0 H5 F300 T18000
M558 H15
G31 P500 X0 Y0 Z0
M556 S50 X0 Y0 Z0
M557 R95 S20

bed.g
M561
G28
G30 P0 X0 Y94.9 H0 Z-99999
G30 P1 X55.78 Y76.78 H0 Z-99999
G30 P2 X90.26 Y29.33 H0 Z-99999
G30 P3 X90.26 Y-29.33 H0 Z-99999
G30 P4 X55.78 Y-76.78 H0 Z-99999
G30 P5 X0 Y-94.9 H0 Z-99999
G30 P6 X-55.78 Y-76.78 H0 Z-99999
G30 P7 X-90.26 Y-29.33 H0 Z-99999
G30 P8 X-90.26 Y29.33 H0 Z-99999
G30 P9 X-55.78 Y76.78 H0 Z-99999
G30 P10 X0 Y47.4 H0 Z-99999
G30 P11 X41.05 Y23.7 H0 Z-99999
G30 P12 X41.05 Y-23.7 H0 Z-99999
G30 P13 X0 Y-47.4 H0 Z-99999
G30 P14 X-41.05 Y-23.7 H0 Z-99999
G30 P15 X-41.05 Y23.7 H0 Z-99999
G30 P16 X0 Y0 H0 Z-99999 S6

What can be the reason why calibration doesn't update geometry?

Edited 3 time(s). Last edit at 07/16/2019 07:00PM by PB.

Quote
sinned

Quote
MKSA
With the Sbase, just use a TFT32 display, cheap and less hassle.

How does that connect to the mainboard?

Discussed earlier in this thread and documented in MKS doc as they produce both.

Time to drop support of outdated stuff. This thread illustrates the trouble of doing so.

Edited 1 time(s). Last edit at 07/17/2019 01:30AM by MKSA.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Quote
PB
When I try auto calibration G32 everything seems to proceed fine, homing and probing all points without any error messages but resulting M665 M666 parameters are completely same like before calibration.

config.g
M208 Z0 S1
M574 X2 Y2 Z2 S0
M558 P5 R0 H5 F300 T18000
M558 H15
G31 P500 X0 Y0 Z0
M556 S50 X0 Y0 Z0
M557 R95 S20

bed.g
M561
G28
G30 P0 X0 Y94.9 H0 Z-99999
G30 P1 X55.78 Y76.78 H0 Z-99999
G30 P2 X90.26 Y29.33 H0 Z-99999
G30 P3 X90.26 Y-29.33 H0 Z-99999
G30 P4 X55.78 Y-76.78 H0 Z-99999
G30 P5 X0 Y-94.9 H0 Z-99999
G30 P6 X-55.78 Y-76.78 H0 Z-99999
G30 P7 X-90.26 Y-29.33 H0 Z-99999
G30 P8 X-90.26 Y29.33 H0 Z-99999
G30 P9 X-55.78 Y76.78 H0 Z-99999
G30 P10 X0 Y47.4 H0 Z-99999
G30 P11 X41.05 Y23.7 H0 Z-99999
G30 P12 X41.05 Y-23.7 H0 Z-99999
G30 P13 X0 Y-47.4 H0 Z-99999
G30 P14 X-41.05 Y-23.7 H0 Z-99999
G30 P15 X-41.05 Y23.7 H0 Z-99999
G30 P16 X0 Y0 H0 Z-99999 S6

What can be the reason why calibration doesn't update geometry?

Have you tried sending M665 and M666 without parameters after running G32, to see if they have changed? Then you can run M500 to save the new values to config-override.g.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42

Have you tried sending M665 and M666 without parameters after running G32, to see if they have changed? Then you can run M500 to save the new values to config-override.g.

Yes I run M665 and M666 without parameters before and after G32 and all values reported are completely same. I even tried to put 1mm thick sheet on half of bed to make it feel differently inclined for probe and no error messages and no changes in values either. Also tried 6,7,8,9 factor calibration and 16 points specified above.

Completely puzzled about that.

Edited 2 time(s). Last edit at 07/17/2019 04:02AM by PB.

Quote
PB

Quote
dc42

Have you tried sending M665 and M666 without parameters after running G32, to see if they have changed? Then you can run M500 to save the new values to config-override.g.

Yes I run M665 and M666 without parameters before and after G32 and all values reported are completely same. I even tried to put 1mm thick sheet on half of bed to make it feel differently inclined for probe and no error messages and no changes in values either. Also tried 6,7,8,9 factor calibration and 16 points specified above.

Completely puzzled about that.

Can you also double check that the controller doesn't reset after the last probe point? Just to rule out the possibility of running out of stack causing a reset. Can you try a 12 point calibration instead? If that works, then I can increase the main task stack size if it needs a bit more space.

Edited 1 time(s). Last edit at 07/17/2019 06:03AM by sdavi.

Quote
sdavi
Can you try a 12 point calibration instead? If that works, then I can increase the main task stack size if it needs a bit more space.

Yes, decreasing number of calibration points to 16 works. In fact it was 17 when it was not working.

Thank you!

After 2 days of silly problems I use this cool firmware!

1. Noticed that DRV8825 steppers which are quite noisy in smoothie, are much less noisy in RRF. Probably is is because in RRF I set recommended pulse delays M569 P0 S0 T1.9:1.9:0.65:0.65 according to DRV8825 datasheet which option is absent in smoothie. Furthermore in low power idle mode (M906 X1000.00 Y1000.00 Z1000.00 E1000.00 I30) it is almost quiet, I notice only hissing of old style TV-set like 10000Hz. Previously I wanted to replace stepper drivers with quiet TMC but now it is not so necessary. Big thanks to developers of RRF and LPC port!

2. Tried to calibrate pressure advance and get final calibration like this . High speed 75 mm/s, low 30 mm/s, pressure advance 1.0-2.5 s (which is big!). Optimum exactly in between: 1.66s. PETG 1.75, 230C, bowden of around 400 mm, volcano hotend 0.4 nozzle, 0.5 print width, 0.2 print height. With that big advance infill lines use a lot of backward extruder moves at their ends, which slows down print significantly if extruder jerk is small, and makes extruder to sound strangely when extruder jerk is 1800-3600mm/min. Maybe I will be happy to use advance only in shell moves and have it zero in infill. Overall experience is that big ('optimum by calibration') advance doesn't make much quality improvement comparing to printing with zero advance and high acceleration and jerk which probably make moves almost constant speed in shell and eliminates need of advance. Sure, ends of infill lines are bulged but I can see it only during print when infill is still not covered by roof. Ends of moves are coasted 0.07mm^3, ends of shell lines are wiped 0.2 mm.

M201 X8000.00 Y8000.00 Z8000.00 E8000.00
M566 X1800 Y1800 Z1800 E1800
M203 X36000.00 Y36000.00 Z36000.00 E1800 (recommended maximum volcano extrusion speed is around 45 mm^3/s which (d=1.75) is around 18 mm/s i.e. E1100 mm/min, but I set E1800 and have no problems as long as maximum extrusion move speed set by slicer is 75 mm/s i.e. extrusion speed 7.5 mm^3/s (h=0.2mm, width=0.5 mm))

3. On temperature plot I notice short but high spikes at all thermistors. Probably only one reading wrong. Not causing problems.

4. Only one question where it is better to place start and stop g-codes universal for each prints like M104 S0
M140 S0
G0 X0 Y0 Z220
M84
at the end. Now I insert it with slicer, but probably it is nice to do it in firmware. Should I place it in stop.g? Will run if job complete or only if stopped job manually? Is there something like start.g (before each print)?

Also I'd like to make G32 every time when printer is turned on but not between prints. If I place it in config.g it works but delays web-interface loading until G32 is complete which is long as I heat bed inside bed.g Now I do it manually.

Again big thanks to developers, I am very satisfied with this firmware.

Edited 2 time(s). Last edit at 07/18/2019 08:54AM by PB.

Quote
PB

3. On temperature plot I notice short but high spikes at all thermistors. Probably only one reading wrong. Not causing problems.
4. Only one question where it is better to place start and stop g-codes universal for each prints like M104 S0
M140 S0
G0 X0 Y0 Z220
M84
at the end. Now I insert it with slicer, but probably it is nice to do it in firmware. Should I place it in stop.g? Will run if job complete or only if stopped job manually? Is there something like start.g (before each print)?

Also I'd like to make G32 every time when printer is turned on but not between prints. If I place it in config.g it works but delays web-interface loading until G32 is complete which is long as I heat bed inside bed.g Now I do it manually.

3. So far it seems only some users of mkssbase 1.3 boards have reported those temperature fluctuations. As mentioned a few times in the thread, also make sure to use the thermistor analog gnd all the way to the thermistor. Are the temps are stable after running M84?

4. There is documentation about the macros here.

Currently you can write custom macros that are just a click away within DWC - you could write a macro to preheat the bed and perform a G32 and anything else you need to do. I think i saw somewhere that DC42 plans to implement conditional gcodes in the future which would be useful.

Edited 2 time(s). Last edit at 07/19/2019 01:06AM by sdavi.

I have tried to map out an example pinout for the RepRap Discount 12864 Full Graphic Display. From looking at the hardware and documentation that I have access to (RreArm, SKR 1.1, 1.3, a few GLCD's) and looking at other people's hardware experiences, the only 'standard' is that there is no standard!

What seems to be 'in the wild' is some controller boards have:
1) LCD+SD on SSP0,
2) LCD on some other pins for software SPI and SD on SSP0, and
3) LCD on SSP0 and SD on SSP1.

Throw in the fact some connectors have rotated keying slots, and it just gets more and more confusing.

Here is the pinmap that I have come up with - [i.imgur.com]



The caveats-

1. The SSP0 pins must be wired as shown. ORANGE notes in the tables.
2. The other pins, which should be assignable pins, have the BLUE notes. These are the other pins on the EXP connectors.


The fact that the display worked under Marlin or Smoothie is not a guarantee that it will work without modification or adjustment with RRF. It will probably work as long as the pin assignments are correct.

DISCLAIMER:I have done my best to verify accuracy,but you should check it yourself and use at your own risk. I don't claim to be an expert, nor do I try to play one on TV. I don't have a youtube channel either so I lack internets credibility too. YMMV.

Quote
sdavi

Quote
PB
3. On temperature plot I notice short but high spikes at all thermistors. Probably only one reading wrong. Not causing problems.

3. So far it seems only some users of mkssbase 1.3 boards have reported those temperature fluctuations. As mentioned a few times in the thread, also make sure to use the thermistor analog gnd all the way to the thermistor. Are the temps are stable after running M84?

When stepper drivers are off and heaters are off fluctuations are absent. If one heater is on then there are fluctuations. Maybe PWM noise somehow make transient.
In Sbase 1.3 thermistors are connected like this

are those really inductors on the sense wires for the therms?
i dont recall them being used on any other board designs. why would they put inductors on the thermistors? are they trying to choke the signal?

Yes, small SMD inductive filters 600Ohm at 100Mhz

I try to use nonlinear extrusion feature but firmware says "M592 not supported".
RepRapFirmware for LPC176x based Boards 2.03+2
Duet Web Control 2.0.0-RC5

According to changelog I think it is already implemented. What can be done to enable it?

Quote
PB
I try to use nonlinear extrusion feature but firmware says "M592 not supported".

What are you other parameters?

Quote
sinned

Quote
PB
I try to use nonlinear extrusion feature but firmware says "M592 not supported".

What are you other parameters?

E.g. M592 D0 A0.1 B0.0 C0.2
The output is same at any parameters "Warning: M592 command is not supported"

Quote
PB
I try to use nonlinear extrusion feature but firmware says "M592 not supported".
RepRapFirmware for LPC176x based Boards 2.03+2
Duet Web Control 2.0.0-RC5

According to changelog I think it is already implemented. What can be done to enable it?

Yeah I just checked the code and it's still disabled in Pins_LPC.h, so it's not currently compiled in. I'll enable it for next version.

Quote
sdavi

Quote
PB
I try to use nonlinear extrusion feature but firmware says "M592 not supported".
RepRapFirmware for LPC176x based Boards 2.03+2
Duet Web Control 2.0.0-RC5

According to changelog I think it is already implemented. What can be done to enable it?

Yeah I just checked the code and it's still disabled in Pins_LPC.h, so it's not currently compiled in. I'll enable it for next version.

Thanks! I tried to make from current source (of course I enabled it in Pins_LPC.h beforehand) and got some syntax errors I consider they are because of older version compiler.

1. ../RepRapFirmware/src//Libraries/Fatfs/port/lpc/diskio.cpp:64:51: error: no matching function for call to 'SDCard::disk_read(BYTE*&, DWORD&, BYTE&)'
     while(_ffs[drv]->disk_read(buff, sector, count) != RES_OK){
                                                   ^
In file included from ../RepRapFirmware/src//Libraries/Fatfs/port/lpc/diskio.cpp:16:0:
CoreLPC2/libraries/SDCard/SDCard.h:60:9: note: candidate: int SDCard::disk_read(uint8_t*, uint32_t)
     int disk_read(uint8_t *buffer, uint32_t block_number);
         ^~~~~~~~~
CoreLPC2/libraries/SDCard/SDCard.h:60:9: note:   candidate expects 2 arguments, 3 provided

2. stdptional is unsupported by compiler (got a recommendation to replace for std::experimentalptional in all code.)
etc.

I use arm-none-eabi-g++ (GNU Tools for ARM Embedded Processors 6-2017-q2-update) 6.3.1 20170620 (release) [ARM/embedded-6-branch revision 249437] (OS X)
maybe I need to use another version?

Edited 2 time(s). Last edit at 07/25/2019 04:29AM by PB.

Quote
PB
Thanks! I tried to make from current source (of course I enabled it in Pins_LPC.h beforehand) and got some syntax errors I consider they are because of older version compiler.

1. ../RepRapFirmware/src//Libraries/Fatfs/port/lpc/diskio.cpp:64:51: error: no matching function for call to 'SDCard::disk_read(BYTE*&, DWORD&, BYTE&)'
     while(_ffs[drv]->disk_read(buff, sector, count) != RES_OK){
                                                   ^
In file included from ../RepRapFirmware/src//Libraries/Fatfs/port/lpc/diskio.cpp:16:0:
CoreLPC2/libraries/SDCard/SDCard.h:60:9: note: candidate: int SDCard::disk_read(uint8_t*, uint32_t)
     int disk_read(uint8_t *buffer, uint32_t block_number);
         ^~~~~~~~~
CoreLPC2/libraries/SDCard/SDCard.h:60:9: note:   candidate expects 2 arguments, 3 provided

2. stdptional is unsupported by compiler (got a recommendation to replace for std::experimentalptional in all code.)
etc.

I use arm-none-eabi-g++ (GNU Tools for ARM Embedded Processors 6-2017-q2-update) 6.3.1 20170620 (release) [ARM/embedded-6-branch revision 249437] (OS X)
maybe I need to use another version?

Make sure to grab the v2-dev branch of the corelpc as well. Maybe you checked out the Master branch which has the old SDCard files?

This is the version of arm toolchain I currently use (also on OSX): gcc version 8.2.1 20181213 (release) [gcc-8-branch revision 267074] (GNU Tools for Arm Embedded Processors 8-2018-q4-major)

Quote
sdavi

Quote
PB
Thanks! I tried to make from current source (of course I enabled it in Pins_LPC.h beforehand) and got some syntax errors I consider they are because of older version compiler.

1. ../RepRapFirmware/src//Libraries/Fatfs/port/lpc/diskio.cpp:64:51: error: no matching function for call to 'SDCard::disk_read(BYTE*&, DWORD&, BYTE&)'
     while(_ffs[drv]->disk_read(buff, sector, count) != RES_OK){
                                                   ^
In file included from ../RepRapFirmware/src//Libraries/Fatfs/port/lpc/diskio.cpp:16:0:
CoreLPC2/libraries/SDCard/SDCard.h:60:9: note: candidate: int SDCard::disk_read(uint8_t*, uint32_t)
     int disk_read(uint8_t *buffer, uint32_t block_number);
         ^~~~~~~~~
CoreLPC2/libraries/SDCard/SDCard.h:60:9: note:   candidate expects 2 arguments, 3 provided

2. stdptional is unsupported by compiler (got a recommendation to replace for std::experimentalptional in all code.)
etc.

I use arm-none-eabi-g++ (GNU Tools for ARM Embedded Processors 6-2017-q2-update) 6.3.1 20170620 (release) [ARM/embedded-6-branch revision 249437] (OS X)
maybe I need to use another version?

Make sure to grab the v2-dev branch of the corelpc as well. Maybe you checked out the Master branch which has the old SDCard files?

This is the version of arm toolchain I currently use (also on OSX): gcc version 8.2.1 20181213 (release) [gcc-8-branch revision 267074] (GNU Tools for Arm Embedded Processors 8-2018-q4-major)

Thanks for advice! Compiled with nonlinear extrusion support. Will test.
arm-none-eabi-gcc (GNU Tools for Arm Embedded Processors 8-2019-q3-update) 8.3.1 20190703 (release) [gcc-8-branch revision 273027]