Hello Peter!

Is RC20 containing the solution for the json problem also?

G@bor

yes

  matherp said  

V6.03.00RC20
UI changes to the FM command - see the chapter in the manual
Fixes bug in LIST SYSTEM I2C
Camera command now supports 320x240 capture for applicable RP2350 versions (now including HDMI and HDMIUSB)

The camera works excellent with HDMIUSB. Colours look really good and natural and the picture is quite sharp with good Autofocus. Thank you for the improvements!
If it would work in terms of speed a buffer would make sense to me if PSRAM is present.

List system i2c works fine for me.

FM and its help look clear. When copying files the status of the copy process should be displayed immediately. I tend to wonder if I pressed the button correctly..



The picture on the HDMI monitor look also very nice in QVGA:

PSRAM is far too slow

This is what my FM help in > HELP looks like. Tastes differ.



Jan.

G'day Peter,
A few observations noted whilst testing various systems and ideas recently and a couple of follow up questions:

RP2350BTH: Several keyboards tested and generally highly successful, however scroll wheel / scroll function and CAPS LOCK LED not working on any version, is this correct?

RP2350HDMI: You have stated the GT911 I2c touch will not be supported on HDMI, is it possible to enable the SPI XPT2046 resisitive touch controller on non-usb versions to take advantage of the numerous HDMI/Res-touch panels available aimed at the raspberry-pi?

Updating Firmware: Several suspected bugs and unusual system behaviour observed after a firmware update / option update has been made good by reloading firmware and manually setting options. A flash nuke is always done between firmware loads. Would you advise to manually update option settings between RC/Beta versions rather than OPTION DISK SAVE/LOAD?

Stepper System: To make system calibration easier, would it be possible to allow an axis to be re-defined (possibly after a STEPPER RESET) as long as the pins used are not changed, IE:

Currently this appears to fail as the pins are not released until STEPPER CLOSE is issued, however I may have missed something obvious as usual.

The only other thing missing is a true low power sleep mode, however I believe that is not possible due to the interaction of MMBasic and the RP-Pico system.

Many thanks once again to you and Geoff for the whole PicoMite/MMite/MMbasic system, amazing how far it has come.

Regards, Lyle.

  matherp said  

V6.03.00RC20
UI changes to the FM command - see the chapter in the manual
Fixes bug in LIST SYSTEM I2C
Camera command now supports 320x240 capture for applicable RP2350 versions (now including HDMI and HDMIUSB)

LIST PINS cuts off the first 6 pins on a PicoCalc LCD display (26x40):


Edited 2026-06-17 01:07 by toml_12953

You probably have option tab set to something other than 4. OPTION TAB 4 then try again

  Quote  

RP2350HDMI: You have stated the GT911 I2c touch will not be supported on HDMI, is it possible to enable the SPI XPT2046 resisitive touch controller on non-usb versions to take advantage of the numerous HDMI/Res-touch panels available aimed at the raspberry-pi?

Sorry, but this is a no. resistive touch is a different compile module and is not compatible with VGA and HDMI builds

The option structure has changed so you can't load a previous option file

What is the requirement on the stepper? Just changing the steps/accelerations without restarting?
Edited 2026-06-17 01:21 by matherp

Hi Peter,
Correct, being able to change the Steps/mm or acceleration without shutting down the complete stepper system and then reloading all from scratch. Wasn't sure if there was a way to do this already. This would certainly make the machine calibration routine far cleaner and allow different machine modes to be set easily.

Thanks for the clarification regarding RES-Touch.

Regards, Lyle.

Peter,

there is an issue with PSRAM display at least ...

What board and what ram chip?

Peter,

I have loaded the latest PicoMiteHDMI MMBasic USB RP2350A Edition V6.03.00RC20 firmware.
What I have noticed is that switching between left and right with FM and an HDMI screen using the Tab key works well with Tera Term, but not on the HDMI screen.
The situation is as follows: There are two cyan bars. One for the cursor and one for the drive.
When I press Tab, I see the two bars switch on Tera Term, while on the HDMI screen only the cursor bar appears.



Jan.

Easily fixed - will be in the next RC or release whichever comes sooner

  matherp said  

What board and what ram chip?

Board: Waveshare RP2350-Plus
Chip: PSRAM64H, retrofitted

The board+chip did work fine with older versions, did not test it with RC19 though.

I will try to find out with which version the change did happen.

  ville56 said  

  matherp said  

What board and what ram chip?

Board: Waveshare RP2350-Plus
Chip: PSRAM64H, retrofitted

The board+chip did work fine with older versions, did not test it with RC19 though.

I will try to find out with which version the change did happen.

Solved ... a power cycle did reset the issue. Sorry Peter, my bad.

  matherp said  

You probably have option tab set to something other than 4. OPTION TAB 4 then try again

Even with OPTION TAB set to 2 I get this:

ad PSRAM issue:

it turns out that the problem is with the Cytron board (the one with the LED for each pin) that holds the pico board. GP19 is also connected to the audio-out network. That seems to spoil the timing on the PSRAM CS signal. Worked with earlier versions though. Have the I/O pin drive levels changed in V6.03.xx ?

Anyway, if out of the Cytron board everything works fine.

Hi Peter,
I'm doing some testing on B20 and think there is a problem with Math(crossing)
Base is set to 0

-1 ok
but shouldn't the first 1 be 0 then the second 1 would be correct?


shouldn't 26.75 be 7
What should I get if I use 26.85 as the crossing value, direction 1, would it be 8 or -1 for outside the array data?
If I use -1 for the direction, I always get -1 as the result.
Thanks for any insights. Kevin.

I'm getting some behaviour from the Stepper system that I don't understand.
I request a move of 1 (on the A-axis) but only seem to get a move of 0.875

I suppose this might be some sort of rounding error, but at 16 steps per unit, there ought not be such.

The stepper driver is set to 16 microsteps, the motor is 200 steps per turn. The goal is to turn the motor such that there are 100 positions per turn. The setting of 16 steps per unit is puzzling, but works. A move of A100 does indeed give one turn of the motor. But a move of A1 does not give 1/100 of a turn.

In the code below, two GCode lines are offered, one with incremental moves, the other with absolute. Both show the same behaviour.

I note too that G90 and G91 are not modal - they don't set the state for following moves. Instead, they (or at least G91?) must be included with each line GCode.


Edited 2026-06-17 08:32 by vegipete

G'day,
Looking at Vegipete's code this might be a better solution:

to catch the last update after motion finishes.

However it does appear that short moves at higher feedrates cause what he is seeing:

X1 move at feedrates < 1896 work ok
X1 move at feedrates > 1898 moves less than requested

This is same with G1 or G0 moves.

Also my request for the change for the Stepper System should be STEPS/MM and MAX_VELOCITY not MAX_ACCEL, however MAX_ACCEL would be nice also. (getting over the FLU/COVID   )

Regards, Lyle.