Jump to content

latigid on

Frequent Writer
  • Content Count

    2,391
  • Joined

  • Last visited

  • Days Won

    132

Posts posted by latigid on

  1. https://github.com/midibox/mios32/blob/master/mios32/STM32F4xx/mios32_ain.c
    https://github.com/midibox/mios32/blob/master/modules/ainser/ainser.c

    https://github.com/midibox/mios32/tree/master/modules/aout

    Basic drivers/assignments:
    https://github.com/midibox/mios32/blob/master/mios32/STM32F4xx/mios32_spi.c
    https://github.com/midibox/mios32/blob/master/mios32/STM32F4xx/mios32_iic.c

     

    Maybe it would be easier for you to use the Arduino and output an analogue voltage to be read by the STM32F4 AIN? That way you can use your working system without much extra coding effort. MBIO supports AIN either on the MCU ports J5A/B or using the AINSER module (typically less noise on the signal).

  2. 13 hours ago, zaordsword said:

    Hi latigid !

    Thanks for your answear :)

    >>have you written a driver for it to be used with MIOS?


    No, I don't know how to wrote the driver working with MIOS, but I have a script for making it working with arduino and python.
    Do you have examples of drivers ? I have see nothing about this topic anywhere on ucapp or on wiki or documentation and forum ...

    As far as I know, there are no I2C peripherals that were coded into MIOS32. The closest thing that I am familiar with is the I2C MIDI modules, that according to TK. were not trivial to code for, and at that on very low-performance 16F PIC chips.

    All the MBHP framework gives you is a hardware port on the MCU. You could consider to compare your arduino sketch with a device running off an STM32F4 chip or similar.

    Alternatively, you might find more code examples of SPI devices. Maybe you can try to code a driver for a similar sensor with an SPI interface?

     

    13 hours ago, zaordsword said:

    The bmp280 is a pressure / temperature and humidity sensor manufactured by boch. I join the datasheet, my python and a arduino script. (c ++)

    Edit: I had upload server -200 error when I try to up the datasheet . It's a project exemple availble here. The datasheet is on le page or  datasheet



     >>How would you expect the velocity values to be modified?
    The sensor give the barometric pressure . I plan to save the atmospheric pressure at initialization as P0 and find the realtime surpressure like P_surpess= ( Press_sensor(t) - P0) 

    Then I would like to map a certain variation of pressure matching with the velocity from 0 ... 127

    >>Have you looked into the MBIO code?

    What is MBIO code ? If you talk about the code on github there is a lot of file and I don't know where to find the file where to modify or to put a driver.
    I did not find documentation of a list of the reference classes / function. 

    The uCapps project is here:
    http://ucapps.de/midio128.html -- it is more targeted towards general users who would like to use the project without writing code for it.

    The git project is here:
    https://github.com/midibox/mios32/tree/master/apps/controllers/midio128_v3

    You would need to modify the code to suit your needs. Note that the code is written in C, not python.

    You can find the documented functions and some examples here:
    http://www.midibox.org/mios32/manual
    http://ucapps.de/mios32_c.html
     

    Best of luck to you if you decide to follow this journey. Others will probably not write the software for you, but if you post your modifications and explain where you are having difficulty, some might be able to help.

  3. Hi Matt, wonder if it is a short circuit to 0V (ground) somewhere. 

    Test each board in turn for shorts on the +5V line. Easiest is across any 100nF cap. Did you solder in the Matias switches yet? Sounds like it is on the Base or Plate PCB?

    Please upload sharp photos of the PCBs.

    Some status LED illumination is expected if the audio jacks and PCB stack are disconnected.

    Best,
    Andy

  4. Hi! 

    Great, we want to see the image (and give you a serial number)! There is a 2MiB size limit, so please try to reduce the size with an image program and save as jpeg or png, should work then! 

  5. :cheers:

    Cool! I think here you will have an easier time wiring up and figuring out things hands-on. Please feel free to post your progress.

    Best of luck!
    Andy

  6. No problem, hope it's helpful!

    You will need to figure out the wiring puzzle. The following links might be of use:

    http://ucapps.de/midibox_ng_manual_ngc.html -> search "LED_MATRIX" and "DOUT_MATRIX"
    There you can understand how things will look in software.


    http://ucapps.de/mbhp/mbhp_dout_8x8leds.pdf
    http://ucapps.de/mbhp/mbhp_dout_8x16leds.pdf
    http://www.midibox.org/dokuwiki/lib/exe/fetch.php?media=home:users:ilmenator:rgb-led:midibox_rgb-led_matrix_8x8_v1.3_schematic.pdf
    https://github.com/midibox/mios32/blob/master/apps/controllers/midibox_ng_v1/cfg/tests/blm16x4.ngc

     

    DOUT_MATRIX n=1  rows=4  inverted=0  sr_dout_sel1=1  sr_dout_r1=2  sr_dout_g1=3

    This is an .NGC example of one four-row matrix. In theory you could share the same sr_dout_sel1=1, though you still need dedicated RGB columns. As you have common-anode LEDS:

    inverted_sel=1 and (optionally) inverted_row=1

    commands should be used when defining the matrix.

     

    6 hours ago, doc.g4 said:

    create one larger matrix using sparks fun type shiftreg break out boards then to a MIDIBOX DOUT BOARD.

    This: https://www.sparkfun.com/products/10680 ?

    You don't need that if you will use DOUTX4 PCBs. You can get creative with how you assemble IDC cables, no need to follow the original order or total number of conductors.

     

    6 hours ago, doc.g4 said:

    i thought i could connect a  second 72 led matrix into SR U12 via q4-q7

    I guess you could consider defining an 8-row matrix?

    If you save one shift register but give yourself a wiring headache, maybe the simpler approach is better. See ilmenator's approach for an 8x8 RGB matrix. 

     

  7. Still not totally easy to follow but I get more where you're coming from. I would prefer to draw it out as a logical table rather than a netlist. Are the LEDs on boards and you want to use DOUTX4 PCBs to drive them? Or are the shift registers also on your LED boards? 

    4x source rows as selection pulses will work, though you get no faster multiplexing as that's limited to a doubled 4x pattern. But colours can only be 16 columns wide. 

    If using DOUTs, you could buss two IDC10 ribbon cables and define them as one matrix but you still need a second DOUT to drive the 'second' matrix (third LED board). I would suggest simply to use 3x DOUTX4 boards and define three separate matrices with 4x source rows each. 

  8. MB_NG supports multiple matrices. You just need to define the shift register where each block of 8 is connected. If any block is less than 8 wide, no problem. It will just take more shift register chips to address them. 

    Best idea is to draw a circuit diagram of how the LEDs are arranged. Then for each contiguous block, that is one shift register with a maximum of 8 rows or columns. You probably need sink transistors (NPN) on the rows unless your current draw is low. 

    Please keep posts public so other travellers can benefit from any knowledge or lessons. 

  9. You can create a matrix with 9x rows (sink side) and 8x3 columns (R/G/B source sides).

    So you need 5x 74HC595 chips because they only have 8x outputs. You also have 7x rows spare as it is a 16x8(x3) matrix.

    If you have buttons you can connect 1x DIN 74HC165 chip to the columns with diodes to the switch inputs. 

     

  10. Hi Simon and a happy new year to you too!

    We try to keep track of the "successfully built" SEQ v4+s in this thread, so I don't think you got a serial number yet as you were posting in the "troubleshooting" thread? Happy to assign you #39 if you want to post a pic here?

    Best,
    Andy

  11. Hi,

    What meters do you have? Some have a "high-res" function that is activated by holding the backlight button (Fluke meters). Another option is to buy resistors with better tolerance (e.g. 0.1%) but they will be more expensive. In the end, the CV is calibrated on the SEQ and has a variable gain with the first op amp stage. So you should be fine.

    For more than eight outputs, simply chain the J19 data from the superDAC board. RC1 should still be set for all DACs. 

    Best of luck and happy new year!

    Best,
    Andy

  12. For time-multiplexed LEDs you can exceed the 20mA current rating or whatever the datasheet says. Some even give a peak current rating that is valid for a given duty cycle (normally 10% rather than our 12.5 or so) and pulse time.

    Totally depends on your LEDs. A tip is to use super bright ones, but probably you already have something installed. In the SEQ v4+ we use 47R but these are superflux RGB LEDs designed for more forward current. Generally the higher the pulse energy, the lower the lifetime of the LED. 

    Keep in mind that the 74HC595 outputs are overdriven in this configuration. Your +5V rail will sag and in the worse case it won't be enough to power the PIC properly. But many MIDIboxes work fine like this :).

     

    • Like 1
×
×
  • Create New...