Programming dsPIC33's with F6 and EB064

[randomrichard]

With the EB085 DSP input board on socket BL of EB064 I have found that the only way to program the PIC easily is to disconnect the Chip Select (RB0) connection between the chip and the input board. The program is then installed on the first or second try every time. Otherwise it is necessary to close and re-open flowcode and/or switch the EB064 off and then on again before the chip will accept a program, and then only after several tries. I use Win7/64bit.

[Benj]

Hello,

What jumper setting do you have the EB085 set to and what features of the EB085 are you using in your program e.g. ADC or Analogue value.

I think Jumper setting A routes the analogue signal to pin 0 which is also the programming pin. You could instead use the Patch jumper setting and use single core wire to join across the signals you require leaving pin 0 free.

[randomrichard]

Hello Ben
I have attached the Flowcode 6 file I was using. The EB085 maps to port BL with the jumper on Patch, with wires connecting CS to pin1 (inexplicably labelled 0 on the board!), MISO to pin5 (labelled 4) and SCK to pin4 (labelled 3). I have checked the outputs with my Picoscope and all works fine, and very fast: 24 bits sent in about 50microsec at Fosc/4 with the internal oscillator. The problems are with the USB connection to RS232, using the EB039, jumpers C and 2.. I was only able to receive data at 115200 baud maximum using my VB.NET program (I tried software bauds up to 921600), necessary to save the data for display in Excel. An error, possibly during transmission, reverses the sign of some values, corrected by the PC software. Excel displays the digitised waveform but it is unacceptably noisy. I need at least 10K lines/sec and preferably as much as 40K lines/sec for my application, which the UART -> USB link precludes. I did try 12Meg and 20Meg crystals to improve baud rate accuracy but to no avail.

Therefore I am once again forced to pension off Flowcode and look elsewhere to find the necessary speed. I am currently evaluating a Microchip PIC32 development board using their new "Harmony" programming environment, so that I can use a true USB high speed link. I tried the 24FJ64GB002's USB through the EB055 but the latter has series resistors on the wrong pins so I built my own version of the EB055 with them on the correct ones (D9 pin3=D+, 4=D-, 5=Vbus, 9=gnd, with the Vbus not connected to the USB socket but to ground through a 220nF). It didn't work; the PC tried to handshake but would only install the phoney 2006 driver.

I expect to have at least as much trouble with the Microchip system but by the end it will still have been cheaper and better than going back to college!
Richard