ATTiny2313/4313 Flowcode development board

[mnfisher]

I made a small ATTiny4313 dev board. The MCI can run at up to 20MHz (or 8MHz internal)

I breadboarded and it worked well, although the connection to the programmer was a little 'delicate'

So using an ElectroCookie solderable breadboard i set to make a version with an ICSP header.

This entailed cutting three tracks - and i also added a jumper to allow external power. Originally i had a 10K pull up resistor to pin 1 (reset) bit I removed, when testing failed. It seems okay on the internal pull-up but might be good to re-add.

PXL_20260527_133639488.MP~2.jpg (212.32 KiB) Viewed 325 times

[mnfisher]

I had most difficulty soldering the header pins - some melting has occurred on my VCC jumper

I hit two problems - it didn't work ( a biggy) - i checked the board carefully and I hadn't bridged the links in the power/ground rails..

Still no joy - and attaching a scope, I see that reset doesn't pull low. I removed the 10K resistor

Programming the chip separately (I knew the socket was a good idea) convinced me that the MCU was running okay (blinkie) - so I checked the wiring again!

Then I noticed that I had plugged the programmer cable into the UDP port - doh..

I wrote a small batch file for Flowcode (I used a Diamex programmer) - and all works well....

I'll post after I've polished it a bit... FC doesn't allow a USB port to be specified for ATTiny chips - and hardcoding it seems excessive ('USB' doesn't do it)

So I'll peek at the definition file (and maybe get some pointers from Steve?) to allow this to be set!

Martin

[mnfisher]

... and I forgot to connect VCC to pin 20...

Which took rather longer to notice than you'd hope

[Steve-Matrix]

Ben is more familiar with that part of FCD files than I am, but I think it's one or both of these 2 sections you need to pay attention to:

The first is the use of the batch files and the "UsesComm" and "$(comport)" of the <programmer> section within <build>.

The second is the <comminfo> entries which allow you to specify a VID/PID pairing

[mnfisher]

I think I'm almost there - I created a fcd file AT4313_Dev_Board - and copied across the 'programmer' lines from a device that has a COM port (I used Arduino Uno) - this seems good - with the proviso that 'upload configuration now' gives fuse memory is undefined. I've not altered the fuse settings - and only one byte is offered in configuration. Should there be a H(igh) and a L(ow) byte?

Martin

[Steve-Matrix]

Sorry, I don't know about this particular chip. But the "E-Blocks2 AVR Programmer" fcd might hold a clue. This one has the following section and maybe you need something similar:

Code: Select all

    <configuration address='0x0000' mult='1' progonly='1' >
        <format><byte address='Low' value='0xFF' /><byte address='High' value='0xDF' /><byte address='Extended' value='0xFF' /></format>
        <setting />
        <settings >
            <setting name='SUT_CKSEL' offset='0' mask='63' bits='6' hint='Select Clock Source' >
                <option name='Ext. Clock, Start-up time PWRDWN/RESET: 6 CK/14 CK + 0 ms' value='0' />
                <option name='Ext. Clock, Start-up time PWRDWN/RESET: 6 CK/14 CK + 4.1 ms' value='16' />

                ...etc...

[mnfisher]

Thanks, I'll have another rummage through the fcd files.

It all seems to work quite smoothly for programs which is good.

Auto-detect doesn't pick the com port, but manually selecting it is fine. Using a long name looks a bit odd on the 'quick start' buttons - so a better name is needed.

Martin

[mnfisher]

The description and programmer files.

I feel - something isn't quite right, but it does work. Oddly - after doing new project with ATTINY4313_dev_board as the project type - in options etc - the type shows up as ATTiny4313. However - if you choose this then the programmer speed etc are incorrect (8MHz instead of 20MHz and no USB port)?? The fcdx file maybe needs a CRC ??

Also - if you do new project - choose one type (say ATTiny4313) then doing 'back' and choosing the other (ATTiny4313_dev_board) - doesn't reset the options. Odd - and this only happens between the 4313 / 4313_dev_board files - and if I close / reopen the file - I then can't set the COM port - but it 'remembers' the previous setting (COM7 here)!

I used a new programmer file (AVR_Diamex) - though it should be possible to use one of the standard files (with stk500v2 as the programmer type in this case) - but I didn't want to mess up my set up whilst fiddling.

The ATTiny4313 fcdx file only seems to define one fuse value - although BOD (brown out detection) and other options are available too in hfuse?

'Download config now' can be used to change the clock (from 20MHz ext to 1MHz (8 + div) internal and works well - I use a simple blinkie and a 'scope to test.

I didn't manage to turn off / get to work the autodetect - so the user must untick the box before selecting the COM port.

Although the MCU has a hardware UART - (and the component allows channel 1) - compiling gives:

C:\ProgramData\MatrixTSL\FlowcodeV11\CAL\AVR\AVR_CAL_UART.c: In function
'FC_CAL_UART_Send_1':
C:\ProgramData\MatrixTSL\FlowcodeV11\CAL\AVR\AVR_CAL_UART.c11: error: lvalue required as left operand of assignment
UDR0 = nChar;
^

Possibly should just be UDR, although the CAL file seems to do #define UDR0 UDR? Software mode is okay though.

Martin

[LeighM]

Quick look on my mobile, re the first issue, I think you need to change device name="ATTINY4313' to something like ATTINY4313_dev_board to make the fcd unique. As I recall, as an alternative, there's also a 'file' option property for the device section.
Hope that helps.

[mnfisher]

Thanks Leigh,

I'll have another play - my first attempt stopped compilation with configuration file not found (or words to that effect) - presumably compilation uses chip:l and changing this to the alternate name should fix the issue?

Martin