--
Has recibido este mensaje porque estás suscrito al grupo "FPGAwars: explorando el lado libre" de Grupos de Google.
Para cancelar la suscripción a este grupo y dejar de recibir sus mensajes, envía un correo electrónico a fpga-wars-explorando-el...@googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/930cbfef-e66e-4175-8c0d-819f4c340ca1n%40googlegroups.com.
Now I have included this board --> Colorlight i5-v7.0 from here --> https://github.com/wuxx/Colorlight-FPGA-Projects
There is another board --> iCESugar-pro from here --> https://github.com/wuxx/icesugar-pro
Please, could you confirm which is your board??
You said that you flash your board with ECPdap from cli
1) What is the command you are using ?? is this your command line (ecpdap program Your_bitstream.bit) ???
2) Are you running in Linux or a Windows OS ?
If you are running in Linux, do you need to include the USB device detected in the OS in the udev rules ??
ecpdap flash write bitstream.bit
https://github.com/adamgreig/ecpdapPara ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/406293f7-9ad8-41ce-9d5c-6c2332d5f449n%40googlegroups.com.
--
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/6db3af3f-8906-48d3-9f92-f63af2b04bb7n%40googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/CAEQ-VCZiOh78JA%2BfzOT_R6ka9Q5GjdbeAnhN1_U0T5juUwunhg%40mail.gmail.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/c4c7b777-128f-4573-b6e8-3d7c8579ec84n%40googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/90c84515-f77a-4a0f-a082-b12ab4c6f05bn%40googlegroups.com.
--
Has recibido este mensaje porque estás suscrito al grupo "FPGAwars: explorando el lado libre" de Grupos de Google.
Para cancelar la suscripción a este grupo y dejar de recibir sus mensajes, envía un correo electrónico a fpga-wars-explorando-el...@googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/9942c070-9dd2-415c-8f27-1c54f4504123n%40googlegroups.com.
--
Has recibido este mensaje porque estás suscrito al grupo "FPGAwars: explorando el lado libre" de Grupos de Google.
Para cancelar la suscripción a este grupo y dejar de recibir sus mensajes, envía un correo electrónico a fpga-wars-explorando-el...@googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/9942c070-9dd2-415c-8f27-1c54f4504123n%40googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/CAEQ-VCYMJFwB%2BARko%2BEArOB08wDbdyV4PmW%3DKhac4tWLYh%3D7Cg%40mail.gmail.com.
Hello Francisco,
Can you please crosscheck if there is a little typo error for a directory in
your instructions.
At the step 5) (or 4) in Spanish)
“
5) In the folder -->
C:\Users\your_user\.icestudio\venv\Lib\site-packages
Overwrite util.py file of util.zip file
"
It seems to me that in the original icestudio installation, util.py is in the folder
C:\Users\your_user\.icestudio\venv\Lib\site-packages\apio
But even with this modification, icestudio can still not communicate with my
colorlight i5.
It seem I have a problem with ecpdap on my win10. Which is weird because ecpdap
is fully working on a linux VM on the same machine.
It will be nice if you can make some precise (for linux noobs) instruction for
the icestudio modification on a linux system.
Anyway, thanks for that good work.
Have a good Sunday
Joaquim
Hello Fernando,
Yes, you are right ecpdap is quite slow(it can take one minute to load a bit file). But as a STM chip is what is actually installed on the dev board!
Also, with this devboard and colorlight i5 board combo, we have to be careful to avoid connecting the jtag outputs of the STM chip and the jtag outputs of another programmer.
When the i5 board is well inserted in the sodim connector the 4 jtag pins are mechanically connected at the same time with the help of 4 spring pins (which connects the i5 board to the STM chip on the dev board).
So if we want to inject another jtag signal on the 4 pins of the i5 board, it may be necessary/safer to add an insulating plate between the spring pins and the i5 board for fully disconnecting the STM chip outputs.
see the following picture it may help understanding.
In the future, I planned make my own pcb for connecting the i5 board(with a sodim +4 “spring” pins). And i will add on the pcb a connector for connecting a USBblaster. So yes, if you can implement the USBblaster it will be nice.
Concerning ecpdap on win10 or win seven (in two different machines).
It is able to see the programmer but not the jtag device.
On a windows terminal I got:
C:\ecpdap>ecpdap
probes
Found 1 CMSIS-DAP probe:
0d28:0204:070000010669ff343832434257103334a5a5a5a597969908
DAPLink CMSIS-DAP
C:\ecpdap>ecpdap scan
Error: specified probe not found.
And on the same win 10 machine but running a linux VM, it works fine :
q@q-virtualbox:~$ ecpdap probes
Found 1 CMSIS-DAP probe:
0d28:0204:070000010669ff343832434257103334a5a5a5a597969908 DAPLink CMSIS-DAP
q@q-virtualbox:~$ ecpdap scan
Detected JTAG chain, closest to TDO first:
- 0: 0x41111043 ECP5 LFE5U-25 [IR length: 8]
q@q-virtualbox:~$
And I can load bitstream with the terminal. But on this linux VM I didn't yet manage to configure well icestudio!
regards
joaquim
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/8e31e294-8e9e-45fc-ad83-e22fe8f41df3n%40googlegroups.com.
Hello Steinor,
Therefore, did you manage to program your colorlight i5 board using the usb connector of the devboard:
- With icestudio on windows 10 (using ecpdap)?
- Just with ecpdap on a win 10 cmd terminal?
I only make it work using ecpdap on linux and it work very well (no need to push on the jtag pins for now :-). but without icestudio for now.
Yes I tried about everything with ZADIG and the usbdrivertool.
Mainly, installed winusb(libusb) driver for the CMSIS-dap daplink(see
picture) but no improvement.
Yes, understood that libusbk is for the FTDI chips and not the stm32 ones.
I just ordered a usbBlaster (it will be there in one month :-) )
Thanks a lot for your replies
Joaquim
Muchas gracias Fernando,
I have just to wait 3 weeks for the delivery of my alliexpress usbBlaster.
In the meantime, I will modify slightly my colourliht i5 devboard to allow it to be programmed by this programmer. I will post a picture here.
Have a good day.
Joaquim
Here are few pictures of my colorlight i5 modification (for USBblaster compatibility).
Good news, i found an FT2232H mini module in one of my drawers. So no need to wait for the delivery of my aliexpress usbblaster to test icestudio with my colourlight i5 board.
Anyway, the update of colorlight i5 (usbblaster) will also be useful because it offers a cheaper solution for the masses ;-)
A little remark Francisco :
In your message of 31 march at around 9h33m49s(+/- few hours depending on your time zone)J
You recommended connecting the 3.3 V on the i5 dev board side.
But when the dev board is powered with its main power source, (like you are also doing with your colourligth 5a-75E board), and another 3.3V is connected in the same time from the programmer side, we may have a short-circuit conflict between two different 3.3V regulator outputs.
This may induce a current from the higher voltage regulator (eg 3.35V) into the lower voltage regulator (eg 3.25V). And as the connection resistance is very low, we can have an unwanted high current consumption and (even a component damage potentially).
Thanks again for all that good stuff Francisco
Regards
Yes Fernando, I did exactly that and got the same result as you .
it is just that a zero which turns on a led is a bit counter-intuitive :-)
But anyway, as this led is mainly used for our test during the devellopment of our code, we just have to get used to it ;-)
Regards
joaquim
Hello Everybody,
I don’t know if it is related to the colorlight i5 toolchain or to Icetudio but I have strange think here.
With the above simple 4 bits counter circuit, if:
- I make an verify/build/upload
- And I just make changes in the ouput numbers. (eg P2_L18 becomes P2_L20)
- I make second verify/build/upload
Then my logic
analyser does not see any wiring change on the ouputs signal.
So it seems thaht the second upload was identical to the first one!
To make it work as expected I had to add a step in the sequence:
- I make an verify/build/upload
- I make changes in the output numbers. (eg P2_L18 becomes P2_L20)
- I add block (eg: constant block left unconnected on the workspace)
- I make second verify/build/upload
Then, my logic
analyser see all the wiring changes on the output signals.
Any ideas? i joined the .ice file
Have a good evening
Joaquim
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/5662c9ab-39c0-42af-bbe7-75d1fe879e29n%40googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/a2c0366a-4dd1-4a4e-a804-161c61e0e4b5n%40googlegroups.com.
Hello Fernando,
I am struggling with those icestudio savings problems.
I isolated a very simple case (see joined file) thaht may be eazy to debug.
On my machine, when:
- I save the file (even with a new file name / “save as” ).
- Quit / close the file
- Then reopen it
I have 4 pins (out of100) that are saved as other pins .
Before saving and quit at reopening
P3_C1 ------> P3_C17
P3_D2 ------> P3_D20
P3_B1 ------> P3_B19
P4_E1 ------> P3_E19
Before opening an issue on the icestudio github page, I would like to tri to see that my icetudio installation is not corrupted.
So could you please check also on your machine by loading the file without converting it to a different board. I am using the ColorLight-i5-v7.0_(FT2232H).
As we actually do not need to have the dev board connected for making this test, I will try to see if I can reproduce this on the Alhambra board pins.
Other question, is there a reason for not inserting also the P1 connector (ETH ports) in your pinout.json file? Even if I do not need those for now J
Thanks
Joaquim
I just tried my .ice file (supplied in the last post) with the Alhambra II board selected and 2 ( out of 51) pin are also wrongly saved.
SDA becomes ADC_SDA
SCL becomes ADC_SCL
I think thaht I have a hint here.
All of this errors have something in commun. Text of the first is also fully present in the second one
P3_C1 --> P3_C17
P3_D2 --> P3_C17
P3_B1 --> P3_C17
P4_E1 --> P3_C17
Alhambra II board
SDA becomes ADC_SDA
SCL becomes ADC_SCL
Ok, I just tried the obvious workaround:
Renaming in file “pinout.json” the problematic pins by adding an ”_”
(eg: P3_C1 changed to P3_C1_ )
And the problem is gone.
It seems evident that we have a weakness in the .Json files attributes reading.
So i will open an issue on icestudio/github because as the software is using plenty of .json files, the problem is maybe also happening with other attributes used in the software (causing potentially other issues).
Oups, Sorry
I just realised that I addressed my message to Benito instead of Steinhor in my message 3 hours ago! I should sleep more :-)
Steinor are you running your icestudio on windows with a "colorlight i5 ecpdap" selected board?
Personally, as I never managed to do that, neither (like you) configure fully icestudio on linux .
I decided to mod my colorlight i5 board (see messages few days ago). And as Fernando (Benito) wrote, the programming of the chip is faster (~1 sec). This is quite nice for debugging sessions.
Joaquim
Ok Fernando,
Looking at your board pinout.json file (ColorLight-5A-75E-V71_(FT2232H)).
I localised that we can expect a problem with the pin “CLK” which will be replaced by "Shared_CLK" .
I tried and, the problem is there.
So for now, if you want, you can apply the workaround i proposed earlier.
Regards
Hello,
For info, I received today my cheap UsbBlaster with a mysterious chip inside (IC ref laser erased)
And after loading the Winusb driver with zadig ,I tested it with my “slightly moded” colorligth i5 board and icestudio and everythink works fine so far on win10.
As Fernando have seen with his colorlight 5A-75 board the upload is slightly slower (13s) than with the FT2232 programmer (8s) for a same design.
Joaquim
Hello Fernando,
After reading your message, I was afraid of having partially damage my 3.3V fpga setup. Because my usb blaster looks like your first image, the one 5v powered.
Visually, my blaster has less components (no external quartz to clock the IC,…).
I measured the blaster and found that It has a 3.3V regulator. So only two ICs on this board ( the regulator and the mysterious main IC).
So maybe a newer and cheaper to produce version of these blaster?! I got mine with the Alliexpress link you supplied us https://fr.aliexpress.com/item/32825776038.html
Great my board is still alive. :-) .
Well in a sense, i am waiting the moment where my 25KLut fpga fries;-)
if it happens, I will try replacing it with the 85KLut one ( with the same 381bga package ).
Remark: On the newer Icesugar-Pro board muselab used a smaller package the 281bga. So the max size upgrade is 45KLut (see ecp5 datasheet).
Why always more ressources?
The machine i want to devellop needs a maximun of ressources:
Regards
Joaquim
--
Has recibido este mensaje porque estás suscrito al grupo "FPGAwars: explorando el lado libre" de Grupos de Google.
Para cancelar la suscripción a este grupo y dejar de recibir sus mensajes, envía un correo electrónico a fpga-wars-explorando-el...@googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/56dcfad0-729c-4c71-8e0e-9433f1c9cd0dn%40googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/7cc3cbb5-dccf-4cca-9b83-af2f139ccaban%40googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/ba950d38-3d72-4c4a-b797-1b8f7fecbbe5n%40googlegroups.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/CAEQ-VCbyZopELR_1uqWz5O%2BC34UXVghXdJgVKtihvndpBY7u2w%40mail.gmail.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/CAGjhNWOV-0WM4UhCaX4nthDXnnsvm%2BrP-%2B0oNjxu44rP5yER7A%40mail.gmail.com.
Para ver esta conversación en el sitio web, visita https://groups.google.com/d/msgid/fpga-wars-explorando-el-lado-libre/CAEQ-VCbahVGZ%3DqA0CQ%2BQKXUP_rtiCB8v0HZAAXW3519nN-hECg%40mail.gmail.com.