Would you mind posting the part number of your eMMC module as well?
It doesn’t matter whether your 5A is pink or not, it’s the eMMC module that matters.
5A corrupts eMMC (was 5A does not boot)
1 Like
Here they are… yeah mine works.
Atm im using the amazingfate armbian image before The monka image all them two worked straight out of the box with this emmc.
I cant however… do that FydeOS .dtbs renaming to boot into this Rock 5A.
This is the one I have:
z.z’
1 Like
Thanks. It’s impossible to see anything in the picture with the module’s top, but from the logs it seems to be FSEIASLD-128G (confirmed via DM).
1 Like
One more data point.
The Samsung module, KLMCG4JETD-B041 (the one that has write corruption with 5A), is now placed in my Rock 5B, and works at HS400ES with HSQ just fine.
As another data point, my Foresee module, FSEIASLD-64G, is now placed in my Rock 5A, and also works at HS400ES with HSQ just fine. It’s a similar one to the module @DarkevilPT is using (apart from being half the size), and I can now also confirm that it works with 5A without issues.
It’s a shame that Allnet has silently changed their eMMC chip supplier (note: both the front and the back of the module still looks exactly the same, the only thing they’ve changed is the chip itself), and it’s a shame that 5A has these weird compatibility issues, but oh well…
2 Likes
Is it still going to take some time to release the images?
I was able to confirm that it boots using the armbian image, but would like to use the radxa official image.
Armbian image is better for you, it’s using full speed of your Foresee eMMC module, whereas Radxa’s official image is capped at half the speed, to help those with a Samsung module.
Thanks for the reply.
The reason I wanted to use the official image is because I have no sound on HDMI and GPU is not enabled in my browser, so I wanted to see if the official image would solve this problem.
I think this is an issue with the Armbian image, so I will ask the question in the Armbian community.
I got the one not compatible with 5A (KLMAG1JETD-B041) - is it worth trying with armbian on 5A or will it just fail?
It’s going to fail, I’m afraid. It’ll probably work with Radxa image, but the speed will be limited.
1 Like
After more testing we found that if we bump eMMC voltage from 1.8V to 1.825V we can have HS400 working with Samsung eMMC module on ROCK 5A.
We will check if this is with-in specs with Rockchip before rolling this change out.
Update: another fix was discovered without the need to increase voltage.
BTW: It seems that I have the same problem with the SPINOR at the EMMC socket. When I try to write it, thw write operation “hangs”, and no data is written.
What’s the other fix?
The other fix disabled HS400ES:
&sdhci {
bus-width = <8>;
no-sdio;
no-sd;
non-removable;
max-frequency = <200000000>;
mmc-hs400-1_8v;
/delete-property/ mmc-hs400-enhanced-strobe;
pinctrl-names = "default";
pinctrl-0 = <&emmc_rstnout &emmc_bus8 &emmc_clk &emmc_cmd>;
//pinctrl-0 = <&emmc_rstnout &emmc_bus8 &emmc_clk &emmc_cmd &emmc_data_strobe>;
status = "okay";
};
I think this information is already unnecessary. I was able to confirm that the following image works in one way.
Ubuntu booting, GPU, HDIMI sound output, etc.
I think this image worked most correctly, although it is a third party.
After some more testing the above 2 fixes can’t fix all HS400 issues. We will keep the HS200 workaround.
2 Likes
I have the same issue. Tried 32gb and 64gb card with no luck including the image above. What’s the current solution for using EMMC?
As an update, Orange Pi eMMC 256GB module FEMDNN256G-A3A56 works just fine, and is cheap as hell.
3 Likes
iNAND CHINA SDINBDV4-256G (same capacity, bought from Orange pi) - so far works well with Armbian on 5a
I’ve been watching this issue for a long time.
When do you think you will be able to resolve it?