|
|
So it looks like it has been confirmed that it's a hardware issue, and not a firmware issue. You can view the 2 different links down below for some more info, it looks like there might be a fix already.
Alvatech
Links:
https://4pda.ru/forum/index.php?s=&s...ost&p=87947976
https://4pda.ru/forum/index.php?s=&s...ost&p=87955256
A quick update. thnks to the user aquagenptz from russian forum, e know here is the lan problem, i will post the original post in russion, maybe someone can tralnslate bettehr that google translator:
Приветствую всех!!! Ну что, нашел я китайский косячёк с портом Ethetnet. Теперь немного подробнее... Только я начал тыкать в приставку вольтметром и осциллографом как она возьми да и сдохни, как оказалось, вернулся самый первый косяк моего Кинга, а именно проблема с отвалом памяти, о нем я писал ранее. Пришлось искать трафарет и реболить eMMC, на это ушло пару дней, после чего приставка ожила. Далее проштудировал весь даташит, снял все параметры и пришел к выводу что проблема в шине RGMII (эта та что между SoC и ETHERNET TRANSCEIVER). Осциллограф показывал хорошие фронты сигналов по этой шине, но что то сбоило. Шина подключена к процу через низкоомные резисторы - перемычки, но один из этих резисторов установлен повышенного сопротивления (22 Ом), это настрожило, хотя в идеальных условиях должно работать и так. Резистор этот стоит по цепи RXC - при приеме пакета трансивер шлет меандр с частотой 125МГц для синхронизации, немаленькая такая частота... Требования к разводке платы под такую частоту соответствующие, вот он косяк, а тут дорога уходит в межслой, и обладает большой паразитной емкостью, и резистор в 22 Ом не хило так гасит сигнал. В общем хватит теории, на практике нужно заменить (закоротить) этот резистор перемычкой, на фото отметил. После этой процедуры LAN порт ожил, и работает как надо. С NAS по SAMBA качает 28MB/s, ура-а-а!!!
Т.к. проблемных боксов много эту инфу можно добавить в шапку. И да, все делаете на свой страх и риск...
Now on Google
Greetings to all!!! Well, I found a Chinese joint with the Ethetnet port. Now a little more detailed ... As soon as I started poking at the console with a voltmeter and an oscilloscope, take it and die, as it turned out, the very first jamb of my King returned, namely, the problem with the memory dump, I wrote about it earlier. I had to look for a stencil and reboot eMMC, it took a couple of days, after which the prefix came to life. Then I studied the entire datasheet, took all the parameters and came to the conclusion that the problem is in the RGMII bus (this is the one between SoC and ETHERNET TRANSCEIVER). The oscilloscope showed good signal fronts on this bus, but something failed. The bus is connected to the process through low-resistance resistors - jumpers, but one of these resistors is installed with a high resistance (22 Ohms), this is alarming, although in ideal conditions it should work like that. This resistor stands on the RXC circuit - when receiving a packet, the transceiver sends a meander with a frequency of 125 MHz for synchronization, such a rather big frequency ... The requirements for wiring the board for such a frequency are appropriate, here it is a jamb, and then the road goes into the interlayer, and has a large stray capacitance , and a 22 ohm resistor is not sickly so it dampens the signal. In general, enough theory, in practice you need to replace (short-circuit) this resistor with a jumper, in the photo I noted. After this procedure, the LAN port came to life, and it works as it should. From NAS to SAMBA, it downloads 28MB / s, hooray !!!
Because You can add a lot of problem boxes to this header. And yes, you do everything at your own peril and risk ...
 ?s=000000005e2e0ddb5d5e8c0c00000000f3bd5b13d14ab2f9f1496f70a302f1e2
 ?s=000000005e2e0ddb5d5e8c09000000008c05c1fe7d1ff866a3c89ac24674e954
In the image you can see the place were it need to make a short circuit do fix that. Now we can only wait to someone make a video tutorial that shows how to make a save short circuit in there ( I don?t know if using a jumper will work).
other user report that this affect also 922x , not only b922.
"While I studied various forums, I realized that a problem with the LAN port occurs in both revisions since this part of the board is identical both there and there. It also depends on the connection speed, at 100Mbit / s everything will be perfect, because the clock frequency of only 25 MHz."
From user bezik66 in russian foruns:
Today I performed an operation to bypass the sprot according to the Beelink GT-King [Android] instructions (Post aquagenptz # 87947976). Super !!! IPTV, YouTube, HD videobox, start up, channels are switched, instantly! Although the Internet 100 decided to re-solder, and the result I was pleased !!! + the author flew away !!! So it helps not only on) gigabit internet but also on 100mb too. (anyone who has trembling hands is better not to climb !!!)
Bee-link should recall this batch of faulty boxs from all the users that have this problem.
|
|
Author: skein147
Post time 2019-07-26 12:14:48
