Get the 300 pieces from factory:
The alpha version have tons of bugs, be happy to catch them 😀
From my first test, there are two main problems:
1. the board is not stable through 3.3V~6V port, many oops when system boot up, but through 3.3V pin(input set to 3.1V~3.2V), the system works normal.
2. From computer/phone we are able to find VoCore wifi hotpot, but after connect to it, it shows a warning, can not ssh to the board. 
First problem, I think it is PCB problem or design problem. But the design is copied from v0.5(some wire changed, I must compare them. Never be too careful, specially on hardware), and v0.5 is normal. So high chance to be PCB problem. Might be crosstalk? From my simulation the max crosstalk is 84mV(3.0V, 1.0ns), in most situation the max allowed value should be 150mV.
Second problem, just no idea, might caused by first problem. Same firmware works well on v0.5 but not this one.
From the factory engineer, he said two points must be improved in next PCB version:
1. V-CUT border is too closed to each other, some pin hole might be broken after SMT.
2. V-CUT is not good for such small board, the copper is easy to be stretched when they cut it into pieces, so it will effect signal.
I thought there might be some small problem such as wifi signal in the alpha version but I never thought these problems are so big…None of that happens to my test version v0.3 or v0.5.
I am really sorry, I planed this version should be at least as good as v0.5… I will also ship a small gift to all the alpha tester(USB<=>TTL/UART device), hope your testers will be a little happier. 🙂
And the boards will be shipped in this week. I am so limited, please help me make it better, really appreciate!
>> 1. but through 3.3V pin(input set to 3.1V~3.2V), the system works normal.
It is a terrible problem. May be it is very simple. Check the LDO for the system. 90% the power design/soldering has problem. Whether you change certain type of chip?
>>2. From computer/phone we are able to find VoCore wifi hotpot, but after connect to it, it shows a warning, can not ssh to the board.
Which kind or warning？Could you ping with the larger package? It may related with 1. Small package working, but larger package is caused current, and then system crashed.
First solve the problem 1. It may be a small problem, but brings serious result.
LDO(in fact it is DCDC) is using the same chip, I will check 🙂
Any more clues as to what the problem could be?
I am trying to find out. Looks like it is PCB problem, it is able to boot to openwrt, but wifi still can not connect, the weird thing it is not signal problem…
It is hard to say PCB problem yet. If factory is no problem, this kind of PCB rarely meets problem.
Hi! Any news?
Now I lower the voltage to 3.2V, it works normal for over one day already. It has a high chance to be that DCDC problem. My BOM order is MT3410L but the reseller send me MT3410, might be that problem.
GPIO, ethernet works normal.
Oh cool, so joyhope was on the right track with the above comments.
Have you tried disconnecting the onboard DCDC circuit completely (cutting tracks or removing components) and powering the rest of the system externally with 3.3V to see if that fixes the instability/wifi problems?