LightningAsic Gridseed Controller Firmware V3

[mazedk]

So i got a new TP-Link WR703N in the mail and tried to figure out how to flash it with the LA firmware..

Its ended up with the unit flashing for around 12 secs, then having the LED off for 4-5 secs and finishing having the blue LED on constantly after.

Anyone have any ideas?

Hey mazedk,

I had the same problem. It turns out after flashing the first image the WR703N looses it's network settings completely and reverts back to DHCP.
To make it worse needed two or three fresh power-cycles before the web UI came up to flash the 2nd stage firmware.

Hope that helps!

   one4many

Thanks for the reply. I think i have powercycled it 20 times so far, with the same end result :/

What i did exactly was (prolly stupid) to try and flash fromthe chinese stock firmware to the english one of the TL-WR703N (according to this guilde: http://wr703n.blogspot.com/ )

That all worked fine. But then i couldnt install the LA firmware, so thought i needed to go to openwrt or so. Then i tried to revert to the stock firmware - as per the same guide.

Which has ended up with the previously mentioned result..

[1715102112]

1715102112

[1715102112]

1715102112

[sandor111]

So i got a new TP-Link WR703N in the mail and tried to figure out how to flash it with the LA firmware..

Its ended up with the unit flashing for around 12 secs, then having the LED off for 4-5 secs and finishing having the blue LED on constantly after.

Anyone have any ideas?

Hey mazedk,

I had the same problem. It turns out after flashing the first image the WR703N looses it's network settings completely and reverts back to DHCP.
To make it worse needed two or three fresh power-cycles before the web UI came up to flash the 2nd stage firmware.

Hope that helps!

   one4many

Thanks for the reply. I think i have powercycled it 20 times so far, with the same end result :/

What i did exactly was (prolly stupid) to try and flash fromthe chinese stock firmware to the english one of the TL-WR703N (according to this guilde: http://wr703n.blogspot.com/ )

That all worked fine. But then i couldnt install the LA firmware, so thought i needed to go to openwrt or so. Then i tried to revert to the stock firmware - as per the same guide.

Which has ended up with the previously mentioned result..

What you want to do is flash directly to the v1 bin from the Chinese firmware, Jack has done it thousands of times and it works.

[leoragraves666]

Try some software like WakeOnLan, that will scan the whole range of IPs. Maybe you will get lucky.

[Xer0]

switched to DOGE Pool, got first YAY:

[2014-04-16 23:13:34.705]
0: Got nonce 66667652, Hash <= Htarget!
[2014-04-16 23:13:34.747] 0: accepted: 3/3 (100.00%), (yay!!!)
[2014-04-16 23:13:35.413] 0: Got nonce 66675520, Hash <= Htarget!
[2014-04-16 23:13:35.433] New job_id: a15b Diff: 128
[2014-04-16 23:13:35.434] Stratum detected new block
[2014-04-16 23:13:35.472] New job_id: a15c Diff: 128
[2014-04-16 23:13:35.473] 0: accepted: 3/4 (75.00%), (booooo)

but then if falls back to doing nothing



miner runs fine on XP-PC with cgminer cm-blog edition
but still cant get it to work with the wr703n box


has it something to do with Fan  beeing on +5V USB?

[jcoin]

Any ideas what could cause this?




They've done this on v1 and v3 firmwares on the LA controller. Stable for while, then they go nuts. Some miners showing up twice with zero accepts. After a bunch of auto restarts they stabilize for a little. Power related maybe?

[Xer0]

finaly got it...



the trick was, to power 12v and 5v on at the exact same time
so i put both miner psu and controller wallwart on a switchable multi outlet

[poopypants]

Any ideas what could cause this?




They've done this on v1 and v3 firmwares on the LA controller. Stable for while, then they go nuts. Some miners showing up twice with zero accepts. After a bunch of auto restarts they stabilize for a little. Power related maybe?

That is weird, what does the log say when this happens? Do you use the restart dead miners?  Do you power your USB hubs?

[jcoin]

Any ideas what could cause this?



They've done this on v1 and v3 firmwares on the LA controller. Stable for while, then they go nuts. Some miners showing up twice with zero accepts. After a bunch of auto restarts they stabilize for a little. Power related maybe?

That is weird, what does the log say when this happens? Do you use the restart dead miners?  Do you power your USB hubs?

A few UART Write errors in the log then auto restart. I have it set to restart dead miners on 5 mins, it loops until it stabilizes. I have auto overclock on but it also did with v1. Hub is powered, it's the one that was included in the LA3M.

[sandor111]

New version coming up very soon, API will be included so you can monitor the miners anywhere you want.

[zyberguy]

So, do I understand it correctly, after upgrading to new fw. I cannot SSH to my controller.

[christian1980]

So, do I understand it correctly, after upgrading to new fw. I cannot SSH to my controller.

Correct. I think they are hiding something.
Even by connecting a serial Adapter to your hardware you cannot login.

The only thing you can do is to protect it with a firewall like the free sophos UTM and allow only the mining traffic.
I hope the developers here will wake up or one day a guy will read out the flash chip to extract the files and create a new image.

[rammy2k2]

So, do I understand it correctly, after upgrading to new fw. I cannot SSH to my controller.

Correct. I think they are hiding something.
Even by connecting a serial Adapter to your hardware you cannot login.

The only thing you can do is to protect it with a firewall like the free sophos UTM and allow only the mining traffic.
I hope the developers here will wake up or one day a guy will read out the flash chip to extract the files and create a new image.

they are protecting the source

[christian1980]

they are protecting the source

The sources are not included in a binary image.
Even the original chinese wiibox software is not blocking every access like that.
 

[bobby35ny]

Protecting the source, Protecting the Image, no SSH.
Who cares???
This software/controller works!!!!
Tried almost everything else and nothing compares!
-Bobby

[Ajeto]

Guys do you have any idea why this is happening?

[2014-04-19 00:09:47.221] 1: rejected 33/45 (73.33%) 219.4/2208.6 KH/s
[2014-04-19 00:09:47.221] DEBUG: reject reason: low difficulty share of 0
[2014-04-19 00:10:5.168] 7@4 850MHz: Got nonce cd096459, Hash <= Htarget!
[2014-04-19 00:10:5.168] 7@4: 544s until frequency adjusts to 875MHz
[2014-04-19 00:10:5.207] 7: rejected 33/46 (71.74%) 144.6/2280.7 KH/s
[2014-04-19 00:10:5.208] DEBUG: reject reason: low difficulty share of 0

Itried different pools:
- waffle: worked in the beginning, not anymore DEBUG: reject reason: low difficulty share of 0
- Clevermining: now working after last FW update (v3d)  DEBUG: reject reason: low difficulty share of 0
- LTC: 100% rejects (DEBUG: reject reason: high-hash)
- Doge: not working

First 12hrs it wokrd very well with Waffle pool
i even tried setting diff with entering "d=512" for password with waffle, but i cant seem to make it work. Suddenly I'm getting a ton of rejects (>50%) and very low HR (1600kH for 10 miners)
Settings:
General freq: 850
Auto overclicking
Using LA controller


The other 10 miners are still on LA controler version 1.0 and working OKayish on 850Mhz (10 miners @ 3300-3400 Kh).

Any ideas appreciated!

[poopypants]

Protecting the source, Protecting the Image, no SSH.
Who cares???
This software/controller works!!!!
Tried almost everything else and nothing compares!
-Bobby

I care but I understand why they are doing it this way, just wish their choice was different.  It is the most stable solution I have found for running these things though.  Just wish the code could be verified and would like to have the ability to fix things and customize on my own time frame.  Sadly I'll have to wait to use it again when its decided to open it up.  Sandor111 did a great job turning around the original version hack job, to what it is now and for that I commend him. 

[mazedk]

So i got a new TP-Link WR703N in the mail and tried to figure out how to flash it with the LA firmware..

Its ended up with the unit flashing for around 12 secs, then having the LED off for 4-5 secs and finishing having the blue LED on constantly after.

Anyone have any ideas?

Hey mazedk,

I had the same problem. It turns out after flashing the first image the WR703N looses it's network settings completely and reverts back to DHCP.
To make it worse needed two or three fresh power-cycles before the web UI came up to flash the 2nd stage firmware.

Hope that helps!

   one4many

Thanks for the reply. I think i have powercycled it 20 times so far, with the same end result :/

What i did exactly was (prolly stupid) to try and flash fromthe chinese stock firmware to the english one of the TL-WR703N (according to this guilde: http://wr703n.blogspot.com/ )

That all worked fine. But then i couldnt install the LA firmware, so thought i needed to go to openwrt or so. Then i tried to revert to the stock firmware - as per the same guide.

Which has ended up with the previously mentioned result..

What you want to do is flash directly to the v1 bin from the Chinese firmware, Jack has done it thousands of times and it works.

Yeah, trying to get around to that, just gotta get through to the webinterface of the router now

[marunga]

Hi,

I've been logging the IP addresses, other than the mining pools, that are being contacted by the Controller and this is what I got:

50.116.38.157:123 (UDP/NTP)
193.136.164.1:123 (UDP/NTP)
194.117.9.136:123 (UDP/NTP)
195.23.33.83:123 (UDP/NTP)
80.172.224.24:123 (UDP/NTP)
194.8.30.16:123 (UDP/NTP)

All connections are system attempts to set the clock, so no bad behaviour.

But I still like to see this software released as open source

regards,
m

[dc17]

Guys do you have any idea why this is happening?

[2014-04-19 00:09:47.221] 1: rejected 33/45 (73.33%) 219.4/2208.6 KH/s
[2014-04-19 00:09:47.221] DEBUG: reject reason: low difficulty share of 0
[2014-04-19 00:10:5.168] 7@4 850MHz: Got nonce cd096459, Hash <= Htarget!
[2014-04-19 00:10:5.168] 7@4: 544s until frequency adjusts to 875MHz
[2014-04-19 00:10:5.207] 7: rejected 33/46 (71.74%) 144.6/2280.7 KH/s
[2014-04-19 00:10:5.208] DEBUG: reject reason: low difficulty share of 0

Itried different pools:
- waffle: worked in the beginning, not anymore DEBUG: reject reason: low difficulty share of 0
- Clevermining: now working after last FW update (v3d)  DEBUG: reject reason: low difficulty share of 0
- LTC: 100% rejects (DEBUG: reject reason: high-hash)
- Doge: not working

First 12hrs it wokrd very well with Waffle pool
i even tried setting diff with entering "d=512" for password with waffle, but i cant seem to make it work. Suddenly I'm getting a ton of rejects (>50%) and very low HR (1600kH for 10 miners)
Settings:
General freq: 850
Auto overclicking
Using LA controller


The other 10 miners are still on LA controler version 1.0 and working OKayish on 850Mhz (10 miners @ 3300-3400 Kh).

Any ideas appreciated!

I am having the same problem with v3.

- Wafflepool: On v1 worked great with freq=850 and hashrate @ about 3400kh/s. Now on v3 I get 70+% rejects and if I get lucky around 1000kh/s
- Coinshift: same problem

I have tried everything (power cycling 10+ times, frequency anywhere from 700-900 and running only one of 10 units connected to the controller). The problem is always there...

Does anyone know how to downgrade to v1 or fix this problem on v3?

Any help is much appreciated. Thanks!!!

[sandor111]

Guys do you have any idea why this is happening?

[2014-04-19 00:09:47.221] 1: rejected 33/45 (73.33%) 219.4/2208.6 KH/s
[2014-04-19 00:09:47.221] DEBUG: reject reason: low difficulty share of 0
[2014-04-19 00:10:5.168] 7@4 850MHz: Got nonce cd096459, Hash <= Htarget!
[2014-04-19 00:10:5.168] 7@4: 544s until frequency adjusts to 875MHz
[2014-04-19 00:10:5.207] 7: rejected 33/46 (71.74%) 144.6/2280.7 KH/s
[2014-04-19 00:10:5.208] DEBUG: reject reason: low difficulty share of 0

Itried different pools:
- waffle: worked in the beginning, not anymore DEBUG: reject reason: low difficulty share of 0
- Clevermining: now working after last FW update (v3d)  DEBUG: reject reason: low difficulty share of 0
- LTC: 100% rejects (DEBUG: reject reason: high-hash)
- Doge: not working

First 12hrs it wokrd very well with Waffle pool
i even tried setting diff with entering "d=512" for password with waffle, but i cant seem to make it work. Suddenly I'm getting a ton of rejects (>50%) and very low HR (1600kH for 10 miners)
Settings:
General freq: 850
Auto overclicking
Using LA controller


The other 10 miners are still on LA controler version 1.0 and working OKayish on 850Mhz (10 miners @ 3300-3400 Kh).

Any ideas appreciated!

I am having the same problem with v3.

- Wafflepool: On v1 worked great with freq=850 and hashrate @ about 3400kh/s. Now on v3 I get 70+% rejects and if I get lucky around 1000kh/s
- Coinshift: same problem

I have tried everything (power cycling 10+ times, frequency anywhere from 700-900 and running only one of 10 units connected to the controller). The problem is always there...

Does anyone know how to downgrade to v1 or fix this problem on v3?

Any help is much appreciated. Thanks!!!

Should be fixed in next version, which will be released today.