[ANN] Technobit HEX8A1 -240 GHs Coincraft A1 board- Finished production

[Gator-hex]

One of my Hex8A1 miners has a habit of crashing then it's really difficult to bring it back up.
I suspect it don't like running at 220/880 or any other low voltage.
It takes out both miners and the PSU. It could be the PSU I guess but they are running < 250W each and it's single rail 684W so something wierd going on.
Any tips? I'm guessing more volts less clock? But why should the fact it crashed make it hard to get booted again?
I thought it was dead, but it came back once, and now it's dead again.


Rebooted and only one miner comes back. Fans spin but heatsink stays cold.
Is it broken? Why does CGminer say 800mV when it should say 880mV? Could it be a software bug?
I've never had a miner act like this before. Help!

So I reset default 260/1000 rebooted again and now both Hex8A1 are MIA!
Fans are spinning, one board shows green+green light, the other show green+amber light.
Maybe I'm just asking too much from the PSU here, so lets try a lower volt again.


Ok so I try 200/900 and reboot again. The green+green light one starts hashing. Although it doesn't flash the lights like the Hex16B do.
It's all so random!


So I unplug the bad one and CGminer detects that, not mining seems to be a different state than not connected/powered.
Sadly when I plug it back in it still refuses to mine with it.

[iongchun]

Thanks for the delivery, Marto.
The fans are really noisy, I replace them with low CFM and quiet ones,
and run the HEX8A1 with only 175 in frequency and 870 in voltage parameter.

When connected to Linux machine, lsusb will show:
Bus 001 Device 017: ID 04d8:000a Microchip Technology, Inc. CDC RS-232 Emulation Demo

And it appears as a tty device:
crw-rw---- 1 root dialout 166, 1 2g  16 08:51 /dev/ttyACM2

Showed in dmesg:
[  161.103406] cdc_acm 1-1.1:1.0: This device cannot do calls on its own. It is not a modem.
[  161.103442] cdc_acm 1-1.1:1.0: ttyACM2: USB ACM device

Running with cgminer:

Code:

 cgminer version 3.12.3 - Started: [2014-02-15 21:11:05]
------------------------------------------------------------------------------------------
 (5s):151.1G (avg):175.5Gh/s | A:1696768  R:28928  HW:64  WU:38.3/m
 ST: 2  SS: 0  NB: 99  LW: 1891862  GF: 0  RF: 0
 Connected to 10.0.0.254 diff 64 with stratum as user 1X.....
 Block: 168ab84a...  Diff:2.62G  Started: [08:53:52]  Best share: 1.86M
------------------------------------------------------------------------------------------
 [P]ool management [S]ettings [D]isplay options [Q]uit
 HEX8 0: 175  870/ 872mV | 173.9G/175.5Gh/s | A:1696448 R:28928 HW:64/0.24% WU:38.3/m
------------------------------------------------------------------------------------------

While I also run bfgminer with other mining hardware, it seems bfgminer could cause the OS to hang when it probe the tty device.
This can be avoided with run HEX8A1's cgminer first (so the tty device would disappear), or plug the HEX8A1 after bfgminer runs.

Oh, and it runs well with p2pool, with less than 2% DOA, compared to 5~6% DOA for BFL or BitFury devices.

[Zich]

One of my Hex8A1 miners has a habit of crashing then it's really difficult to bring it back up.
I suspect it don't like running at 220/880 or any other low voltage.
It takes out both miners and the PSU. It could be the PSU I guess but they are running < 250W each and it's single rail 684W so something wierd going on.
Any tips? I'm guessing more volts less clock? But why should the fact it crashed make it hard to get booted again?
I thought it was dead, but it came back once, and now it's dead again.

Maybe you can try one board only with 260/1000 setting without other HEX16B and set "Hex8A1 Set ASIC difficulty to one" : "0".

You had:
 6 HEX16B x 5 A  = 30 A
 2 HEX8    x 17 A = 34 A set at 200/870

Total 64 A X 12 V = 768 watt

So i guess your PSU can not deliver the power required by the board

[Gator-hex]

One of my Hex8A1 miners has a habit of crashing then it's really difficult to bring it back up.
I suspect it don't like running at 220/880 or any other low voltage.
It takes out both miners and the PSU. It could be the PSU I guess but they are running < 250W each and it's single rail 684W so something wierd going on.
Any tips? I'm guessing more volts less clock? But why should the fact it crashed make it hard to get booted again?
I thought it was dead, but it came back once, and now it's dead again.

Maybe you can try one board only with 260/1000 setting without other HEX16B and set "Hex8A1 Set ASIC difficulty to one" : "0".

You had:
 6 HEX16B x 5 A  = 30 A
 2 HEX8    x 17 A = 34 A set at 200/870

Total 64 A X 12 V = 768 watt

So i guess your PSU can not deliver the power required by the board

The Hex16b's are on 2x 350W PSUs.
Only the Hex8A1 have been on the Thermaltake 730W

Anyway, one last try, put problem child on the Thermaltake alone, has both PCI-E power cables. I tried 100/1000 and default 260/1000 and 1/0...

...but it seems to be dead for good now. How does one get Marto to fix this?

I'll leave it disconnected overnight and see if it works in the morning, but has been a problem all day since it arrived. I think it's a DOA.
What does the orange LED mean?

[Zich]

The Hex16b's are on 2x 350W PSUs.
Only the Hex8A1 have been on the Thermaltake 730W

Anyway, one last try, put problem child on the Thermaltake alone, has both PCI-E power cables. I tried 100/1000 and default 260/1000 and 1/0...

...but it seems to be dead for good now. How does one get Marto to fix this?

I'll leave it disconnected overnight and see if it works in the morning, but has been a problem all day since it arrived. I think it's a DOA.
What does the orange LED mean?

I see, so different PSU on one host.

[Gator-hex]

I see, so different PSU on one host.

Yeah, I figure out when powering 6x Hex16b, if you use a power protected hub, multiple PSU works fine.

I noticed some lose PolyFuse legs on this bad Hex8A1. Maybe they got broken in packing or shipping but the one on the right looks like a horrible solder joint. Not sure if this would stop it hashing though. I'm not an expert at electronics. Wish I was!

[Zich]

I see, so different PSU on one host.

Yeah, I figure out when powering 6x Hex16b, if you use a power protected hub, multiple PSU works fine.

I noticed some lose Capacitor legs on this bad Hex8A1. Maybe they got broken in packing or shipping but the one on the right looks like a horrible solder joint. Not sure if this would stop it hashing though. I'm not an expert at electronics. Wish I was!

Look like the voltage stand on 800mv even with 1000mv setting.

Actually my board also had same broken capacitor leg but cause by my self  .
But i just join it with his friend there( the one you point by single red arrow)

[stiffla]

I have a question.

I have ~0.01% HW errors, but my Rejects or Refuses are quite high.



How can I get rid of them? Do I use the wrong clock setting?
Does this cause my variable btc reward at my pool, because normally it was really stable.

[Zich]

I have a question.

I have ~0.01% HW errors, but my Rejects or Refuses are quite high.



How can I get rid of them? Do I use the wrong clock setting?
Does this cause my variable btc reward at my pool, because normally it was really stable.

Your reject ratio only 0.095%
That's not high but super low

[loshia]

I have a question.

I have ~0.01% HW errors, but my Rejects or Refuses are quite high.



How can I get rid of them? Do I use the wrong clock setting?
Does this cause my variable btc reward at my pool, because normally it was really stable.

As zich said you are golden

[Zich]

After playing with various setting, this is my final result 

[loshia]

After playing with various setting, this is my final result 

Locking good zich
congrats

[Zich]

After playing with various setting, this is my final result 

Locking good zich
congrats

Yeah, thanks 
After two days in same room with HEX8, i become familiar with the noise 

[stiffla]

After playing with various setting, this is my final result 

Locking good zich
congrats

Yeah, thanks 
After two days in same room with HEX8, i become familiar with the noise 

Thx for your answers guys
So Zich, nice settings, do you use any extra cooling or can you touch your board longer than 5 sec?

[Zich]

Thx for your answers guys
So Zich, nice settings, do you use any extra cooling or can you touch your board longer than 5 sec?

Nope 
The heatsink is very hot to touch 

[eiabea]

Thx for your answers guys
So Zich, nice settings, do you use any extra cooling or can you touch your board longer than 5 sec?

Nope 
The heatsink is very hot to touch 

does the miner have a overheating shutdown?

[Zich]

Thx for your answers guys
So Zich, nice settings, do you use any extra cooling or can you touch your board longer than 5 sec?

Nope 
The heatsink is very hot to touch 

does the miner have a overheating shutdown?

No, there is no temperature sensor on board.
You need to set the board at appropriated setting which don't make the board over temperature 

[eiabea]

Thx for your answers guys
So Zich, nice settings, do you use any extra cooling or can you touch your board longer than 5 sec?

Nope 
The heatsink is very hot to touch 

does the miner have a overheating shutdown?

No, there is no temperature sensor on board.
You need to set the board at appropriated setting which don't make the board over temperature 

the question is what isn't too hot? ^^

[Zich]

Thx for your answers guys
So Zich, nice settings, do you use any extra cooling or can you touch your board longer than 5 sec?

Nope 
The heatsink is very hot to touch 

does the miner have a overheating shutdown?

No, there is no temperature sensor on board.
You need to set the board at appropriated setting which don't make the board over temperature 

the question is what isn't too hot? ^^

you can decide it your self 

[Gator-hex]

I see, so different PSU on one host.

Yeah, I figure out when powering 6x Hex16b, if you use a power protected hub, multiple PSU works fine.

I noticed some lose Capacitor legs on this bad Hex8A1. Maybe they got broken in packing or shipping but the one on the right looks like a horrible solder joint. Not sure if this would stop it hashing though. I'm not an expert at electronics. Wish I was!

Look like the voltage stand on 800mv even with 1000mv setting.

Actually my board also had same broken capacitor leg but cause by my self  .
But i just join it with his friend there( the one you point by single red arrow)

Did yours stop working with broken capacitor leg? I was really careful with unpacking and handling, I know caps are always the weakest electronic part.
I do have a soldering iron, if I knew what needed soldering to what I could maybe fix it, but it's unclear where these legs were attached.
I pushed the legs to touch the nearest micro resistor. It powered up with it's sister 220/880 ran for 5 mins, then crashed the power supply.

The bad one had green and orange light as always. The good one had two green lights and a red light after crashing.
After rebooting the good one went back to two green lights and started working. The bad one refused to reboot again.

I think there is some charge building up in the bad one after 5mins which causes the PSU to trip failsafe, then that charge is still held for some time, which stops it rebooting back up.

I'm going to give it one last try and run it on it's own TP-Link at 260/1000 next time it works. How do I upgrade an old TP-Link from 0_1_0 to 0_2_0? Do I flash the "factory.bin" file or do I use the "sysupgrade.bin" file. It's not in factory condition anymore, so I guess it's an upgrade, but I've only seen people mention the factory file?

Bad Boy & Good Girl... one hashes 210GH/220W quietly... the other just don't hash for more than a few minutes!