HASHRA - LUNAR SERIES - ASIC SCRYPT MINERS

[wellfried]

I finally got my y Lunar Lander today and it isn't working. I tried with a laptop (downloaded the software), and with the RaspPi that is included (which I had to power with another cable because it wouldn't join the internet - thanks to the thread for heads up on that); and on both *AS SOON* as CGminer starts, the unit shuts down. I get this error in the log on the Pi:

Icarus Detect: Test failed at /dev/ttyUSB0: get 00000000, should: 00038d26

Not feeling great about this.

My PSU is rated platinum and at 1300W - pretty sure that's not the problem. I have each plug isolated, so I don't think that is the problem.  Since the error is USB related, I tried swapping out the USB cable - no dice.

I am open to suggestions.

[jjj0923]

Don't forget LIGHTNINGASIC 28nm Scrypt Blade 30MH/S @ 260W

OMG those guys are a nightmare - there are people who ordered over a month ago from those guys, were promised 5 day delivery and still don't have their units - it's why I bought Hashra

thats a lie dude, i ordered from them and got my order in time, i have seen u post same shit everywhere but u never bought.
LA worked just fine for me

you can read all the complaints here. I was ready to order from them - they promised me 5 day delivery. I'm glad I did a little research because people waited for at least 3 weeks go thet their units - were promised money for lack of mining - never got the money either.

it's all here...

I really wanted one of their units but could not bring myself to order after I learned of their deceptive statements. I would have been fine if they had simply told me they're unable to deliver for a few weeks, but don;t tell me 5 day delivery knowing full well that you can't deliver - that is called a lie. I really tried to talk myself into it but the more I researched the more skeptical I got. I am sure that some of their most recent customers did get their units in the time frame promised, but I lost confidence in their statements and they were clearly shipping the A2's in the time frame promised but not the 30mh unit which is what I wanted.

https://bitcointalk.org/index.php?topic=421921.msg7188265#msg7188265

[jjj0923]

I finally got my y Lunar Lander today and it isn't working. I tried with a laptop (downloaded the software), and with the RaspPi that is included (which I had to power with another cable because it wouldn't join the internet - thanks to the thread for heads up on that); and on both *AS SOON* as CGminer starts, the unit shuts down. I get this error in the log on the Pi:

Icarus Detect: Test failed at /dev/ttyUSB0: get 00000000, should: 00038d26

Not feeling great about this.

My PSU is rated platinum and at 1300W - pretty sure that's not the problem. I have each plug isolated, so I don't think that is the problem.  Since the error is USB related, I tried swapping out the USB cable - no dice.

I am open to suggestions.

well - don't use their power cable to power the pi - myself and other have had problems. In addition the ethernet cable they supplied did not work. (at least for me and I got two units). In addition make sure you're not plugging in the additional 2 wire plug from the pcie cable into the back of the unit.

send a message to support@hashra.com - their tech guy is very responsive.

if you can sssh - tail the /var/log/syslog file (it's also displayed on the diagnostic area of the screen)

[wellfried]

I finally got my y Lunar Lander today and it isn't working. I tried with a laptop (downloaded the software), and with the RaspPi that is included (which I had to power with another cable because it wouldn't join the internet - thanks to the thread for heads up on that); and on both *AS SOON* as CGminer starts, the unit shuts down. I get this error in the log on the Pi:

Icarus Detect: Test failed at /dev/ttyUSB0: get 00000000, should: 00038d26

Not feeling great about this.

My PSU is rated platinum and at 1300W - pretty sure that's not the problem. I have each plug isolated, so I don't think that is the problem.  Since the error is USB related, I tried swapping out the USB cable - no dice.

I am open to suggestions.

well - don't use their power cable to power the pi - myself and other have had problems. In addition the ethernet cable they supplied did not work. In addition make sure you're nit plugging in the additional 2 wire plug from the pcie cable.

send a message to support@hashra.com - their tech guy is very responsive.

Thanks - I did take the forum's advice and used a different power cable for the Pi, so thanks for that.  I am using my own ethernet cable, and I can browse to the Controla interface.  And no, I am not plugging in the +2 on the PCI-E connections.  The thing powers up fine... makes lots of noise and even some heat... but then instantly shuts down once I ask to to do some mining...

[davecoin]

At clock speed 291, LL hashes about 12.8mh/s, 444 watts, still 6% HW errors.

Also, < 1% rejects on clevermining.

[jjj0923]

if it's making heat then it's mining....

what's in the syslog file?

can you tail it and post the results or the log file on the screen?

[wellfried]

if it's making heat then it's mining....

what's in the syslog file?

can you tail it and post the results or the log file on the screen?

Here is a little more on the error (below).  Physically, it seems that once I get this error, the unit shuts down.  so the fans shut off, the PSU powers off, and I have to cycle everything to get it back up and running.

Jun 10 21:11:00 cgminer: Started cgminer 3.1.1

Jun 10 21:11:01 cgminer: Icarus Detect: Test failed at /dev/ttyUSB0: get 00000000, should: 00038d26

Jun 10 21:11:01 cgminer: All devices disabled, cannot mine!

Jun 10 21:11:01 kernel: [   80.162232] usb 1-1.2: USB disconnect, device number 4

Jun 10 21:11:01 kernel: [   80.162905] cp210x ttyUSB0: cp210x converter now disconnected from ttyUSB0

Jun 10 21:11:01 kernel: [   80.162985] cp210x 1-1.2:1.0: device disconnected

[geetash]

At clock speed 291, LL hashes about 12.8mh/s, 444 watts, still 6% HW errors.

Also, < 1% rejects on clevermining.

6% HW errors =  800~kH/s waste

[r0ach]

At clock speed 291, LL hashes about 12.8mh/s, 444 watts, still 6% HW errors.

Also, < 1% rejects on clevermining.

I don't have a power meter, could you try at 10.4mh to see what the power use is compared to gridseed blades?  And are you using original firmware, or the upgraded one availabe?  I haven't upgraded myself yet because jjj0923 said there was some kind of problem with it.

[davecoin]

I have to leave to work soon, but my napkin math guess is somewhere around 350 watts.  I'll try to mess with it tomorrow and post results again.

[r0ach]

I have to leave to work soon, but my napkin math guess is somewhere around 350 watts.  I'll try to mess with it tomorrow and post results again.

Can't really use napkin math for this because gridseed blade power use isn't even close to linear for overclocking.  I assume the Hashra chips are the same way and they are shipped to us already in that outside of efficiency power band state.

[jjj0923]

if it's making heat then it's mining....

what's in the syslog file?

can you tail it and post the results or the log file on the screen?

Here is a little more on the error (below).  Physically, it seems that once I get this error, the unit shuts down.  so the fans shut off, the PSU powers off, and I have to cycle everything to get it back up and running.

Jun 10 21:11:00 cgminer: Started cgminer 3.1.1

Jun 10 21:11:01 cgminer: Icarus Detect: Test failed at /dev/ttyUSB0: get 00000000, should: 00038d26

Jun 10 21:11:01 cgminer: All devices disabled, cannot mine!

Jun 10 21:11:01 kernel: [   80.162232] usb 1-1.2: USB disconnect, device number 4

Jun 10 21:11:01 kernel: [   80.162905] cp210x ttyUSB0: cp210x converter now disconnected from ttyUSB0

Jun 10 21:11:01 kernel: [   80.162985] cp210x 1-1.2:1.0: device disconnected

I assume you've tried both usb ports on the pi?

once it's running and you get the error message - disconnect the usb cable and reconnect it - see if the pi detects it?

[brian_23452]

At clock speed 291, LL hashes about 12.8mh/s, 444 watts, still 6% HW errors.

Also, < 1% rejects on clevermining.

I don't have a power meter, could you try at 10.4mh to see what the power use is compared to gridseed blades?  And are you using original firmware, or the upgraded one availabe?  I haven't upgraded myself yet because jjj0923 said there was some kind of problem with it.

I took one of the Zeus 1.3Mh/s machines down to stock speed of 280mhz (the default OC is 328 I believe) and it was still pulling 50 watts from the wall on a platinum PSU.  Consider that 8 gridseed blades pull 737 watts from the wall on the same PSU yet will do 45.6Mh/z at 838mhz and you see the problem.  What would it take on one of these things, around 1600ish from the wall to get similar performance?  It isn't a clock issue, or a firmware issue, or a configuration issue the chips in these things simply pull way too much power.  There is no way around that. 

[Brassguy]

You would need to lower the voltage going to the chips to find the most efficient ratio of power input to hashrate output... It seems efficiency is really becoming the name of the game. Ahhh the old days when only the hashrate mattered.... moar moar!!!

[brian_23452]

You would need to lower the voltage going to the chips to find the most efficient ratio of power input to hashrate output... It seems efficiency is really becoming the name of the game. Ahhh the old days when only the hashrate mattered.... moar moar!!!

I'm guessing their design was just messed up.  Consider, One of their 1.3Mh/s units was originally slated to do 1Mh/s at 12 watts, at 280Mhz.  The finished product does 1.3Mh/s at 328Mhz and consumes 45ish watts.  Underclocking it to 280Mhz or even lower does, at least in my experiments, virtually nothing to reduce power consumption (maybe 1 watt).  Of course undervolting it would reduce consumption but I have to assume they tried that and were not able to get it stable at anywhere near 1Mh/s, or else they would have done that to begin with. 
Of course, this is the Zeus machines, but the Hashra devices are using the same chips just in different multiples, so outside of 1-2 percent differences they have the exact same issue. 

[r0ach]

You would need to lower the voltage going to the chips to find the most efficient ratio of power input to hashrate output...

Yea, lowering voltage without changing any other variable on GPU mining rigs got large results.  I can't remember the numbers, but I think it was something like 20-30% less power use.  I would be very surprised if you couldn't get 300w or less power use out of these at 10.4mh on the launcher with voltage lowered.  The question is, will Hashra provide an easy solution to do so in the future if people want to keep these plugged in a long time.

[samsonn25]

Thats worse than the BFL mess up with the power on their 1st gens

[r0ach]

1.2v is default on Gridseed blade, what's default on hashra?

[anarchoatheist]

Is there any way to turn the fan down on these? Can you please work on the decibel levels on the next version. It kinda sounds like a wind tunnel. I think maybe its the design of the heatsinc. The air passing over the fins creates some kind of resonance. I have been doing my mining at a friends house because he has free electricity. I brought this asic over and hooked it up and he said get that thing out of here. I need to find a way to dampen the noise, so I can get some free electricity.

[iikun]

Is there any way to turn the fan down on these? Can you please work on the decibel levels on the next version. It kinda sounds like a wind tunnel. I think maybe its the design of the heatsinc. The air passing over the fins creates some kind of resonance. I have been doing my mining at a friends house because he has free electricity. I brought this asic over and hooked it up and he said get that thing out of here. I need to find a way to dampen the noise, so I can get some free electricity.

I'm thinking to switch the fan to a much quieter but still high flow fan I have at home but no idea of the windflow of the stock fan. I wonder if you're right about the fins though, that would be a real pain as there's no way to get around that obviously.
Not sure about risking $2,000 on switching the fan tho...