Cairnsmore1 - Quad XC6SLX150 Board

[yohan]

yup


Uploaded with ImageShack.us

If you are running the default bitstreams (makomk) then your you shoulf be using Icarus worker in MPBM. The Cairnsmore setting is for hashvoodoo bitstreams.

[1714072182]

1714072182

[1714072182]

1714072182

[yohan]

We are often asked how high can we stack. Does this answer the question?

[gyverlb]

Note for people with USB stability problems: try a piece of clear tape covering the 5V connector (on the right when you look at the connectors from above) in the plug you connect to your USB hub or server, I was fed up trying to work out which combination of power supply and USB hub/port would work and it seems it helps (no disconnect for 2 days now) and is easily reversible.

Make it longer than the full length of the connector and make it wrap around the exterior so that inserting the plug won't push it inside (sorry no picture, you'll have to try it for yourself).

[glowkeeper]

I'm trying to run 9 Cairnsmore's off a rPi, using cgminer and the bitstream they came shipped with. They all work fine, but I can only seem to get 8 of the 9 running SIMULTANEOUSLY; /dev/ttyUSB10 /dev/ttyUSB11 just won't work, giving the ol' Icarus detect problem. I'm Scratching my head a bit. Any pointers on how I might resolve it?

[elchorizo]

Note for people with USB stability problems: try a piece of clear tape covering the 5V connector (on the right when you look at the connectors from above) in the plug you connect to your USB hub or server, I was fed up trying to work out which combination of power supply and USB hub/port would work and it seems it helps (no disconnect for 2 days now) and is easily reversible.

Make it longer than the full length of the connector and make it wrap around the exterior so that inserting the plug won't push it inside (sorry no picture, you'll have to try it for yourself).

I'm totally not understanding your post?

[glowkeeper]

I'm trying to run 9 Cairnsmore's off a rPi, using cgminer and the bitstream they came shipped with. They all work fine, but I can only seem to get 8 of the 9 running SIMULTANEOUSLY; /dev/ttyUSB10 /dev/ttyUSB11 just won't work, giving the ol' Icarus detect problem. I'm Scratching my head a bit. Any pointers on how I might resolve it?

Actually, just checked dmesg properly, which shows:

[ 1900.005211] ftdi_sio ttyUSB8: FTDI USB Serial Device converter now disconnected from ttyUSB8
[ 1900.005329] ftdi_sio 1-1.2.6:1.0: device disconnected
[ 1900.005997] ftdi_sio ttyUSB9: FTDI USB Serial Device converter now disconnected from ttyUSB9
[ 1900.006100] ftdi_sio 1-1.2.6:1.1: device disconnected
[ 1900.014160] ftdi_sio ttyUSB10: FTDI USB Serial Device converter now disconnected from ttyUSB10
[ 1900.014268] ftdi_sio 1-1.2.6:1.2: device disconnected
[ 1900.014981] ftdi_sio ttyUSB11: FTDI USB Serial Device converter now disconnected from ttyUSB11
[ 1900.015077] ftdi_sio 1-1.2.6:1.3: device disconnected

...the offending card

But as I said, I get any card to do this, it's not tied to one specifically, but seems to be just tied to ttyUSB8-11. Any ideas?

[elchorizo]

I'm trying to run 9 Cairnsmore's off a rPi, using cgminer and the bitstream they came shipped with. They all work fine, but I can only seem to get 8 of the 9 running SIMULTANEOUSLY; /dev/ttyUSB10 /dev/ttyUSB11 just won't work, giving the ol' Icarus detect problem. I'm Scratching my head a bit. Any pointers on how I might resolve it?

Actually, just checked dmesg properly, which shows:

[ 1900.005211] ftdi_sio ttyUSB8: FTDI USB Serial Device converter now disconnected from ttyUSB8
[ 1900.005329] ftdi_sio 1-1.2.6:1.0: device disconnected
[ 1900.005997] ftdi_sio ttyUSB9: FTDI USB Serial Device converter now disconnected from ttyUSB9
[ 1900.006100] ftdi_sio 1-1.2.6:1.1: device disconnected
[ 1900.014160] ftdi_sio ttyUSB10: FTDI USB Serial Device converter now disconnected from ttyUSB10
[ 1900.014268] ftdi_sio 1-1.2.6:1.2: device disconnected
[ 1900.014981] ftdi_sio ttyUSB11: FTDI USB Serial Device converter now disconnected from ttyUSB11
[ 1900.015077] ftdi_sio 1-1.2.6:1.3: device disconnected

...the offending card

But as I said, I get any card to do this, it's not tied to one specifically, but seems to be just tied to ttyUSB8-11. Any ideas?

I had the same issue but Yohan sent me a PM and the most important instruction he gave me was as follows:

1. Power up the units.
2. Power up your PC, miner, etc.
3. Plug in the units to the USB hub/turn the USB hub on. Wait until all the units have the double orange LED.
4. Run the modprobe. And this part is important. PHYSICALLY LOOK at the units and make sure after you run the modprobe that they all activate. It can take a couple minutes.

This was my problem, they were in another room so I would run the modprobe, wait 2 seconds and start mining and it was super flaky.

I even had the modprobe in my mining script followed by a sleep 2;

I never saw anywhere that said the USB might take a while to pick them up.

So make sure you do this, this is the most important step.

Then, do #5, mine with your program. Only after you verify step #4 and that all devices have the proper lights and have had time to settle will the miner pick them all up.

Hope this helps.

[jml]

I have noticed that while using BFGminer 3.0.2, one of my CM1's would get detected and my Ztex would be registered. The other CM1 would not be registered even after executing modprobe. I found out that by removing Ztex first during boot up and allowing the CM1 script to run with modprobe as root, I can see that the are all registered. Plugging in the Ztex would be picked up afterwards.

[glowkeeper]

I'm trying to run 9 Cairnsmore's off a rPi, using cgminer and the bitstream they came shipped with. They all work fine, but I can only seem to get 8 of the 9 running SIMULTANEOUSLY; /dev/ttyUSB10 /dev/ttyUSB11 just won't work, giving the ol' Icarus detect problem. I'm Scratching my head a bit. Any pointers on how I might resolve it?

Actually, just checked dmesg properly, which shows:

[ 1900.005211] ftdi_sio ttyUSB8: FTDI USB Serial Device converter now disconnected from ttyUSB8
[ 1900.005329] ftdi_sio 1-1.2.6:1.0: device disconnected
[ 1900.005997] ftdi_sio ttyUSB9: FTDI USB Serial Device converter now disconnected from ttyUSB9
[ 1900.006100] ftdi_sio 1-1.2.6:1.1: device disconnected
[ 1900.014160] ftdi_sio ttyUSB10: FTDI USB Serial Device converter now disconnected from ttyUSB10
[ 1900.014268] ftdi_sio 1-1.2.6:1.2: device disconnected
[ 1900.014981] ftdi_sio ttyUSB11: FTDI USB Serial Device converter now disconnected from ttyUSB11
[ 1900.015077] ftdi_sio 1-1.2.6:1.3: device disconnected

...the offending card

But as I said, I get any card to do this, it's not tied to one specifically, but seems to be just tied to ttyUSB8-11. Any ideas?

I had the same issue but Yohan sent me a PM and the most important instruction he gave me was as follows:

1. Power up the units.
2. Power up your PC, miner, etc.
3. Plug in the units to the USB hub/turn the USB hub on. Wait until all the units have the double orange LED.
4. Run the modprobe. And this part is important. PHYSICALLY LOOK at the units and make sure after you run the modprobe that they all activate. It can take a couple minutes.

This was my problem, they were in another room so I would run the modprobe, wait 2 seconds and start mining and it was super flaky.

I even had the modprobe in my mining script followed by a sleep 2;

I never saw anywhere that said the USB might take a while to pick them up.

So make sure you do this, this is the most important step.

Then, do #5, mine with your program. Only after you verify step #4 and that all devices have the proper lights and have had time to settle will the miner pick them all up.

Hope this helps.

It did! Based on what you told me, I have all 9 fpga's working! Yay!

So the essence of my working solution is that I power everything up, let it all settle, then last of all I physically plug in the Pi and let it boot up. Lo and behold, my mining software finds the cards correctly.....

For some reason, a soft power cycle on the Pi via "shutdown -r now", with everything else running, doesn't work; I only get 7 or 8 of the 9 cards mining. But whatever, for now, I'm celebrating the whole rig doing its stuff

[glowkeeper]

So the essence of my working solution is that I power everything up, let it all settle, then last of all I physically plug in the Pi and let it boot up. Lo and behold, my mining software finds the cards correctly.....

For some reason, a soft power cycle on the Pi via "shutdown -r now", with everything else running, doesn't work; I only get 7 or 8 of the 9 cards mining. But whatever, for now, I'm celebrating the whole rig doing its stuff

ps. I too do a modprobe followed by a sleep 10 in the mining script that I run from init.d.....

[elchorizo]

So the essence of my working solution is that I power everything up, let it all settle, then last of all I physically plug in the Pi and let it boot up. Lo and behold, my mining software finds the cards correctly.....

For some reason, a soft power cycle on the Pi via "shutdown -r now", with everything else running, doesn't work; I only get 7 or 8 of the 9 cards mining. But whatever, for now, I'm celebrating the whole rig doing its stuff

ps. I too do a modprobe followed by a sleep 10 in the mining script that I run from init.d.....

I think the time you need to sleep depends on how many units you have. With 6 units, they don't all settle in completely even in 10 seconds. It takes more like 30....

[shawtux]

hi!
had to reinstall on my pi due to a user password error...
did all the steps i supposedly did... but i get the following error on all cm1

Traceback (most recent call last):
  File "/home/pi/Modular-Python-Bitcoin-Miner-testing/modules/theseven/icarus/icarusworker.py", line 203, in main
    if not self.checksuccess: raise Exception("Timeout waiting for validation job to finish")
Exception: Timeout waiting for validation job to finish

any ideas?



edit.
no worries wrong dip switch settings 

[LazyOtto]

edit.
no worries wrong dip switch settings  

<lol>

Yep. 'User error' is surely 80% of the problems of life.  

[glowkeeper]

So the essence of my working solution is that I power everything up, let it all settle, then last of all I physically plug in the Pi and let it boot up. Lo and behold, my mining software finds the cards correctly.....

For some reason, a soft power cycle on the Pi via "shutdown -r now", with everything else running, doesn't work; I only get 7 or 8 of the 9 cards mining. But whatever, for now, I'm celebrating the whole rig doing its stuff

ps. I too do a modprobe followed by a sleep 10 in the mining script that I run from init.d.....

I think the time you need to sleep depends on how many units you have. With 6 units, they don't all settle in completely even in 10 seconds. It takes more like 30....

Thanks again. modprobe; sleep 60 works a treat

[SnitraM]

The 0x6014 is what I don't have the full details for yet for Cairnsmore1 but I presume it will be the same as BFL but with a different baud rate.

My Cairnsmores are 0x8350. Do you still want me test something?

[shawtux]

where can i find stock firmware for quad?
makomk? this one works with cgminer right?

i'm asking because bfgminer detects it but cgminer doesnt

[tigerbit]

where can i find stock firmware for quad?
makomk? this one works with cgminer right?

i'm asking because bfgminer detects it but cgminer doesnt

That won't help, its not a firmware issue.  One of the cgminer devs called kano has included some new USB direct code in cgminer as of 3.2 onwards IIRC - I don't fully understand why, but it means that the CM1s won't work any longer.

Either go for cgminer 3.1.x which uses the old USB serial drivers, or stick with BFGMiner.  BFG works well for me and I haven't noticed much difference.

[shawtux]

hmmm that the rpc calls of BFG dont have the current hashrate.
a friend of mine is currently using cgminer with CM1 so i dont know whats up. i currently run 3.3.1
my cm1 is second hand.

before running do i necessarily need to run modprobe? (i run it on boot)


edit:
duh!! its right there... testing 3.3.1 and not 3.1 lol... as soon as i get home i'll try it out.

[Milan77]

Yohan, is there a cut-off price for Cairnsmore1 ?

[yohan]

Yohan, is there a cut-off price for Cairnsmore1 ?

Did you mean the discount structure or when Bitcoin return falls to zero?

Discount structure is still in place.

If you mean Bitcoin return then that depends on your electricity cost and the Bitcoin exchange rate if you are measuring your return in flat currency, Bitcoin currency, or social gain. On the CM1s that I am running personally I estimate on my electricity cost and current exchange rate I will remain profitable until the network reaches to about 1500 TH/s if indeed it reaches that number. What will happen to the network size and exchange rates are a pure guess. I think we might hit an equlibrium or plateau on network size after August/September whilst people take stock of the expected market changes. That might last for 6 months. I expect at that some of the quick buck crowd, and any remaining GPU based miners, will dissappear from the network and it will settle down to something a little more normal. It is going to be interesting at that point as well because transaction processing could be very quick and that will further increase the use of Bitcoin maybe. I expect a period of gentle oscillation between usage growth and network growth.

We are going to offer a memory add-on to CM1 to help a potential Litecoin implementation running on CM1 and I expect if we do that and then either us or someone else does the Litecoin IP profit lifetime could be extended. It is only an well educated guess that the memory add-on will help a Litecoin mining implementation so don't use that fact for any financial analysis.