[Work in progess] Burnins Avalon Chip to mining board service

[alexuk]

Tried cgminer 3.2, no change :/

Tried just 2 boards / 3 , no combination seems to work....

Tried the horizontal CAN-BUS ports, no change...


put my multimeter up to the jumpers, it showed no resistance over the jumpers in the direction of the photo example on the website (front to back), however, in the other direction (left to right), resistance was there .... moved the jumpers into that configuration, and no change......

Code:

Aug 17 10:58:06 sol kernel: [431166.353607] usb 6-2: new full-speed USB device number 34 using uhci_hcd
Aug 17 10:58:06 sol mtp-probe: checking bus 6, device 34: "/sys/devices/pci0000:00/0000:00:1d.0/usb6/6-2"
Aug 17 10:58:06 sol mtp-probe: bus: 6, device: 34 was not an MTP device
Aug 17 10:58:06 sol kernel: [431166.523836] generic-usb 0003:04D8:003C.000D: hiddev0,hidraw0: USB HID v1.11 Device [Microchip Technology Inc. USB HID Bootloader] on usb-0000:00:1d.0-2/input0
Aug 17 10:58:18 sol kernel: [431178.140626] usb 6-2: USB disconnect, device number 34
Aug 17 10:58:19 sol kernel: [431179.372112] usb 6-2: new full-speed USB device number 35 using uhci_hcd
Aug 17 10:58:19 sol mtp-probe: checking bus 6, device 35: "/sys/devices/pci0000:00/0000:00:1d.0/usb6/6-2"
Aug 17 10:58:19 sol mtp-probe: bus: 6, device: 35 was not an MTP device
Aug 17 10:58:19 sol kernel: [431179.536550] ftdi_sio 6-2:1.0: FTDI USB Serial Device converter detected
Aug 17 10:58:19 sol kernel: [431179.536585] usb 6-2: Detected FT232RL
Aug 17 10:58:19 sol kernel: [431179.536588] usb 6-2: Number of endpoints 2
Aug 17 10:58:19 sol kernel: [431179.536591] usb 6-2: Endpoint 1 MaxPacketSize 64
Aug 17 10:58:19 sol kernel: [431179.536593] usb 6-2: Endpoint 2 MaxPacketSize 64
Aug 17 10:58:19 sol kernel: [431179.536596] usb 6-2: Setting MaxPacketSize 64
Aug 17 10:58:19 sol kernel: [431179.538614] usb 6-2: FTDI USB Serial Device converter now attached to ttyUSB0

This *looks* ok .... stumped....

I would take photos, but they look exactly like the the website + other photos here,

Tried my spare can-bus cable .... no change ....

Code:

root@sol:~/cgminer# ./cgminer -n
 [2013-08-17 11:09:09] USB all: found 9 devices - listing known devices
.USB dev 0: Bus 6 Device 38 ID: 0403:6001
  Manufacturer: 'Burnin Electronics'
  Product: 'BitBurner'
 [2013-08-17 11:09:09] 1 known USB devices
root@sol:~/cgminer#

[kano]

A completely aside comment ... just in case you didn't spot the obvious
Until the problem is sorted, you can of course mine with 4 independent BTBs with 4 USB cables

The later firmware is supposed to report a version number, I'll track down that to find out how I can check for it in cgminer so to be able to check exactly what the version number is.
I don't know the control transfer codes to read it.
Burnin, if you happen to know, could you post/PM me the info.

[burnin]

The current firmware doesn't report the fw-version.

[kano]

The current firmware doesn't report the fw-version.

OK, yep got the email from Mr Firmware
I'll try adding it so that it reports 0.0.0 if it is missing or whatever it is if it is there.
(I can also try with some of the older firmware I have to make sure it works)

[SebastianJu]

Is it normal that the last miner in a row of 5 is having the red led shining nearly all the time? Only blinking going out instead like the others that go on blinking?

Its this way:

Miner 1: Green normal blinking, Yellow not blinking, Red normal blinking
Miner 2-4: Green normal blinking, Yellow normal blinking, Red normal blinking
Miner 5: Green normal blinking, Yellow normal blinking, Red constantly on and going out blinking occassionally

Is this how it should be?

By the way. Till now i didnt find a way to get above a 42.34GH/s (The amount reaching the pool. No rejected shares and >1% HW) Settings are 29ms 440MHz 1330mV for this. Running at 50°C (Less when its night and its cooler.)

[JHenderson]

Is it normal that the last miner in a row of 5 is having the red led shining nearly all the time? Only blinking going out instead like the others that go on blinking?

Its this way:

Miner 1: Green normal blinking, Yellow not blinking, Red normal blinking
Miner 2-4: Green normal blinking, Yellow normal blinking, Red normal blinking
Miner 5: Green normal blinking, Yellow normal blinking, Red constantly on and going out blinking occassionally

Is this how it should be?

By the way. Till now i didnt find a way to get above a 42.34GH/s (The amount reaching the pool. No rejected shares and >1% HW) Settings are 29ms 440MHz 1330mV for this. Running at 50°C (Less when its night and its cooler.)

What is the power consumption?

[burnin]

I have new firmware for you, that will increase the effective hashrate.
But it requires the newest cg-miner build from source to function.

https://www.dropbox.com/s/0lo1t40yhkbsq25/BB_Firmware_1.0.2.hex

This one now has a  number  

And don't mix different FW-revisions in a cluster.

[eraziel]

I have new firmware for you, that will increase the effective hashrate.
But it requires the newest cg-miner build from source to function.

https://www.dropbox.com/s/0lo1t40yhkbsq25/BB_Firmware_1.0.2.hex

This one now has a  number  

And don't mix different FW-revisions in a cluster.

Source from ckolivas or Kano's repository?

[kano]

I have new firmware for you, that will increase the effective hashrate.
But it requires the newest cg-miner build from source to function.

https://www.dropbox.com/s/0lo1t40yhkbsq25/BB_Firmware_1.0.2.hex

This one now has a  number  

And don't mix different FW-revisions in a cluster.

Source from ckolivas or Kano's repository?

I've committed it into ckolivas master git now
Though it does only add a "version" to the API stats
We don't use it for anything quite yet
However, ckolivas has been making some sweeping changes tonight so I'll post again here when we've finished with what he's doing - and thus when it's back to ideal to get current git

Edit: it shows 1.0.0 if your firmware doesn't have it

[BenTuras]

Tried the horizontal CAN-BUS ports, no change...
...

Try a canbus cable that only has as many connectors as you have boards.

[dddbtc]

Burnin, any word on when order #824 will ship.  I placed an order for a BitBurnerXX with zefir batch #1 chips and attached heatsink on August 6th.  I lost access to my lavabit account because of Edward Snowden and the NSA unfortunately.  It still shows my order as "Processing" in my burninmining account.  I would really like to receive a tracking number or at least hear that it has or hasn't shipped.

Thanks!

[eraziel]

When your order is shipped your order status will change to Complete and you will receive a mail with a tracking number, if applicable

[LordTheron]

Tried the horizontal CAN-BUS ports, no change...
...

Try a canbus cable that only has as many connectors as you have boards.

Thanks, I got it working but only with latest firmware and as you mentioned with cable with  as many connectors as boards. if I have latest firmware and more connectors than boards, can-bus is not working. Bit strange but its working so don't care.  


Burnin, could you respond to my email sent to you  regarding 3 faulty bitburners? Bus-can is not working on them and I cant upgrade the firmware as all 3 usb ports are faulty. Is there any other way to update the firmware than usb?

Lastly, when I have 7 bitburners connected via can-bus, I get only single BTB0 device. Its hashing with the speed of 7 but I was wondering how do I see temp on each of them. is it normal that they show as single device?

Many thanks in advance.

Edit: Also I have 1 miner that is doing half the work what the other ones do with the same settings. Any idea why that may be?
My settings are cgminer.exe --avalon-options 115200:2:10:35:350 --avalon-fan 100 --avalon-temp 50 -u xxx -p 1234 -o stratum.bitcoin.cz:3333

 (5s):4.032G (avg):3.517Gh/s | A:324  R:1  HW:0  WU:49.1/m
 ST: 2  SS: 0  NB: 2  LW: 593  GF: 0  RF: 0
 Connected to stratum.bitcoin.cz diff 8 with stratum as user xxx
 Block: 004791794de937d6...  Diff:50.8M  Started: [22:46:27]  Best share: 1.62K
-------------------------------------------------------------------------------
 [P]ool management ettings [D]isplay options [Q]uit
 BTB 0: 41/ 41C 1229mV | 3.427G/3.517Gh/s | A:332 R:1 HW:0 WU: 49.7/m
-------------------------------------------------------------------------------

 [2013-08-19 22:45:59] Accepted 07001753 Diff 36/8 BTB 0
 [2013-08-19 22:46:09] Accepted 03ac9151 Diff 69/8 BTB 0
 [2013-08-19 22:46:09] Accepted 0766bc93 Diff 34/8 BTB 0
 [2013-08-19 22:46:16] Accepted 0618e2e4 Diff 41/8 BTB 0
 [2013-08-19 22:46:22] Accepted 0ee012ba Diff 17/8 BTB 0
 [2013-08-19 22:46:26] Accepted 047a2465 Diff 57/8 BTB 0
 [2013-08-19 22:46:27] Stratum from pool 0 detected new block
 [2013-08-19 22:46:33] Accepted 10738936 Diff 15/8 BTB 0
 [2013-08-19 22:46:33] Accepted 0f8fa0e5 Diff 16/8 BTB 0
 [2013-08-19 22:46:43] Accepted 002850ed Diff 1.62K/8 BTB 0
 [2013-08-19 22:46:48] Accepted 15dae9f6 Diff 11/8 BTB 0
 [2013-08-19 22:47:06] Accepted 01cf081d Diff 141/8 BTB 0
 [2013-08-19 22:47:11] Accepted 11c46aaf Diff 14/8 BTB 0

[SebastianJu]

By the way. Till now i didnt find a way to get above a 42.34GH/s (The amount reaching the pool. No rejected shares and >1% HW) Settings are 29ms 440MHz 1330mV for this. Running at 50°C (Less when its night and its cooler.)

What is the power consumption?

475W for 5 Miners without rpi. Only miners and PSU.

[-Redacted-]

Interesting post, I think I see how to make strikeout work now!

[kano]

...
Lastly, when I have 7 bitburners connected via can-bus, I get only single BTB0 device. Its hashing with the speed of 7 but I was wondering how do I see temp on each of them. is it normal that they show as single device?

No, there are no commands in the firmware yet to get details off the other boards.
I'm not sure yet what might be added to the firmware in the near future.

...
Edit: Also I have 1 miner that is doing half the work what the other ones do with the same settings. Any idea why that may be?
My settings are cgminer.exe --avalon-options 115200:2:10:35:350 --avalon-fan 100 --avalon-temp 50 -u xxx -p 1234 -o stratum.bitcoin.cz:3333
...

I know one possible cause of this, but I was given a special firmware to resolve it.
If it is the same problem, and burnin is OK with it, I'll forward the firmware I got.
I'll wait see what his reply is here.

[LordTheron]

No, there are no commands in the firmware yet to get details off the other boards.
I'm not sure yet what might be added to the firmware in the near future.

So is it safe to run them in cluster?
 

I know one possible cause of this, but I was given a special firmware to resolve it.
If it is the same problem, and burnin is OK with it, I'll forward the firmware I got.
I'll wait see what his reply is here.

Cheers, that would be great. Its weird one as  overclocking doesn't seem to do much in terms of speed. At 400 and timeout 30, it only peaks at 5.2gh and average is   3.54gh. On 360 clock and 35 timeout it peaks at 4.7gh and  average is 3.47gh. Do you know how to check if all the chips are hashing? 

[SebastianJu]

I have new firmware for you, that will increase the effective hashrate.
But it requires the newest cg-miner build from source to function.

https://www.dropbox.com/s/0lo1t40yhkbsq25/BB_Firmware_1.0.2.hex

This one now has a  number  

And don't mix different FW-revisions in a cluster.

I flashed the firmware and recompiled cgminer. Its running fine and it looks like the hashrate is a bit higher now. Might be variance so i have to observe first.
Can you explain what you changed so that i maybe could use the info to tweak the settings better?

In your news to flashing you say only to disconnect after flashing, not Run Application anymore?

[g83]

Still running on windows or did you change to linux ?

[SebastianJu]

Still running on windows or did you change to linux ?

In case you mean me, i flashed in windows again but mine at rpi for some time now.

@Burnin... great Firmware Update. Before the biggest hashrate i got was 42.3GH but your firmware gave me 6.03% more. So that my pool tells me i run at an average of 44.85GH/s. Cgminer claims the average is at 44.01GH after 10 hours but i used the hashrate of my pool before to determine that (the pools one was a bit lower mostly) so i will take it now too. At least the pools hashrate is what matters. My pool collects data for 1hour so i  measure after 65 minutes normally, when testing, and take the average of 10 values the pool gives me.

So im happy with the new hashrate. My settings are 450MHz 1340mV and 28ms. Rejected at 0.01% and HW at 0.7%. 491W.

But do you have an explaination why its not possible to go above this? Is it a hardware part that is limiting? When i go above 450MHz the hashrate drops, and when i use more voltage the hashrate drops too. So when using 470MHz and 1400mV i only get 17GH anymore. Wattage is jumping around between 330 and 410W. Normal would be to stay relatively at 491W with the new settings. In fact im impressed how exact it remains at 491W. Its not even changing anymore like it did with previous firmware. So i guess thats one point you changed in firmware.

Are the problems with higher MHz and voltage fixable or by design?