X6500 Custom FPGA Miner

[Dexter770221]

Which revision will have I/O routed out? It's not that much work to do, and this should increase resale value and usefulness. Lack of I/O is one of not so many cons of this great mining board.

[1714124968]

1714124968

[1714124968]

1714124968

[1714124968]

1714124968

[hashking]

Hey guys. I finally put together an installation package of sorts for Windows:

X6500 Easy Package for Windows
Download


This should make it simple and easy take an X6500 and go from 0 to Bitcoin in no time (on Windows). It includes a driver installation program for the libusb driver, MPBM with compiled EXE, and all the current mining firmware. I tested it out on a clean VM install of Windows 7 64-bit, Windows 7 32-bit, and Windows XP 32-bit.

If that's too convenient for you, the individual pieces can also be downloaded:

• FPGA Mining Driver Install
• Modular Python Bitcoin Miner
• Firmware

Let me know if there are any problems. Though I did take a lot of time testing it, this is still fresh off the presses

Very nice!  Will test it out asap and post up any issues.

When I run the web interface http://127.0.0.1:8832.  Some of the information is written over other informationi and I can't use it.  I'm using mozilla firefox.  Any help would be great.

[TheSeven]

When I run the web interface http://127.0.0.1:8832.  Some of the information is written over other informationi and I can't use it.  I'm using mozilla firefox.  Any help would be great.

Use Chrome. It's known broken in IE/FF/Opera, and I don't really know why this happens.
I'll happily accept a patch if someone figures out how to fix that.

[hashking]

When I run the web interface http://127.0.0.1:8832.  Some of the information is written over other informationi and I can't use it.  I'm using mozilla firefox.  Any help would be great.

Use Chrome. It's known broken in IE/FF/Opera, and I don't really know why this happens.
I'll happily accept a patch if someone figures out how to fix that.

Thanks.  It worked, but now to of my miners arent loading.  All of my other ones are. I keep getting traceback errors.

[hashking]

Got it working.  Had to unplug the usb cable and plug it back in   Thanks.

[Garr255]

That just cycles power, unloading the firmware from the fpga. Ticking the box mentioned above will give you the same result most of the time.

[Pipesnake]

Have a board stuck at 50mh/s per core.  Any ideas as to why?

[fizzisist]

Have a board stuck at 50mh/s per core.  Any ideas as to why?

Using the new overclocker firmware? That is the starting frequency if you don't send the FPGA any commands to set the clock. I'm guessing an MPBM bug that caused that command to not go out. Can you restart the worker in MPBM? You might not even need to upload the firmware again, although that couldn't hurt.

If you're on a fixed-clock firmware, like ztexmerge, then I don't know what's going on. First thing to try is reprogramming it. If you've done that already, then I'm pretty stumped.

[nbtcminer]

BTW I've been testing and playing around with my X6500 Rev 3.0 and I got to say: this thing is solid! I'm using the Zalman "flower" coolers and so far getting a fairly consistent hash power from my unit! Pics to come later once I've gotten it all settled!


-n

[Pipesnake]

Have tried rebooting the board several times....



 

[nbtcminer]

Have tried rebooting the board several times....



 

@ Pipesnake:
A couple of questions for you:

1.) Which rev of the x6500 are you using?
2.) Which bitstream / firmware are you using?
3.) Have you tried 0.4 of MPBM or only 1.0 MPBM?

[TheSeven]

Which bitstream are you using?
Can you post a screenshot of the worker configuration?
Can you paste the full MPBM log, from startup until the first couple of shares were found?

[Pipesnake]

Wow I'm a 'tard.  Forgot to update bitstream from default.  Thanks for the quick response guys.

[freshzive]

So does program.py not work to program the new rev of boards? I get this error on both

Code:

2012-04-16 21:45:07 | Device 3 opened (A4014PHJ)
Traceback (most recent call last):
  File "program.py", line 91, in <module>
    fpga.detect()
  File "/Users/christian/x6500-6/fpga.py", line 82, in detect
    with self.ft232r.lock: self.jtag.detect()
  File "/Users/christian/x6500-6/jtag.py", line 81, in detect
    raise NoDevicesDetected

Shows up in listDevices.py fine though?

[fizzisist]

So does program.py not work to program the new rev of boards? I get this error on both

Code:

2012-04-16 21:45:07 | Device 3 opened (A4014PHJ)
Traceback (most recent call last):
  File "program.py", line 91, in <module>
    fpga.detect()
  File "/Users/christian/x6500-6/fpga.py", line 82, in detect
    with self.ft232r.lock: self.jtag.detect()
  File "/Users/christian/x6500-6/jtag.py", line 81, in detect
    raise NoDevicesDetected

Shows up in listDevices.py fine though?

It should... Have you tried MPBM? NoDevicesDetected means no FPGAs were detected, which usually happens when you don't have power on the board. Can you double check your power supply?

[freshzive]

So does program.py not work to program the new rev of boards? I get this error on both

Code:

2012-04-16 21:45:07 | Device 3 opened (A4014PHJ)
Traceback (most recent call last):
  File "program.py", line 91, in <module>
    fpga.detect()
  File "/Users/christian/x6500-6/fpga.py", line 82, in detect
    with self.ft232r.lock: self.jtag.detect()
  File "/Users/christian/x6500-6/jtag.py", line 81, in detect
    raise NoDevicesDetected

Shows up in listDevices.py fine though?

It should... Have you tried MPBM? NoDevicesDetected means no FPGAs were detected, which usually happens when you don't have power on the board. Can you double check your power supply?

Yeah, you were right. I was able to upload the firmware, but now I'm having issues connecting in MPBM 0.4.

I get the error:

2012-04-17 08:35:58.714317: X6500 board A4014PHJ: Can not open the specified device

Ztex200 is definitely loaded on the board, so I'm not sure what's up? All of my old rev work fine with this version.

[freshzive]

Yeah, and the old mine.py works fine with it as well...strange!

[TheSeven]

Yeah, and the old mine.py works fine with it as well...strange!

Maybe ftdi_sio got loaded accidentally? x6500-miner tries to remove that automatically, MPBM doesn't.

[bitcowok]

Just a photo of the latest progress on my rig (4 x x6500's so far). TIP: if you live in Australia, order these individually otherwise customs will slug you if it's worth over A$1000 (even if you paid in bitcoins).  These are the new rev3 boards with old style heatsinks, attached with arctic alumina thermal adhesive. The new boards are very slightly too big for my original design, but with a bit of shoving and wiggling, they fit fine.

http://jabawok.net/gallery/d/29298-2/20120418_003.jpg

Heres the rest of the build photos, if interested: http://jabawok.net/gallery/v/random/201201-x6500-30x-enclosure/

The boards are all running at 200Mhz (1.6Gh/s total), the fans are @ 7V and total power consumption at the AC is 95W

Temperature at 25C ambient is 36C on lower heatsink and 42C on upper heatsink. I'm not getting any invalids, if I start getting them when the ambient temperature is higher, i'll increase fan voltage a bit (which costs a few watts).

@fizzisist: if you have any blue heatsinks left, save them for me! What was your supplier for those blue heatsinks? On the rev3 boards sold with heatsnk/fan, how high do the standoffs need to be to stack the boards and allow adequate airflow? I can work with it, but I think i wont be able to fit 30 boards in my enclosure any more with the new heatsink/fan: blowing air down rather than across is not really optimal for stacking.

[fizzisist]

Just a photo of the latest progress on my rig (4 x x6500's so far). TIP: if you live in Australia, order these individually otherwise customs will slug you if it's worth over A$1000 (even if you paid in bitcoins).  These are the new rev3 boards with old style heatsinks, attached with arctic alumina thermal adhesive. The new boards are very slightly too big for my original design, but with a bit of shoving and wiggling, they fit fine.



Heres the rest of the build photos, if interested: http://jabawok.net/gallery/v/random/201201-x6500-30x-enclosure/

The boards are all running at 200Mhz (1.6Gh/s total), the fans are @ 7V and total power consumption at the AC is 95W

Temperature at 25C ambient is 36C on lower heatsink and 42C on upper heatsink. I'm not getting any invalids, if I start getting them when the ambient temperature is higher, i'll increase fan voltage a bit (which costs a few watts).

@fizzisist: if you have any blue heatsinks left, save them for me! What was your supplier for those blue heatsinks? On the rev3 boards sold with heatsnk/fan, how high do the standoffs need to be to stack the boards and allow adequate airflow? I can work with it, but I think i wont be able to fit 30 boards in my enclosure any more with the new heatsink/fan: blowing air down rather than across is not really optimal for stacking.

Awesomeness! Glad you were able to get the rev3 to fit.

I have a whole box of those heatsinks with your name on it. We bought them from Digikey, but you're welcome to them at a nice discount.

The new heatsinks are about 2 inches tall with the fans. You need another 1/2 inch or so at least if you want to stack them and leave the fans on. But, you can remove the fans (4 screws) and supply airflow from the side, just like you're doing now. The height of the heatsink then is a little under 1.5 inch. I've been using 2 inch standoffs between boards, but you could get by with less, maybe 1 3/4 inch.