Pages:
Author

Topic: The Chili – 30+GH/s BFL based Bitcoin Miner Assembly - page 19. (Read 137904 times)

sr. member
Activity: 280
Merit: 250
Helperizer
That would imply that either it's not accessing the EEPROM correctly, or that it wasn't successful when you programmed the FTDI chip.

You did program the FTDI chip, correct?
Aha, I'm guessing that I missed that part of swapping it out.  I don't have a programmer for such things, let alone the program I'd need (assuming you're not talking about the firmware update...).

The question then becomes, what now?  Wink
legendary
Activity: 1274
Merit: 1004
Also, following up on the Chili that died during a reset gone bad (short pins 5-6 but must have also touched something adajacent)...  I replaced the FTDI, so now two parts are working:

1.) The boot up light sequence appears to be functioning correctly, so I would imagine that most of the Chili should be in good order (was working right after I lost comms/hashing, too).
2.) The FTDI seems to work.  I can see the device in the USBs and in Windows I can *try* to read or even flash the firmware, but I get garbage in the results window.

Something else must be dead, but I'm not sure where to go from here.  Ideas, anyone?
I would concur with lightfoot, it sounds like the microcontroller is running fine and the FTDI chip is also working or at least partially working.
When the device shows up in Windows, is it a serial port / serial convertor, or does it show up as a Bitforce SC? If it doesn't show up in Windows and you're getting garbage from the EEPROM in the FT_PROG window, I'd look for a cold solder joint on the FTDI chip.
Good deal - thanks, I'll take a look when I get home from work.  My eyes aren't that good for close up stuff (old!), but I'll see what I can do.  I was wondering if it could be something like that, but couldn't see anything amiss earlier.  This gives me impetus to look again or borrow my wife's eyes.  ;-)

BTW the FTDI shows up in windows as a serial converter IIRC.  Does that help?
That would imply that either it's not accessing the EEPROM correctly, or that it wasn't successful when you programmed the FTDI chip.

You did program the FTDI chip, correct?
member
Activity: 71
Merit: 10
For those who have assembled their chilis with thermal pads instead of paste -

Do you cut small squares for each chip or have you been using one large pad that also covers the open area between the chips?

Thanks!

I used one large pad which also covers the open area between the chips.  I have had no problems with that technique.
sr. member
Activity: 280
Merit: 250
Helperizer
Also, following up on the Chili that died during a reset gone bad (short pins 5-6 but must have also touched something adajacent)...  I replaced the FTDI, so now two parts are working:

1.) The boot up light sequence appears to be functioning correctly, so I would imagine that most of the Chili should be in good order (was working right after I lost comms/hashing, too).
2.) The FTDI seems to work.  I can see the device in the USBs and in Windows I can *try* to read or even flash the firmware, but I get garbage in the results window.

Something else must be dead, but I'm not sure where to go from here.  Ideas, anyone?
I would concur with lightfoot, it sounds like the microcontroller is running fine and the FTDI chip is also working or at least partially working.
When the device shows up in Windows, is it a serial port / serial convertor, or does it show up as a Bitforce SC? If it doesn't show up in Windows and you're getting garbage from the EEPROM in the FT_PROG window, I'd look for a cold solder joint on the FTDI chip.
Good deal - thanks, I'll take a look when I get home from work.  My eyes aren't that good for close up stuff (old!), but I'll see what I can do.  I was wondering if it could be something like that, but couldn't see anything amiss earlier.  This gives me impetus to look again or borrow my wife's eyes.  ;-)

BTW the FTDI shows up in windows as a serial converter IIRC.  Does that help?
hero member
Activity: 868
Merit: 1000
I have just assembled my 7th Chili

On power up, the leds 5,6,7,8, come on and stay on, then leds 1,2,3,4, flash once.
The com port is not found. nothing else happens.

Does anyone have any ideas?
This is one of Lucko's boards?

LEDS 5-8 being on indicate that it is in the hardware initialization phase and if there is a hardware fault it keep those on. This is usually due to the 1V power supply not being able to turn on for whatever reason.

Thanks for the info.
full member
Activity: 195
Merit: 100
For those who have assembled their chilis with thermal pads instead of paste -

Do you cut small squares for each chip or have you been using one large pad that also covers the open area between the chips?

Thanks!
legendary
Activity: 1274
Merit: 1004
I have just assembled my 7th Chili

On power up, the leds 5,6,7,8, come on and stay on, then leds 1,2,3,4, flash once.
The com port is not found. nothing else happens.

Does anyone have any ideas?
This is one of Lucko's boards?

LEDS 5-8 being on indicate that it is in the hardware initialization phase and if there is a hardware fault it keep those on. This is usually due to the 1V power supply not being able to turn on for whatever reason.
hero member
Activity: 868
Merit: 1000
I have just assembled my 7th Chili

On power up, the leds 5,6,7,8, come on and stay on, then leds 1,2,3,4, flash once.
The com port is not found. nothing else happens.

Does anyone have any ideas?
legendary
Activity: 1274
Merit: 1004
Also, following up on the Chili that died during a reset gone bad (short pins 5-6 but must have also touched something adajacent)...  I replaced the FTDI, so now two parts are working:

1.) The boot up light sequence appears to be functioning correctly, so I would imagine that most of the Chili should be in good order (was working right after I lost comms/hashing, too).
2.) The FTDI seems to work.  I can see the device in the USBs and in Windows I can *try* to read or even flash the firmware, but I get garbage in the results window.

Something else must be dead, but I'm not sure where to go from here.  Ideas, anyone?
I would concur with lightfoot, it sounds like the microcontroller is running fine and the FTDI chip is also working or at least partially working.
When the device shows up in Windows, is it a serial port / serial convertor, or does it show up as a Bitforce SC? If it doesn't show up in Windows and you're getting garbage from the EEPROM in the FT_PROG window, I'd look for a cold solder joint on the FTDI chip.
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Are you sure all the pins are soldered and there are no bridges? It took me the ass end of ever to get the FTDI chip working on a jally; class one pain in the ass.

C
sr. member
Activity: 280
Merit: 250
Helperizer
Also, following up on the Chili that died during a reset gone bad (short pins 5-6 but must have also touched something adajacent)...  I replaced the FTDI, so now two parts are working:

1.) The boot up light sequence appears to be functioning correctly, so I would imagine that most of the Chili should be in good order (was working right after I lost comms/hashing, too).
2.) The FTDI seems to work.  I can see the device in the USBs and in Windows I can *try* to read or even flash the firmware, but I get garbage in the results window.

Something else must be dead, but I'm not sure where to go from here.  Ideas, anyone?
sr. member
Activity: 267
Merit: 250
So, with my rooms being nice and cold now, I finally ran into the situation where my Chili would overvolt.  I installed the e1v1.hex to limit the voltage, and it works, but I definitely take a hit - about 5% or 2 GH/s.  I see 4 places where the e.hex differs from the e1v1.hex file, and I was wondering if I could just get a bit of the voltage back since my units run so well.  But it's not clear what to change.  Thoughts?  I'm thinking 1.14V would be a better limit than 1.1V since mine runs well up to 1.18V.

Can a firmware file be posted that has a different limit?  Or a guide so that we can put in the limit of our own choosing?

Thanks much!

Most of my best performers are working well between 1.14v and 1.16v
sr. member
Activity: 280
Merit: 250
Helperizer
So, with my rooms being nice and cold now, I finally ran into the situation where my Chili would overvolt.  I installed the e1v1.hex to limit the voltage, and it works, but I definitely take a hit - about 5% or 2 GH/s.  I see 4 places where the e.hex differs from the e1v1.hex file, and I was wondering if I could just get a bit of the voltage back since my units run so well.  But it's not clear what to change.  Thoughts?  I'm thinking 1.14V would be a better limit than 1.1V since mine runs well up to 1.18V.

Can a firmware file be posted that has a different limit?  Or a guide so that we can put in the limit of our own choosing?

Thanks much!
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Thanks for that picture. Your errors and submit ratios almost exactly match my unit, so I can say mine is working normally. I still am getting a good 39/36gh rate out of it with the water cooling heatsink, best $50 I have spent.

C

Mine are just using the EVO212 heatsinks with the Recommended Padding (not even arctic silver)
*nod* And running a few gh slower, matches what I saw with the cooler master. Oddly enough I am running two 7 chip jallies, one with the cooler master and a big heat sink on the back (60-70c), and one with a corsair water block on the back and the cheap BFL AL sink on the front (40c). The temp difference is quite impressive, but so is the value of cooling the back of the board as well as the chips.

Fun stuff.
hero member
Activity: 518
Merit: 500
Every man is guilty of all the good he did not do.
Thanks for that picture. Your errors and submit ratios almost exactly match my unit, so I can say mine is working normally. I still am getting a good 39/36gh rate out of it with the water cooling heatsink, best $50 I have spent.

C

Mine are just using the EVO212 heatsinks with the Recommended Padding (not even arctic silver)
hero member
Activity: 1246
Merit: 501
i have win xp

i must install something else first?HuhHuh?

How about installing an OS that's less than 13 years old?   Roll Eyes
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)


Almost 36 Days uptime. Solid Units.

Looks like you would have almost solved a solo share (1.14G)!  Or depending on the difficulty at the time maybe you would have.
Thanks for that picture. Your errors and submit ratios almost exactly match my unit, so I can say mine is working normally. I still am getting a good 39/36gh rate out of it with the water cooling heatsink, best $50 I have spent.

C
sr. member
Activity: 267
Merit: 250


Almost 36 Days uptime. Solid Units.

Looks like you would have almost solved a block there (best share = 1.14G)!  Or depending on the difficulty at the time maybe you would have.
hero member
Activity: 518
Merit: 500
Every man is guilty of all the good he did not do.


Almost 36 Days uptime. Solid Units. As always I recommend MrTeal/ChipGeek.
legendary
Activity: 2408
Merit: 1004
i have win xp

try to install   win usb drivers with zadig

but failed to install;

why that happen??
i must install something else first?HuhHuh?
Pages:
Jump to: