Hacking the S7 - improving efficiency through minor hardware manipulation

[sidehack]

You sure it's not reading from the SO-8 TMP75 in the center of the board using I2C via the TSCL and TSDA lines on the cable connector?

[fanatic26]

You sure it's not reading from the SO-8 TMP75 in the center of the board using I2C via the TSCL and TSDA lines on the cable connector?

Ill have to dig out my documentation to confirm but that is what I remember from the hardware training I received from bitmain last year in China.

[sidehack]

You're not thinking the S9, are you?

[surferdudemi]

You sure it's not reading from the SO-8 TMP75 in the center of the board using I2C via the TSCL and TSDA lines on the cable connector?

Ill have to dig out my documentation to confirm but that is what I remember from the hardware training I received from bitmain last year in China.

All of the chips show "o", so it seems odd that the temp value would be bad / wrong on two boards when the chips appear to be functioning properly.  There are multiple posts about the temp problem (I Googled it, not on here), but I could not find any info about where the sensor is so I could inspect it on the board.

People have mentioned that Bitmain had the schematics available, and that they were pulled.  Did copies survive anywhere?  Didn't see any with a Google search - even in the images which sometimes show stuff you don't see in the regular search results.

[fanatic26]

You're not thinking the S9, are you?

Apologies, it seems that I was thinking s9. I pulled out my s7 maintenance guide and confirmed I was mixing the two up.

[Sweminer777]

any form to, revive this boards, when they show 00000000000000000000000000000 but do not hash? what can be done or work around?

[fanatic26]

any form to, revive this boards, when they show 00000000000000000000000000000 but do not hash? what can be done or work around?

There are many reasons for it not hashing, we would need more info to help you. I would recommend making a post in the support section with your full kernel log so people can help.

[sidehack]

To the guy whose post was recently deleted, asking about D3 hacking, someone on the previous page mentioned looking into it.

I started a topic on the D3 with a hardware mod. This is for steady hands only for now  
https://bitcointalk.org/index.php?topic=2346161.0

(Moderator's note: This post was edited by frodocooper to remove an off-topic portion of the post.)

[FNT]

Hi sidehack,

Thanks for replay... and sorry for off topic posting.
I checked the post of D3 completly, but I found no hex file of D3.

About the S7, how have you made the settings in the hex file for the S7?
Have you just modified the memory where the voltage settings are stored or did you
have done the work from scratch do code the voltage settings?

Thanks for info!

Bye, FNT

[NotFuzzyWarm]

Hmm, might this explain my donated old Ant45s7b6 briefly showing up on Kano last night? Someone fiddling with it and a PIC programmer?

[sidehack]

Why would someone have bought it six months ago or whatever and not run it or change pool configs? That's weird.

[NotFuzzyWarm]

No idea unless they only changed pool-1 and it failed over to one of the other original settings for pool-2 or 3. It last reported to Kano's pool 13 hrs ago.

[FNT]

Hi,

@notfuzzywarm: not that miner, as hashboard is here on the table,
all pools set to mine. It's an other ant45s7.

Any hints about the PIC?

Bye

[Breeze]

Anybody have an idea why i cant connect?
https://www.dropbox.com/s/xocgink55xkclrq/fail.JPG?dl=0
https://www.dropbox.com/s/xocgink55xkclrq/fail.JPG?dl=0


Moderator's note: This post was edited by frodocooper to remove inline image tags.

[sidehack]

Plenty of ideas, but it's not too helpful without knowing what you've done to get there.

Is your kit plugged in backward? Did you disconnect the 18-pin cable to the control board?

[Breeze]

Plenty of ideas, but it's not too helpful without knowing what you've done to get there.

Is your kit plugged in backward? Did you disconnect the 18-pin cable to the control board?

I set the voltage 3.3. Arrow on the pickit is down, like in the picture. Have a pin cable. Yes, have dissconnected the 18 pin cable from the boars.
Now i'm gonna try older version of the IPE 3.35

[Breeze]

Code:

2018-04-17 14:38:44 +0200 - Completed loading IPE.


*****************************************************

Connecting to MPLAB PICkit 3...

Currently loaded firmware on PICkit 3
Firmware Suite Version.....01.51.08 *
Firmware type..............Enhanced Midrange

Now Downloading new Firmware for target device: PIC12F1572 
Downloading bootloader
Bootloader download complete
Programming download...
Downloading RS...
RS download complete
Programming download...
Downloading AP...
AP download complete
Programming download...

Currently loaded firmware on PICkit 3
Firmware Suite Version.....01.43.35
Firmware type..............Enhanced Midrange

Programmer to target power is enabled - VDD = 3,300000 volts.
Unable to connect to the target device.
Failed to get Device ID

[Breeze]

Now it connect when i use "Low Voltage Power" in the Power menu.
But have this error:

Code:

*****************************************************

Connecting to MPLAB PICkit 3...

Currently loaded firmware on PICkit 3
Firmware Suite Version.....01.43.35
Firmware type..............Enhanced Midrange

Programmer to target power is enabled - VDD = 3,300000 volts.
Target Device ID (0x0) is an Invalid Device ID. Please check your connections to the Target Device.

EDIT: Got it working now. It works when i have the 18-pin cable connected to the board. It's some random shit.....
Testing board #1 now  

[House Xylander]

Having issues with programming my S7 (45 Asic) boards.

I am using the MPLAB IPE v3.35

The Low voltage checkbox has to be checked as well or it will not connect for me.

The target id keeps coming up as (0x0) and says its invalid. If I select ok it connects but still has the target ID error.

I can read and export original firmware

When I try to Program the board with any of the Hex files, I get the following:

"2018-06-21 22:57:42 -0500 - Programming...

The following memory area(s) will be programmed:
program memory: start address = 0x0, end address = 0xaf
configuration memory

Device Erased...

Programming...
program memory
Address: 0 Expected Value: 2805 Received Value: 0
Failed to program device
2018-06-21 22:57:45 -0500 - Programming failed
Pass Count: 0"

I would appreciate any help on this matter.

[House Xylander]

Having issues with programming my S7 (45 Asic) boards.

I am using the MPLAB IPE v3.35

The Low voltage checkbox has to be checked as well or it will not connect for me.

The target id keeps coming up as (0x0) and says its invalid. If I select ok it connects but still has the target ID error.

I can read and export original firmware

When I try to Program the board with any of the Hex files, I get the following:

"2018-06-21 22:57:42 -0500 - Programming...

The following memory area(s) will be programmed:
program memory: start address = 0x0, end address = 0xaf
configuration memory

Device Erased...

Programming...
program memory
Address: 0 Expected Value: 2805 Received Value: 0
Failed to program device
2018-06-21 22:57:45 -0500 - Programming failed
Pass Count: 0"

I would appreciate any help on this matter.

Resolved this by unchecking the LVP box and increasing the VDD power to 4.65
At 4.65 it connected successfully and produced the correct Target ID

I now have them running 640,630,630 respectfully at 600Mhz and kill a watt is showing a 634w average at 110v.
This cut the cost to run by about half but I fear the continued drop in the value of BTC is the next obstacle in the near future.

Any suggestions on a PSU, preferably a quieter one than the 1300w server PSU?