Pages:
Author

Topic: [GUIDE] BitFury Miner Support/Tuning - page 18. (Read 148062 times)

hero member
Activity: 681
Merit: 500
November 10, 2013, 04:35:56 PM
For all those who measure the voltage applied to chip on H-Boards, as a minus (GND) use for this GND on the H-Board instead power suply or M-boards. These differences arising from the resistance of sockets and contacts on a card. They range tenths of volts. It turns out that, instead of 0.865V is 0.825 or even 0.760
Therefore, the chip receives voltage measured preferably on next chip capacitors or capacitors next to coils (PULSE) of DC-DC converter labeled C12 to C16.

https://www.dropbox.com/s/rxhufkcy1tgtwx2/20131110_220257.jpg

Check voltage on top of the inductor (grey box with PULSE printed on it and any of the GND pin from the PCIE power supply)

Just FYI... There is a small problem with measuring voltage like that, but it's still the easiest way in a crowded rig. There's a significant and varying voltage drop on the ground line between the power supply and a card. Just between the M-board's GND terminal and a slot's ground pins, there's a drop of between 0.023v (slot 0) and 0.067v (slot F) on my v1 rig. So while it may look like you're running a card at 0.900v when measured the easy way, it could actually be as low as 0.833v. And I didn't even account for any further drop from the slot to the chips' ground plane.
sr. member
Activity: 259
Merit: 250
Dig your freedom
November 10, 2013, 04:10:27 PM
For all those who measure the voltage applied to chip on H-Boards, as a minus (GND) use for this GND on the H-Board instead power suply or M-boards. These differences arising from the resistance of sockets and contacts on a card. They range tenths of volts. It turns out that, instead of 0.865V is 0.825 or even 0.760
Therefore, the chip receives voltage measured preferably on next chip capacitors or capacitors next to coils (PULSE) of DC-DC converter labeled C12 to C16.

https://www.dropbox.com/s/rxhufkcy1tgtwx2/20131110_220257.jpg
legendary
Activity: 1593
Merit: 1004
November 08, 2013, 05:26:08 PM
I just received two new h-boards, V1.2 as I have a V1 M-board.
Both cards are running at 33-34 Gh/s out of the box.
I measured the R02F resistor and both boards read 2.1KOhm.
My other boards read 1.3KOhm.
Can I still attempt to mod the two new boards?  If so, I don't want to move them much because they are already running high.  But do have very good cooling, 4 fans on 6 cards.
Thoughts?  Input from the wise men?



Yes, you can pencil mod them and they respond as well as the previous boards, do not go as low tough. I've found that any board can take 1.72 and ones with very good cooling (3 fans per bank and heatsinks) can go as low as 1.63KOhm. I'd try to lower the resistance by 0.05KOhm at a step and see how the cards respond.
Makes sense.  Thanks guys.
hero member
Activity: 910
Merit: 1000
Items flashing here available at btctrinkets.com
November 08, 2013, 05:02:35 PM
I just received two new h-boards, V1.2 as I have a V1 M-board.
Both cards are running at 33-34 Gh/s out of the box.
I measured the R02F resistor and both boards read 2.1KOhm.
My other boards read 1.3KOhm.
Can I still attempt to mod the two new boards?  If so, I don't want to move them much because they are already running high.  But do have very good cooling, 4 fans on 6 cards.
Thoughts?  Input from the wise men?



Yes, you can pencil mod them and they respond as well as the previous boards, do not go as low tough. I've found that any board can take 1.72 and ones with very good cooling (3 fans per bank and heatsinks) can go as low as 1.63KOhm. I'd try to lower the resistance by 0.05KOhm at a step and see how the cards respond.
hero member
Activity: 616
Merit: 500
November 08, 2013, 03:36:18 PM
Lol, i am electronics technician, so i repair/change Ti-chip,I do not want to break the warranty...
you will lose the warranty. But so long as you know what you're doing you wont need it. Smiley *high five* I'm an electronics technician too.

Can I still attempt to mod the two new boards?  If so, I don't want to move them much because they are already running high.  But do have very good cooling, 4 fans on 6 cards.
Thoughts?  Input from the wise men?
Iirc some where reporting problems when mixing old pencil modded cards with modded new ones. I'd say do it Smiley you can always clean the grphite off if problems arise.
legendary
Activity: 1593
Merit: 1004
November 08, 2013, 02:17:20 PM
I just received two new h-boards, V1.2 as I have a V1 M-board.
Both cards are running at 33-34 Gh/s out of the box.
I measured the R02F resistor and both boards read 2.1KOhm.
My other boards read 1.3KOhm.
Can I still attempt to mod the two new boards?  If so, I don't want to move them much because they are already running high.  But do have very good cooling, 4 fans on 6 cards.
Thoughts?  Input from the wise men?

newbie
Activity: 29
Merit: 0
November 06, 2013, 10:52:42 AM
I tried, no help... pheww... 0.0V


you went from 0.3v to 0.0v?  That seems wrong. try a different m-board slot and ensure proper insertion.

 If all else fails, ill buy your board for $200 shipped to toronto

Lol, i am electronics technician, so i repair/change Ti-chip,I do not want to break the warranty...
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
November 05, 2013, 07:27:40 PM
I tried, no help... pheww... 0.0V


you went from 0.3v to 0.0v?  That seems wrong. try a different m-board slot and ensure proper insertion.

 If all else fails, ill buy your board for $200 shipped to toronto
newbie
Activity: 29
Merit: 0
November 05, 2013, 03:13:27 PM
Okay, i made changes to settings, problem not solved...

And this is what i get...:

speed:2592 noncerate[GH/s]:96.651 (2.014/chip) hashrate[GH/s]:100.665 good:6751 errors:1585 spi-errors:1 mis$
4:      864     30.695  34.045  2144    339     0       1
8:      864     31.897  32.481  2228    849     0       0
C:      864     34.059  34.140  2379    397     1       0

Voltage still 0.3V

I measured resistors,
R01F:  2,120K ohm
R02F:  2,120K ohm

..so Ti chipis broken...?
 

1) switch to autotuning. clearly setting manual tuning to 54 across every chips is working terribly based on your 15-30% error rates.
2) R02F is quite high at 2.12K(ohm). try using the pencil mod to reduce it to 1.85K(ohm) and see what happens. If lucky, this is the issue and youll see the voltage rise to >0.6V. If it does, maybe the wrong resistor was placed and you should see if it can be changed for something in the 1.65-1.8K range

1. As you can see, m-board don't even detect BANK1 h-card(s). It is autotune mode now, but no help..
2. If I do pencil mod that, that will broke h-card warrantly...?

pencil modding is techincally messing with the warranty, but a tiny marking of pencil lead is not very hard to clean off and, unless you seriously manage to do it wrong, shouldnt cause anything worse then putting a bit too much voltage that in turn turns off the chips until you restart the miner (after dabbing away a fraction of the graphite with a fingertip)

try pulling out and re-inserting the board. i had a board dissapear the other day (after almost a week in place) and pulled it out, checked the connector and reinserted it. works fine again. I think if there is vibration from fans or weight from heatsinks it may be possible it simply isnt seated perfectly with all the connectors making contact.

I tried, no help... pheww... 0.0V
member
Activity: 85
Merit: 10
November 04, 2013, 07:18:01 PM
If anyone want to take a tracking of his Bitfury based on performance data of each card, here you find a macro for Excel which track all JSON-data each 5min. It delivers you some results as table and chart. You can analyze as you want, e.g. by card, bus, slot, SPIerr or MISOerr.
Just open in excel, enter IPadress and click start.

im getting debug errors through this whole chunk of code:

Workbooks(sFile).Worksheets("Config").Unprotect "bitcoin"
    Workbooks(sFile).Worksheets("Config").Cells(2, 2).Value = "logging stopped"
    Workbooks(sFile).Worksheets("Config").Cells(2, 2).Font.Bold = False
    Workbooks(sFile).Worksheets("Config").Cells(2, 2).Font.Color = RGB(0, 0, 0)
    Workbooks(sFile).Worksheets("Config").Protect "bitcoin"

and the similar chunk below for "logging started" ... "Bold = True" ... etc

am i missing something, or is there an underlying macro or feature i need to instigate?

Not your fault. Maybe you opened a different excel-file and work in it while the macro runs in background? Thats the only case i found for this problem.

I fixed it. Now you can work in different excel-files while data logging runs at background. But this is not a recommended using. If you get again an (other) error, please try at first this macro run alone without other files in background.
To get the new version please download here at http://www.bansemer.net/dl/BF-Datalogger.xlsm

Ciao

Enrico
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
November 04, 2013, 08:50:22 AM
If anyone want to take a tracking of his Bitfury based on performance data of each card, here you find a macro for Excel which track all JSON-data each 5min. It delivers you some results as table and chart. You can analyze as you want, e.g. by card, bus, slot, SPIerr or MISOerr.
Just open in excel, enter IPadress and click start.

Ciao

Enrico

im getting debug errors through this whole chunk of code:

Workbooks(sFile).Worksheets("Config").Unprotect "bitcoin"
    Workbooks(sFile).Worksheets("Config").Cells(2, 2).Value = "logging stopped"
    Workbooks(sFile).Worksheets("Config").Cells(2, 2).Font.Bold = False
    Workbooks(sFile).Worksheets("Config").Cells(2, 2).Font.Color = RGB(0, 0, 0)
    Workbooks(sFile).Worksheets("Config").Protect "bitcoin"

and the similar chunk below for "logging started" ... "Bold = True" ... etc

am i missing something, or is there an underlying macro or feature i need to instigate?
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
November 04, 2013, 08:45:59 AM
Okay, i made changes to settings, problem not solved...

And this is what i get...:

speed:2592 noncerate[GH/s]:96.651 (2.014/chip) hashrate[GH/s]:100.665 good:6751 errors:1585 spi-errors:1 mis$
4:      864     30.695  34.045  2144    339     0       1
8:      864     31.897  32.481  2228    849     0       0
C:      864     34.059  34.140  2379    397     1       0

Voltage still 0.3V

I measured resistors,
R01F:  2,120K ohm
R02F:  2,120K ohm

..so Ti chipis broken...?
 

1) switch to autotuning. clearly setting manual tuning to 54 across every chips is working terribly based on your 15-30% error rates.
2) R02F is quite high at 2.12K(ohm). try using the pencil mod to reduce it to 1.85K(ohm) and see what happens. If lucky, this is the issue and youll see the voltage rise to >0.6V. If it does, maybe the wrong resistor was placed and you should see if it can be changed for something in the 1.65-1.8K range

1. As you can see, m-board don't even detect BANK1 h-card(s). It is autotune mode now, but no help..
2. If I do pencil mod that, that will broke h-card warrantly...?

pencil modding is techincally messing with the warranty, but a tiny marking of pencil lead is not very hard to clean off and, unless you seriously manage to do it wrong, shouldnt cause anything worse then putting a bit too much voltage that in turn turns off the chips until you restart the miner (after dabbing away a fraction of the graphite with a fingertip)

try pulling out and re-inserting the board. i had a board dissapear the other day (after almost a week in place) and pulled it out, checked the connector and reinserted it. works fine again. I think if there is vibration from fans or weight from heatsinks it may be possible it simply isnt seated perfectly with all the connectors making contact.
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
November 04, 2013, 08:40:55 AM
A better test would be to power the card and measure voltage at the top of the Pulse inductor. If you get ~0.75v there (or whatever it's adjusted to), the power circuit (including that capacitor) is probably fine. I only say "probably" because I just fixed a board that was missing a cap on the back side. Before replacing that tiny back-side cap, the voltage at the Pulse top was correct, but the card wouldn't hash stably.

okay, i measured, it was 0.3V or 0V, what i can do next?

okay, lets start at the top:
1) you are forcing tuning to run at 55 when you use "aIfDSo 55". the fact you are seeing more than 10% errors means you should not have them higher than 53 or 54. change them to auto-tuning. (AIfDSo)
2) a few chips (such as 52) are set as "aifdso". this wont work in the bitfury system unless you were forcing custom timing clocks and other info to the chip (you aren't). change them to AIfDSo.
3) where are 1-30?
4) Your chips are producing way to many errors. half of their hash rate is actually errors. You need to cool your chips better. The chips currently at zero likely switched off because the autotuning was overwhelmed by the high error rate. you need to have good cooling fans, and likely add some heatsinks.
5) if you havent yet, update your chainminer following the method mentioned several times in this thread.

do the above and then check again. it should resolve 90% of your hashrate issues. as for the 0.3V you are measuring, try again after the changes. if its still <0.6V, check the values of the resistors R01F and R02F. in a board with capacitors these should be 1.0 and 1.25 k(ohm) respectively. In a board without capacitors they should be -?- and 1.7k(ohm) respectively (i dont know the R01F value for the board, but its likely not 1.0k like the first version)

if your resistors are not these values, thats the issue. if they are correct, then its onto the next step. probably the small TI chip found right beside them (it has about 12 or 14 pins) that is responsible for bringing 12v->0.7v

Wait...   are you saying that for chips that don't work/generate too many errors, setting them as aifdso doesn't work?? if not, what additional information needs to be provided to actually turn them off?? I can see a bunch of zero's in that line, which that seems to indicate they are technically switched off, so I don't understand why you're saying that aifdso alone doesn't fix the problem.... Thx for clarifying...

i meant for all the chips that are working, you have seriously high error rates with your manual tuning. autotuning will work a lot better.
legendary
Activity: 892
Merit: 1002
1 BTC =1 BTC
November 03, 2013, 10:05:40 AM
@Jumpsu, what happens when you leave the first one out and plug A2 in slot A1 ?

Did you try another psu ?
newbie
Activity: 29
Merit: 0
November 03, 2013, 07:08:18 AM
sr. member
Activity: 408
Merit: 250
November 02, 2013, 08:38:46 PM
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
November 02, 2013, 08:34:04 PM
newbie
Activity: 29
Merit: 0
November 02, 2013, 07:50:11 AM
Okay, i made changes to settings, problem not solved...
this what my rig looks,



Uploaded with ImageShack.us

And this is what i get...:


1       aIfDSo  54      1.990   2.124   139     2       0       1       201     [4:0]   629     9 9 9 9 9 9 $
2       aIfDSo  54      1.976   2.146   138     6       0       0       203     [4:1]   632     9 9 9 9 9 9 $
3       aIfDSo  54      2.090   2.114   146     0       0       0       200     [4:2]   622     9 9 9 9 9 9 $
4       aIfDSo  54      0.172   2.051   12      115     0       0       194     [4:3]   745     0 1 0 1 0 0 $
5       aIfDSo  54      1.675   1.998   117     80      0       0       189     [4:4]   650     7 7 7 7 7 7 $
6       aIfDSo  54      2.105   2.209   147     1       0       0       209     [4:5]   618     9 9 9 9 8 9 $
7       aIfDSo  54      2.162   2.156   151     3       0       0       204     [4:6]   617     10 10 10 9 9$
8       aIfDSo  54      2.205   2.241   154     20      0       0       212     [4:7]   620     10 10 10 10 $
9       aIfDSo  54      1.861   2.167   130     0       0       0       205     [4:8]   636     8 8 8 8 8 8 $
10      aIfDSo  54      1.976   2.220   138     5       0       0       210     [4:9]   627     9 9 9 9 9 8 $
11      aIfDSo  54      2.319   2.220   162     2       0       0       210     [4:A]   608     10 11 11 11 $
12      aIfDSo  54      1.861   2.051   130     14      0       0       194     [4:B]   639     8 8 8 8 8 8 $
13      aIfDSo  54      1.890   1.860   132     30      0       0       176     [4:C]   634     8 8 8 8 7 8 $
14      aIfDSo  54      2.119   2.135   148     3       0       0       202     [4:D]   623     9 9 9 9 9 9 $
15      aIfDSo  54      2.019   2.156   141     16      0       0       204     [4:E]   627     9 9 9 9 9 9 $
16      aIfDSo  54      2.276   2.198   159     42      0       0       208     [4:F]   614     10 10 10 10 $
17      aIfDSo  54      2.233   2.135   156     0       0       0       202     [8:0]   616     9 9 9 10 10 $
18      aIfDSo  54      2.033   1.966   142     46      0       0       186     [8:1]   632     9 9 8 8 8 8 $
19      aIfDSo  54      2.176   1.945   152     21      0       0       184     [8:2]   618     10 10 10 9 9$
20      aIfDSo  54      2.162   2.051   151     29      0       0       194     [8:3]   615     9 9 9 9 9 9 $
21      aIfDSo  54      2.147   2.156   150     19      0       0       204     [8:4]   625     9 9 9 9 9 9 $
22      aIfDSo  54      1.933   1.934   135     47      0       0       183     [8:5]   644     8 8 8 8 8 8 $
23      aIfDSo  54      1.761   1.934   123     20      0       0       183     [8:6]   640     8 8 8 8 8 8 $
24      aIfDSo  54      0.601   2.146   42      555     0       0       203     [8:7]   719     3 3 4 1 2 2 $
25      aIfDSo  54      2.348   2.156   164     22      0       0       204     [8:8]   605     10 10 10 10 $
26      aIfDSo  54      2.520   2.051   176     0       0       0       194     [8:9]   590     11 11 11 11 $
27      aIfDSo  54      2.119   1.945   148     51      0       0       184     [8:A]   625     9 9 9 9 9 9 $
28      aIfDSo  54      2.033   2.051   142     18      0       0       194     [8:B]   626     9 9 9 9 9 9 $
29      aIfDSo  54      1.933   2.040   135     0       0       0       193     [8:C]   628     8 8 8 9 9 9 $
30      aIfDSo  54      2.062   1.934   144     11      0       0       183     [8:D]   625     9 9 9 9 9 9 $
31      aIfDSo  54      1.775   2.008   124     0       0       0       190     [8:E]   638     7 7 8 8 8 8 $
32      aIfDSo  54      2.062   2.029   144     10      0       0       192     [8:F]   625     9 9 9 9 9 9 $
33      aIfDSo  54      2.004   2.230   140     32      0       0       211     [C:0]   634     8 8 8 8 8 8 $
34      aIfDSo  54      2.176   2.146   152     35      0       0       203     [C:1]   619     10 9 9 9 9 9$
35      aIfDSo  54      1.761   1.966   123     56      0       0       186     [C:2]   648     7 7 7 7 7 7 $
36      aIfDSo  54      2.348   2.177   164     44      0       0       206     [C:3]   614     10 10 10 10 $
37      aIfDSo  54      2.334   2.167   163     8       0       0       205     [C:4]   603     11 10 10 10 $
38      aIfDSo  54      1.546   1.522   108     0       0       0       144     [C:5]   652     7 7 7 7 7 7 $
39      aIfDSo  54      2.147   2.283   150     52      0       0       216     [C:6]   624     9 9 9 9 9 9 $
40      aIfDSo  54      2.190   2.167   153     33      0       0       205     [C:7]   621     9 9 9 9 9 9 $
41      aIfDSo  54      1.331   2.272   93      71      0       0       215     [C:8]   677     6 5 6 8 4 6 $
42      aIfDSo  54      2.276   2.241   159     23      1       0       212     [C:9]   613     10 10 10 10 $
43      aIfDSo  54      2.076   2.135   145     1       0       0       202     [C:A]   628     9 10 10 9 9 $
44      aIfDSo  54      2.634   2.167   184     25      0       0       205     [C:B]   593     11 11 11 11 $
45      aIfDSo  54      2.448   2.177   171     3       0       0       206     [C:C]   598     11 11 11 10 $
46      aIfDSo  54      2.391   2.124   167     12      0       0       201     [C:D]   613     11 11 11 11 $
47      aIfDSo  54      2.233   2.262   156     0       0       0       214     [C:E]   613     10 10 10 10 $
48      aIfDSo  54      2.162   2.103   151     2       0       0       199     [C:F]   617     9 9 9 9 9 9 $
speed:2592 noncerate[GH/s]:96.651 (2.014/chip) hashrate[GH/s]:100.665 good:6751 errors:1585 spi-errors:1 mis$
4:      864     30.695  34.045  2144    339     0       1
8:      864     31.897  32.481  2228    849     0       0
C:      864     34.059  34.140  2379    397     1       0

Voltage still 0.3V

I measured resistors,
R01F:  2,120K ohm
R02F:  2,120K ohm
R01S:  3,3      ohm
R01L:  14,66K ohm
R01A:  98,80K ohm
R01B:  2,5      ohm
R02E:  68,90K ohm
R01E:  68,80K ohm
R01T1:187,3K ohm
R01T:  146,7K ohm
R01M:  219,8K ohm

..so Ti chipis broken...?
 
legendary
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
November 02, 2013, 03:48:12 AM
My FrankenFury

Got my hands on one of H board no M board so way not create your own FrankenFury  Wink




 Grin


Don't you need a level shifter on the comm lines?

I guess  not  Grin

Only problem that I run across is 1.8v on pin 1 of H board but luckily RPi has 1.8v (not on GPIO).I did this Tuesday night and board is running between 29 and 31GH (after pencil mod).


I do need to give credit to sveetsnelda for his help.






Hey cool man great job. Any chance you got the schematics?
newbie
Activity: 23
Merit: 0
November 02, 2013, 12:49:12 AM
My FrankenFury

Got my hands on one of H board no M board so way not create your own FrankenFury  Wink

http://imageshack.us/a/img824/1081/750f.jpg


 Grin


Don't you need a level shifter on the comm lines?

I guess  not  Grin

Only problem that I run across is 1.8v on pin 1 of H board but luckily RPi has 1.8v (not on GPIO).I did this Tuesday night and board is running between 29 and 31GH (after pencil mod).


I do need to give credit to sveetsnelda for his help.




Pages:
Jump to: