Bitcoin Forum
November 23, 2017, 09:00:18 PM *
News: Latest stable version of Bitcoin Core: 0.15.1  [Torrent].
 
   Home   Help Search Donate Login Register  
Pages: « 1 2 3 4 [5] 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 »
  Print  
Author Topic: Fury/Blizzard tuning and mods  (Read 104104 times)
J4bberwock
Hero Member
*****
Offline Offline

Activity: 840


View Profile
June 14, 2014, 06:10:31 PM
 #81

At clock 515, coinshift says i'm hashing at 4.8-5.0 MH including rejected for 2 Furys
I'll add 2 more overclocked ones and let them run all night long to confirm.
If it's real, I have created the first Zeus mini-blade.

Custom Server PSU breakout boards, 1200w, 1300w, 2000w, 2880w https://bitcointalk.org/index.php?topic=738527.0
Btc: 1J4bberWs6f6XVQ53gn3BNc8qHAicmm7wu
Ltc: LNDmMBtzD3cduwjfxANm2wDrxt9jXRk5ZP
1511470818
Hero Member
*
Offline Offline

Posts: 1511470818

View Profile Personal Message (Offline)

Ignore
1511470818
Reply with quote  #2

1511470818
Report to moderator
Join ICO Now Coinlancer is Disrupting the Freelance marketplace!
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1511470818
Hero Member
*
Offline Offline

Posts: 1511470818

View Profile Personal Message (Offline)

Ignore
1511470818
Reply with quote  #2

1511470818
Report to moderator
1511470818
Hero Member
*
Offline Offline

Posts: 1511470818

View Profile Personal Message (Offline)

Ignore
1511470818
Reply with quote  #2

1511470818
Report to moderator
1511470818
Hero Member
*
Offline Offline

Posts: 1511470818

View Profile Personal Message (Offline)

Ignore
1511470818
Reply with quote  #2

1511470818
Report to moderator
happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 14, 2014, 06:27:15 PM
 #82

Now that we have schematics, should we refer to these resistors as R10 and R9?

Texas Instruments pdf on top
Zeusminer PDF on bottom


Zeus schematic


J4bberwock - could you please confirm that this looks accurate.

If so we can update the formula to:
R10=0.591xR9/(desired voltage-0.591)

Zeus doc indicates R10=10.5k and R9=10k
We don't know if they have made any changes so maybe best to measure and plug into the formula I think.

TI doc seems to indicate R10 (TI R7) as being the resistor to change. Seems like R9 (TI R8) has additional function - Error Amplifier Compensation - I don't enough about this stuff to know for sure.  J4bberwock hope you can help me out here.

It's so great that we have the schematics to refer to now.
happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 14, 2014, 06:39:25 PM
 #83

J4bberwock & nst6563 are you running them at the same or similar voltage?   Any idea of watts (at x volts)?

nst6563
Sr. Member
****
Offline Offline

Activity: 252


View Profile
June 14, 2014, 07:06:23 PM
 #84

J4bberwock & nst6563 are you running them at the same or similar voltage?   Any idea of watts (at x volts)?



The voltage mine is reading now is 1.54v (original reading was 1.26v) and I can run it at 520Mhz and stay in the 3%-4% error rate.  For whatever reason, it seems that with Hashra it stops displaying the correct hashrate above 520Mhz.  I've looked through the code a bit and couldn't really find anything that would limit what it's displaying so I think it may be in whatever version of cgminer it is using.  I know in a litecointalk thread Maxzilla (who supports Hashra) asked if he could use the new builds of bfgminer that are being worked on but I haven't had the time to get the files on there and do a manual run of it to see if bfgminer has the same issue with clocks above 520.  

As for power usage....I have no idea.  A Kilowatt and decent clamp ammeter is something I've been meaning to get for a while now and always end up getting distracted by something else lol
happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 14, 2014, 07:14:08 PM
 #85

Still using 7k resistor value?
nst6563
Sr. Member
****
Offline Offline

Activity: 252


View Profile
June 14, 2014, 07:30:06 PM
 #86

actually it's around 6k now.  I adjusted it just slightly.  It seemed that with my fury the max I could get with 7k was 500 with less than 4% error.  I adjusted the pot down .2k and can now hit 520.  Possibly higher, but I won't know with the current version of hashra/cgminer - I'll have to try bfgminer.
ZiG
Sr. Member
****
Offline Offline

Activity: 406


View Profile
June 14, 2014, 08:07:29 PM
 #87

Now that we have schematics, should we refer to these resistors as R10 and R9?

Texas Instruments pdf on top
Zeusminer PDF on bottom


Zeus schematic


J4bberwock - could you please confirm that this looks accurate.

If so we can update the formula to:
R10=0.591xR9/(desired voltage-0.591)

Zeus doc indicates R10=10.5k and R9=10k
We don't know if they have made any changes so maybe best to measure and plug into the formula I think.

TI doc seems to indicate R10 (TI R7) as being the resistor to change. Seems like R9 (TI R8) has additional function - Error Amplifier Compensation - I don't enough about this stuff to know for sure.  J4bberwock hope you can help me out here.

It's so great that we have the schematics to refer to now.

Yep...you are correct...

R10 ( Zeus ) is the one to modify...

DON"T TOUCH R9 ...it's part of the Compensation filter as well...could create stability problems...

ZiG

EDIT...:

"Feedback Divider (R7, R8)
Select R8 to be between 10 kΩ and 100 kΩ. For this design, select 20 kΩ. R7 is then selected to produce the
desired output voltage when VFB = 0.591 V using Equation 30

R7 = ( R8 * Vfb ) / ( Vout - Vfb )

VFB = 0.591 V and R8 = 20 kΩ for VOUT = 1.8 V, R7 = 9.78 kΩ, so the value of 9.76 kΩ is selected as the closest standard value.
A slightly lower nominal value increases the nominal output voltage slightly to compensate for some trace impedance at load."

Ti R8 is "Zeus" R9 ...Ti R7 is "Zeus" R10...capisco... Grin


nst6563
Sr. Member
****
Offline Offline

Activity: 252


View Profile
June 14, 2014, 10:06:41 PM
 #88

updated first post with link to new bfgminer 4.1.99

https://litecointalk.org/index.php?topic=16301.5760

Sources:
https://www.dropbox.com/sh/fygluxm446cjctn/AAAFUAJDeIEYiWnuEefimKjWa
happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 14, 2014, 10:44:50 PM
 #89

This is more recent: https://litecointalk.org/index.php?topic=16301.msg182643#msg182643
It gives a little higher hash rate and less HW errors.

Darkwinde also mentioned he will merge the work Jstefanop did and have a new version soon.  Maybe this weekend!
nst6563
Sr. Member
****
Offline Offline

Activity: 252


View Profile
June 15, 2014, 12:36:46 AM
 #90

Newer?  Wow...the one I linked to was from today.  They're working fast.

ManeBjorn
Legendary
*
Offline Offline

Activity: 1064



View Profile
June 15, 2014, 12:52:19 AM
 #91

I know I kept going through the thread and boom another and another.  LOL
It rocks though.  Soon it will be in the main branch of BFG that will be awesome.
I like the stability and speed increases they are getting.


Newer?  Wow...the one I linked to was from today.  They're working fast.



happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 15, 2014, 01:16:48 AM
 #92

Here's Darkwinde's newest: https://litecointalk.org/index.php?topic=16301.msg183015#msg183015

Note: --chips-count is now --zeus-cc
 --ltc-clk is now --zeus-clk
nst6563
Sr. Member
****
Offline Offline

Activity: 252


View Profile
June 15, 2014, 01:30:12 AM
 #93

Is there an rpi image like Hashra/Starminer/Minera that supports the bfgminer software and not cgminer (or that supports both)?
ManeBjorn
Legendary
*
Offline Offline

Activity: 1064



View Profile
June 15, 2014, 03:54:04 AM
 #94

So fast.  That is why I love crypto currencies.  We have the best minds always working and improving at a fast rate.
 Grin

Here's Darkwinde's newest: https://litecointalk.org/index.php?topic=16301.msg183015#msg183015

Note: --chips-count is now --zeus-cc
 --ltc-clk is now --zeus-clk

J4bberwock
Hero Member
*****
Offline Offline

Activity: 840


View Profile
June 15, 2014, 06:10:05 AM
 #95

Now that we have schematics, should we refer to these resistors as R10 and R9?

Texas Instruments pdf on top
Zeusminer PDF on bottom


Zeus schematic


J4bberwock - could you please confirm that this looks accurate.

If so we can update the formula to:
R10=0.591xR9/(desired voltage-0.591)

Zeus doc indicates R10=10.5k and R9=10k
We don't know if they have made any changes so maybe best to measure and plug into the formula I think.

TI doc seems to indicate R10 (TI R7) as being the resistor to change. Seems like R9 (TI R8) has additional function - Error Amplifier Compensation - I don't enough about this stuff to know for sure.  J4bberwock hope you can help me out here.

It's so great that we have the schematics to refer to now.

R9 and R11 are swapped on my boards compared to the data from Zeus. the 900R is on the right side on my pictures. And the 9.3k is on the left. You can also see that they were supposed to be around 10k and we have 9k on board. check the markings on your resistor before swapping and use the formula if needed

Custom Server PSU breakout boards, 1200w, 1300w, 2000w, 2880w https://bitcointalk.org/index.php?topic=738527.0
Btc: 1J4bberWs6f6XVQ53gn3BNc8qHAicmm7wu
Ltc: LNDmMBtzD3cduwjfxANm2wDrxt9jXRk5ZP
J4bberwock
Hero Member
*****
Offline Offline

Activity: 840


View Profile
June 15, 2014, 09:05:55 AM
 #96

Regarding the power draw that everybody have asked.
2 overclocked Zeus mini-blades (Copyright to me Grin ) are currently drawing 200 watt at the wall when idle 400 watt when hashing.


It's 100 watt for each  Fury/Blizzard overvolted to 1.65V from 1.26V when hashing
I haven't measured stock power draw.

Next step for me will be to overclock the third mini-blade to 1.8-1.9V, maybe 2v to check for increased power draw and hashrate.
And after the limits are reached on Zeus, move to a new overvolting/overclocking challenge. Unfortunately, A2 is still too expensive for me to buy one simply to play with it.

Custom Server PSU breakout boards, 1200w, 1300w, 2000w, 2880w https://bitcointalk.org/index.php?topic=738527.0
Btc: 1J4bberWs6f6XVQ53gn3BNc8qHAicmm7wu
Ltc: LNDmMBtzD3cduwjfxANm2wDrxt9jXRk5ZP
happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 15, 2014, 11:10:55 AM
 #97

R9 and R11 are swapped on my boards compared to the data from Zeus. the 900R is on the right side on my pictures. And the 9.3k is on the left. You can also see that they were supposed to be around 10k and we have 9k on board. check the markings on your resistor before swapping and use the formula if needed

I figured they must have been swapped on yours.  I'll check mine when I take them apart.
happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 15, 2014, 11:20:07 AM
 #98

Regarding the power draw that everybody have asked.
2 overclocked Zeus mini-blades (Copyright to me Grin ) are currently drawing 200 watt at the wall when idle 400 watt when hashing.


It's 100 watt for each  Fury/Blizzard overvolted to 1.65V from 1.26V when hashing
I haven't measured stock power draw.

Next step for me will be to overclock the third mini-blade to 1.8-1.9V, maybe 2v to check for increased power draw and hashrate.
And after the limits are reached on Zeus, move to a new overvolting/overclocking challenge. Unfortunately, A2 is still too expensive for me to buy one simply to play with it.

Thanks for the info.  Stock is pretty close to 50 watts so it doubles.  Looks like both of mine are able to maintain 1.4 mh/s at around 50 watts.

I'm curious to know what hash rate you and nst6563 are getting.
nst6563
Sr. Member
****
Offline Offline

Activity: 252


View Profile
June 15, 2014, 12:04:35 PM
 #99

Right now (with Hashra and cgminer) it seems to top out at around 1.5.  It also depends on the pool I'm mining at.  If I mine at Nicehash, I get a slightly lower hashrate (usually around 1.4-1.5) with a LOT higher number of rejects.  If I mine at Manicminer I get  nearly 0 rejects and my rate locally shows 1.51mhs-1.52mhs while poolside shows a range of 1.5mhs-1.95mhs (and it stays above 1.5mhs quite a bit).

I found Minepeon which uses bfgminer as it's mining software so I'm getting ready to try that this morning and put the new version of bfgminer from Darkwinde on there.  People in that thread at litecointalk were reportedly getting higher hashrates with less errors at lower speeds than we are (something like 1.5-1.6mhs at 340Mhz clock).

I updated the first post with the Minepeon link.
happydaze
Full Member
***
Offline Offline

Activity: 140


View Profile
June 15, 2014, 12:57:23 PM
 #100

J4bberwock - what's your hash rate at?

nst6563 - is this summary right? 6k resistor, 1.54v,  520 clk, 1.5mh/s +
Wish you had a way to measure amps or watts.

I'm excited about the mod but with LTC today for example if the gain is 100 kh/s for 50 watts then we'd actually be losing money at 10 cents kWh.
I think your volts (and watts) might be lower than J4bberwock's.







Pages: « 1 2 3 4 [5] 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!