Bitcoin Forum
November 20, 2017, 07:56:47 AM *
News: Latest stable version of Bitcoin Core: 0.15.1  [Torrent].
 
   Home   Help Search Donate Login Register  
Pages: « 1 ... 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 [2137] 2138 »
  Print  
Author Topic: Swedish ASIC miner company kncminer.com  (Read 3008434 times)
lightfoot
Legendary
*
Offline Offline

Activity: 1330

I fix broken miners. And make holes in teeth :-)


View Profile
August 08, 2017, 01:44:56 AM
 #42721

Tarkin, something to be aware of: I think each supply is programmed slightly differently to take advantage of a master clock signal. The supplies beat at 180 degrees out of phase with each other, which is why you can't swap a top and bottom supply in a pair for a die. However I think each supply actually beats off an 8 phase clock. In other words:

First pair beats 0,180
Second pair beats 45,225
Third pair beats 90,270
Fourth pair beats 135,225

Result is the board runs like a friggin 8 phase power supply. No wonder the filter caps are small.

But it means each supply is programmed to be what it is based on it's position on the board. very clever.

C
Join ICO Now A blockchain platform for effective freelancing
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1511164607
Hero Member
*
Offline Offline

Posts: 1511164607

View Profile Personal Message (Offline)

Ignore
1511164607
Reply with quote  #2

1511164607
Report to moderator
1511164607
Hero Member
*
Offline Offline

Posts: 1511164607

View Profile Personal Message (Offline)

Ignore
1511164607
Reply with quote  #2

1511164607
Report to moderator
1511164607
Hero Member
*
Offline Offline

Posts: 1511164607

View Profile Personal Message (Offline)

Ignore
1511164607
Reply with quote  #2

1511164607
Report to moderator
GenTarkin
Legendary
*
Offline Offline

Activity: 2170


View Profile
August 08, 2017, 05:47:50 AM
 #42722

Tarkin, something to be aware of: I think each supply is programmed slightly differently to take advantage of a master clock signal. The supplies beat at 180 degrees out of phase with each other, which is why you can't swap a top and bottom supply in a pair for a die. However I think each supply actually beats off an 8 phase clock. In other words:

First pair beats 0,180
Second pair beats 45,225
Third pair beats 90,270
Fourth pair beats 135,225

Result is the board runs like a friggin 8 phase power supply. No wonder the filter caps are small.

But it means each supply is programmed to be what it is based on it's position on the board. very clever.

C
Im pretty sure you are correct. From what Ericsson told me, each pair of DCDC's is configured as its own common current group, which I can see evidence of in certain registers of each DCDC. Also, it seems they use interleaving for each of the 4 groups, the interleaving w/ the combined common setting of automatic phase control in each group ... I think gives it the unique phases per board =)

GenTarkin's MOD Kncminer Titan custom firmware! v1.0.4! <--- CLICK HERE
Donations: bitcoin- 1Px71mWNQNKW19xuARqrmnbcem1dXqJ3At || litecoin- LYXrLis3ik6TRn8tdvzAyJ264DRvwYVeEw
swedenrt
Jr. Member
*
Offline Offline

Activity: 50

feeldiam


View Profile WWW
August 08, 2017, 10:47:07 PM
 #42723

still this is working_

[MULTIBOT] ▬ Powerful multithreaded trading platform
❰▬▬ Token Sale Start 20th Oct[/url
Searing
Legendary
*
Offline Offline

Activity: 1540


Tips for help? 1BzbfMHCrTeLjc7eCGrYVhH3QXSRodSuke


View Profile
August 09, 2017, 04:47:46 AM
 #42724

Tarkin, something to be aware of: I think each supply is programmed slightly differently to take advantage of a master clock signal. The supplies beat at 180 degrees out of phase with each other, which is why you can't swap a top and bottom supply in a pair for a die. However I think each supply actually beats off an 8 phase clock. In other words:

First pair beats 0,180
Second pair beats 45,225
Third pair beats 90,270
Fourth pair beats 135,225

Result is the board runs like a friggin 8 phase power supply. No wonder the filter caps are small.

But it means each supply is programmed to be what it is based on it's position on the board. very clever.

C
Im pretty sure you are correct. From what Ericsson told me, each pair of DCDC's is configured as its own common current group, which I can see evidence of in certain registers of each DCDC. Also, it seems they use interleaving for each of the 4 groups, the interleaving w/ the combined common setting of automatic phase control in each group ... I think gives it the unique phases per board =)



Does any of this help the 'magic' gen tarkin firmware from going in and doing some kinda diagnostic ..to get dead dies back in some manner (or say comes back with a list
of recommended settings to attempt)

long shot here...but I'm 'barely' following the conversation..so maybe all this work you are doing on this...that the above idea is 'implied' and I'm too far behind the curve

on the tech speak to realize this Smiley

as an aside

you and lightfoot really need to get out more....find some new asics to toy with...contact bitmain or something ...god knows enough of the stuff could use some tweaks

or how to fix ..that you guys know how to do..if given the proper specs and motivations




       ▀
   ▄▄▄   ▄▀
   ███ ▄▄▄▄  ██
       ████
    ▄  ▀▀▀▀
▄▄
      ██    ▀▀
██▄█▄▄▄████████
▄▄▄▄▄▄▄▄▀▀███▀▀▀
██████████████████
████▄▀▄▀▄▀███▀▀▀▀▀
████▄▀▄▀▄▀███ ▀
████▄▀▄▀▄▀████████
▀█████████████████
]
CoinPayments
█████
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████
█████
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████
█████
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████ ██
█████
GenTarkin
Legendary
*
Offline Offline

Activity: 2170


View Profile
August 09, 2017, 05:51:18 AM
 #42725

Tarkin, something to be aware of: I think each supply is programmed slightly differently to take advantage of a master clock signal. The supplies beat at 180 degrees out of phase with each other, which is why you can't swap a top and bottom supply in a pair for a die. However I think each supply actually beats off an 8 phase clock. In other words:

First pair beats 0,180
Second pair beats 45,225
Third pair beats 90,270
Fourth pair beats 135,225

Result is the board runs like a friggin 8 phase power supply. No wonder the filter caps are small.

But it means each supply is programmed to be what it is based on it's position on the board. very clever.

C
Im pretty sure you are correct. From what Ericsson told me, each pair of DCDC's is configured as its own common current group, which I can see evidence of in certain registers of each DCDC. Also, it seems they use interleaving for each of the 4 groups, the interleaving w/ the combined common setting of automatic phase control in each group ... I think gives it the unique phases per board =)



Does any of this help the 'magic' gen tarkin firmware from going in and doing some kinda diagnostic ..to get dead dies back in some manner (or say comes back with a list
of recommended settings to attempt)

long shot here...but I'm 'barely' following the conversation..so maybe all this work you are doing on this...that the above idea is 'implied' and I'm too far behind the curve

on the tech speak to realize this Smiley

as an aside

you and lightfoot really need to get out more....find some new asics to toy with...contact bitmain or something ...god knows enough of the stuff could use some tweaks

or how to fix ..that you guys know how to do..if given the proper specs and motivations




I dont think theres much more tweaking that can be done to Titan, aside from switching speed to 200khz to shave another few watts off the meter. But, at this point in Titan's life cycle ... it may be a moot point to write all the code to implement the changes so its easy for end users =/ time will tell tho. I still got my Titan =)  ... Ill be getting an L3+ in late september and Ill dig around in the firmware see if anything can stand to be optimized.

GenTarkin's MOD Kncminer Titan custom firmware! v1.0.4! <--- CLICK HERE
Donations: bitcoin- 1Px71mWNQNKW19xuARqrmnbcem1dXqJ3At || litecoin- LYXrLis3ik6TRn8tdvzAyJ264DRvwYVeEw
lightfoot
Legendary
*
Offline Offline

Activity: 1330

I fix broken miners. And make holes in teeth :-)


View Profile
August 09, 2017, 10:31:41 AM
 #42726

Does any of this help the 'magic' gen tarkin firmware from going in and doing some kinda diagnostic ..to get dead dies back in some manner (or say comes back with a list
of recommended settings to attempt)
At this point, the only thing I can't fix is boards with a specific type of hard-shorted hashing die caused by major voltage spikes or way too much heat. Everything else (blown supplies, shorted 12v rails, blown drivers (a lot of these), blown fpga's, blown power controllers, dead crystals, etc) can be fixed.

Quote
you and lightfoot really need to get out more....find some new asics to toy with...contact bitmain or something ...god knows enough of the stuff could use some tweaks or how to fix ..that you guys know how to do..if given the proper specs and motivations

Hm.....

I've thought about that. I have worked on Bitmain's S7/S9 series, Avalon A6/A7's, Bakali's, and even some MAT systems in addition to KNC. Vendors don't seem to want to talk (I don't know, maybe liability, maybe trade secrets, maybe because I publish what I find to help others on this forum) but if a vendor wanted me to help them I'd be happy to do so. As long as it's with the understanding that my goal is to help miners mine, and if that includes telling miners to slow down their systems, watch out for these issues, etc so be it.

Otherwise I can reverse-engineer enough to figure out pretty much any problem on my own (as evidenced by my fixing all these kinds of systems). Takes time, but keeps me out of trouble so to speak, and to be honest it's an intellectual challenge.... :-)

C

boosti
Newbie
*
Offline Offline

Activity: 4


View Profile
September 02, 2017, 06:29:09 PM
 #42727

Hi GenTarkin

I loaded today your firmware and started the energy save mode but somehow the first step don't finish.

I get the following message like three times in 5h:
http://up.picr.de/30251662kw.jpg


After the last time the miner didnt load anymore the settings, just came this picture:

http://up.picr.de/30251672kk.jpg


I shows me BFG Miner 5.0.0 and I recognized in some of your pictures on the webpage I visible 5.1.0.  Is there a newer update available I might should try?

Thanks for your help
GenTarkin
Legendary
*
Offline Offline

Activity: 2170


View Profile
September 02, 2017, 11:12:33 PM
 #42728

Hi GenTarkin

I loaded today your firmware and started the energy save mode but somehow the first step don't finish.

I get the following message like three times in 5h:



After the last time the miner didnt load anymore the settings, just came this picture:




I shows me BFG Miner 5.0.0 and I recognized in some of your pictures on the webpage I visible 5.1.0.  Is there a newer update available I might should try?

Thanks for your help

Looks to me like your Titan image is borked .... BFGMiner being version 5.0.0 tells me you probably loaded my firmware upgrade on KNC firmware older than 2.00 ... my firmware is only officially tested and working on that version or 2.02

GenTarkin's MOD Kncminer Titan custom firmware! v1.0.4! <--- CLICK HERE
Donations: bitcoin- 1Px71mWNQNKW19xuARqrmnbcem1dXqJ3At || litecoin- LYXrLis3ik6TRn8tdvzAyJ264DRvwYVeEw
boosti
Newbie
*
Offline Offline

Activity: 4


View Profile
September 11, 2017, 11:20:37 AM
 #42729

Thanks man it workes fine now, I totally missed the information to update from 2x, I did from a 1.9x KNC firmware. Now all is running smoothly Smiley
pacnw
Newbie
*
Offline Offline

Activity: 12


View Profile
October 15, 2017, 12:45:46 AM
 #42730

I'm confused what state I'm at in the process. My Titan has been running flawlessly since 2014 getting app 350 Mh/s with the stock settings of 325 MHz and -.0366 V but has always run hot. I never saw any real benefit to running the GenTarkin firmware but after recently adding an L3+ to my home mining, I thought I try the GenTarkin firmware Energy Saver feature in an attempt to get some heat out of my environment. I started the Energy Saver and it immediately made some changes including turning off dies with temps over the 90 degree threshold (e.g.: Dies currently throttled due to DCDC over temp threshold) and reducing a few to 300 MHz

Based on a reading of the documentation, I was expecting nothing to happen immediately since it says:

When energy saver is turned on. Stage 1 of 3 begins: It will restart bfgminer, have an initial 1.5hr cooldown then the first voltage adjustments will be made. All enabled die voltages will be lowered a single “notch”.

I didn't experience a 1.5 hr cooldown before adjustments were made but experienced immediate adjustments.

Also, the Energy Saver immediately went into an "IDLE" state which I interpret as nothing happening based on Status – IDLE – when no tuning is in progress from the documentation

Is something taking place? If not, how do I leave the IDLE state and restart the Energy Saver process?

Thanks
lightfoot
Legendary
*
Offline Offline

Activity: 1330

I fix broken miners. And make holes in teeth :-)


View Profile
October 15, 2017, 01:35:15 AM
 #42731

I don't know, but running the DC-DC's at 90c is the best way to get a unit ready for shipment to me :-) High temps like that can cause the gates to short, resulting in an instant shorted cube with possibly a short to the SCL lines as well.

Not too bad to fix, but run the cubes with die temps under 45c and supply temps under 80 and they will live long and happy little lives.

Meantime I have been working on antminers, S7's and S9's have pretty much given up their key secrets, now just need to play with an A4....

C
pacnw
Newbie
*
Offline Offline

Activity: 12


View Profile
October 15, 2017, 02:34:42 AM
 #42732

I think what might have happened is that when I moved the miners for the first time in about 1.5 years the other day and blew out the huge amount of dust inside of them, I might have removed the insulating factor provided by the dust. The cubes have always run hot in the 3 years of flawless - no maintenance - operation  up to this point in time, but they seemed to be running hotter lately so that's why I ran the GenTarkin firmware which currently seems to be doing nothing/is idle and about which I'm unclear as to next steps. I might just end up restarting.

Based on past reading, your quite deep into these units. Do you have any how-to info about opening - and being able to reclose - the cubes and any info about the thermal paste application process? It looks like rivets to me that would have to be cut.

Thanks
lightfoot
Legendary
*
Offline Offline

Activity: 1330

I fix broken miners. And make holes in teeth :-)


View Profile
October 15, 2017, 02:43:38 AM
 #42733

I think what might have happened is that when I moved the miners for the first time in about 1.5 years the other day and blew out the huge amount of dust inside of them, I might have removed the insulating factor provided by the dust. The cubes have always run hot in the 3 years of flawless - no maintenance - operation  up to this point in time, but they seemed to be running hotter lately so that's why I ran the GenTarkin firmware which currently seems to be doing nothing/is idle and about which I'm unclear as to next steps. I might just end up restarting.

Based on past reading, your quite deep into these units. Do you have any how-to info about opening - and being able to reclose - the cubes and any info about the thermal paste application process? It looks like rivets to me that would have to be cut.

Thanks
Nope. Those are 10 or 15 sized torx bolts. The thermal paste is only for the main core chip. Swapping it won't help the DC-DC's.

Just run them at 250mhz per die on 4 die units and 300mhz on 3 die units. Solid 65mh will also keep the power plugs from melting.
GenTarkin
Legendary
*
Offline Offline

Activity: 2170


View Profile
October 15, 2017, 08:03:02 AM
 #42734

I'm confused what state I'm at in the process. My Titan has been running flawlessly since 2014 getting app 350 Mh/s with the stock settings of 325 MHz and -.0366 V but has always run hot. I never saw any real benefit to running the GenTarkin firmware but after recently adding an L3+ to my home mining, I thought I try the GenTarkin firmware Energy Saver feature in an attempt to get some heat out of my environment. I started the Energy Saver and it immediately made some changes including turning off dies with temps over the 90 degree threshold (e.g.: Dies currently throttled due to DCDC over temp threshold) and reducing a few to 300 MHz

Based on a reading of the documentation, I was expecting nothing to happen immediately since it says:

When energy saver is turned on. Stage 1 of 3 begins: It will restart bfgminer, have an initial 1.5hr cooldown then the first voltage adjustments will be made. All enabled die voltages will be lowered a single “notch”.

I didn't experience a 1.5 hr cooldown before adjustments were made but experienced immediate adjustments.

Also, the Energy Saver immediately went into an "IDLE" state which I interpret as nothing happening based on Status – IDLE – when no tuning is in progress from the documentation

Is something taking place? If not, how do I leave the IDLE state and restart the Energy Saver process?

Thanks

Sorry for the confusion, The documentation on "energy saver" states this as well:
"Is there anything I should do prior to enabling “Energy Saver”?
To get the quickest and most reliable results: Please turn off dies which are very unstable or problematic. Please set dies listed in auto-byass state to “OFF” and clear the bypass list. Please ensure that NO dies are in a throttled state."

So, those dies which were throttled at the onset of your "ENABLE ENERGY SAVER" .... probably caused some weird behaviour w/ the dies that were throttled.

I would recommend ... doing as the documentation says. If ur running 90C on the DCDC temps .... thats quite a risk, you can adjust the maximum DCDC threshold up higher if you want to start energy saver reliably ... but I dont recommend it at your temps at all.

GenTarkin's MOD Kncminer Titan custom firmware! v1.0.4! <--- CLICK HERE
Donations: bitcoin- 1Px71mWNQNKW19xuARqrmnbcem1dXqJ3At || litecoin- LYXrLis3ik6TRn8tdvzAyJ264DRvwYVeEw
SupremeGoose
Jr. Member
*
Offline Offline

Activity: 56


View Profile WWW
October 15, 2017, 01:52:09 PM
 #42735

Hello. Probably stupid question, but, I will try anyway:

What to do with my old Jupiter ASIC miner? Is it just a trash now or I can use it for something?

PS: (To heat the house in winter - I know Wink ).

lightfoot
Legendary
*
Offline Offline

Activity: 1330

I fix broken miners. And make holes in teeth :-)


View Profile
October 15, 2017, 02:19:17 PM
 #42736

Offer the controller board for sale on Ebay or something. With my Lightfoot code it can run Titans.

(Just a data point: I think I got one donation for writing/promoting that code. Fixed a lot of people who had Pi's blow up, just noted)
pacnw
Newbie
*
Offline Offline

Activity: 12


View Profile
October 15, 2017, 06:54:13 PM
 #42737

I'm confused what state I'm at in the process. My Titan has been running flawlessly since 2014 getting app 350 Mh/s with the stock settings of 325 MHz and -.0366 V but has always run hot. I never saw any real benefit to running the GenTarkin firmware but after recently adding an L3+ to my home mining, I thought I try the GenTarkin firmware Energy Saver feature in an attempt to get some heat out of my environment. I started the Energy Saver and it immediately made some changes including turning off dies with temps over the 90 degree threshold (e.g.: Dies currently throttled due to DCDC over temp threshold) and reducing a few to 300 MHz

Based on a reading of the documentation, I was expecting nothing to happen immediately since it says:

When energy saver is turned on. Stage 1 of 3 begins: It will restart bfgminer, have an initial 1.5hr cooldown then the first voltage adjustments will be made. All enabled die voltages will be lowered a single “notch”.

I didn't experience a 1.5 hr cooldown before adjustments were made but experienced immediate adjustments.

Also, the Energy Saver immediately went into an "IDLE" state which I interpret as nothing happening based on Status – IDLE – when no tuning is in progress from the documentation

Is something taking place? If not, how do I leave the IDLE state and restart the Energy Saver process?

Thanks

Sorry for the confusion, The documentation on "energy saver" states this as well:
"Is there anything I should do prior to enabling “Energy Saver”?
To get the quickest and most reliable results: Please turn off dies which are very unstable or problematic. Please set dies listed in auto-byass state to “OFF” and clear the bypass list. Please ensure that NO dies are in a throttled state."

So, those dies which were throttled at the onset of your "ENABLE ENERGY SAVER" .... probably caused some weird behaviour w/ the dies that were throttled.

I would recommend ... doing as the documentation says. If ur running 90C on the DCDC temps .... thats quite a risk, you can adjust the maximum DCDC threshold up higher if you want to start energy saver reliably ... but I dont recommend it at your temps at all.


To get out of the idle state, I manually set the Mhz rate on all dies to 300 MHz and Applied changes. After applying the changes the Energy Saver start button became active and I started/restarted the Energy Saver. Stage 1 was chugging along last night and Stage 2 was running when I checked this morning. It appears all is good. Thanks for the replies

SupremeGoose
Jr. Member
*
Offline Offline

Activity: 56


View Profile WWW
October 21, 2017, 09:29:19 AM
 #42738

Offer the controller board for sale on Ebay or something. With my Lightfoot code it can run Titans.

(Just a data point: I think I got one donation for writing/promoting that code. Fixed a lot of people who had Pi's blow up, just noted)

Thank you!

Can you give me a price recommendation please? Because I was searching, and I cannot find any even approximate price.

Or maybe if you want (or anybody) this controller board, DM me with an offer.

lightfoot
Legendary
*
Offline Offline

Activity: 1330

I fix broken miners. And make holes in teeth :-)


View Profile
October 21, 2017, 02:09:15 PM
 #42739

Offer the controller board for sale on Ebay or something. With my Lightfoot code it can run Titans.

(Just a data point: I think I got one donation for writing/promoting that code. Fixed a lot of people who had Pi's blow up, just noted)

Thank you!

Can you give me a price recommendation please? Because I was searching, and I cannot find any even approximate price.

Or maybe if you want (or anybody) this controller board, DM me with an offer.
Toughie. There are only two in the recent Ebay sold history, so I'd say put it up and see what happens.
(Note those prices included a Pi so they would run Titans natively)

LongAndShort
Legendary
*
Offline Offline

Activity: 1022


View Profile
October 25, 2017, 11:52:19 AM
 #42740

Offer the controller board for sale on Ebay or something. With my Lightfoot code it can run Titans.

(Just a data point: I think I got one donation for writing/promoting that code. Fixed a lot of people who had Pi's blow up, just noted)

Did you only get a donation from me? Damn, that makes me sick!
Pages: « 1 ... 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 [2137] 2138 »
  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!