Author

Topic: Swedish ASIC miner company kncminer.com - page 317. (Read 3049515 times)

sr. member
Activity: 434
Merit: 250
Quick update:
KnC still haven't agreed refund or shipped my order and it's now the Friday of the first week of Q3 so no shipping until Monday at least.
Nobody has confirmed that they have received a Neptune or even a shipping number from KfC since the Q2 deadline.
All signs point towards them having shipped all the chips they had yielded in Q2 and screw the rest of us.

PLEASE GIVE ME MY MONEY BACK KnC!

they learned that one from ScamFast  Roll Eyes
sr. member
Activity: 266
Merit: 250
Quick update:
KnC still haven't agreed refund or shipped my order and it's now the Friday of the first week of Q3 so no shipping until Monday at least.
Nobody has confirmed that they have received a Neptune or even a shipping number from KfC since the Q2 deadline.
All signs point towards them having shipped all the chips they had yielded in Q2 and screw the rest of us.

https://bitcointalksearch.org/topic/m.7619406

PLEASE GIVE ME MY MONEY BACK KnC! (requested as before you started shipping, as soon as you released specs)
 
legendary
Activity: 1848
Merit: 1001
Not true... Many of us mined around 50 coin per Jupiter... just so you know....

And it cost, what, 60 coins?   
... and Yifu Guo was no angel. There were a lucky few who got Avalons early, but make no mistake,
Avalon is just as guilty(or more) when it comes to orders which were delivered late.

Most of batch-1 that was a big winner. Batch-2 was a toss up. Everything after that was a disaster.

Jupiturds cost around 78!! meaning the ONLY way you could be happy with a 50 coin return is if....they were bought with STOLEN coins/coins bought with STOLEN bank details!!  (or you are retarded...)

Agree about Avalon Batch 1 - gutted i missed that by fear, uncertainty & doubt.

Batch 2 was pretty shit though all things considered - april delivery turned into mid august shipping etc...


I paid with PayPal.... Idiot; at a time I had a single coin I mined with GPU's.
and BTW, worked very hard to be able to get a few Saturns. I sold My fucking bad ass 2004 Subaru Outback, sold my entire comic book collection, and  several hundred orchids along with sacrificing "Good meals" for Top Ramen to be able to mine you Jackass. I endured every hardship there has been to deal with on those machines as well. They were shipped to the wrong place, had the wrong pool and password information programmed in, had bad URL's programmed into it, and had the "Dead Die problem to deal with, practically pioneering the wifi networking situation, discovering heat woke them up... lightning strikes (literally) and got slammed by both GOX and the CTSminer scam, and gave out about 5 grand in total reward money
You know what? I still turned a profit, so suck my chocolate salted balls and get a life.
I'd do it all again. I'm a die-hard tech junky addict that loves to tear things apart and make them better, but you wouldn't know about such things, because you are too busy having your head up your ass.
Besides, June 3rd, 2013, when ordering began, the price was 115 dollars, which makes a Jupiter 60 coin, not 78 again, idiocy
Your antics are futile.


damn phoenix...don't hold back...tell us how you really feel Smiley wow to still claim profit after all that is a very nice...now had you paid in btc, you'd be screw hehe...wish i had the money to pay in usd at the time and i'd be sitting pretty right now too...oh well what can ya do...gratz on the profit


Thieves, liars & scammers will ALWAYS find some obscure garbage to 'prove' themselves 'right' until they cant deny facts anymore, then they 'turn' nasty.

perhaps all that 'top' ramen has rotted your insignificant cranial matter...?

either way you remain a liar & a scammer

keep trying tho, eventually everyone who knows the truth will either die or get bored & leave you to it, then you can set your next scam/s in motion...

'Ow! My Orchids'...

member
Activity: 61
Merit: 10
Good news guys.  I have reduced my VRM temps by 20 degrees C with my new delta fan:



Heres a comparison with the other cubes that still have the stock fans:



Only problem is that it sounds like a Boeing 747 but i would rather be deaf than having burnt VRMs.

weird...the cooler one is higher amps...that doesn't make sense...usually the cooler the vrm the lower the amps...at least that is what i see on my jup units...oh well good job on lowering the temps Smiley
weird indeed.(more amps)    Any performance increase from that?
thx crash

Hard to tell at this point if it hashes faster now since ive only modded one cube.  I have 9 more delta fans, but my 120mm fan grills wont arrive until Saturday.  Once I install these fans on all my other cubes then ill be able to see if im getting any performance increase overall.  Its possible that this cube just consumes more amps than my other ones.  I know that this cube was the hottest one so I picked it as the guinea pig for the new fan.  Unfortunately I didn't pay attention the the amps before I installed the delta.  Guess we will know more in a few days.
legendary
Activity: 938
Merit: 1000
LIR DEV
Good news guys.  I have reduced my VRM temps by 20 degrees C with my new delta fan:



Heres a comparison with the other cubes that still have the stock fans:



Only problem is that it sounds like a Boeing 747 but i would rather be deaf than having burnt VRMs.

weird...the cooler one is higher amps...that doesn't make sense...usually the cooler the vrm the lower the amps...at least that is what i see on my jup units...oh well good job on lowering the temps Smiley
weird indeed.(more amps)    Any performance increase from that?
thx crash
hero member
Activity: 784
Merit: 504
Dream become broken often
Good news guys.  I have reduced my VRM temps by 20 degrees C with my new delta fan:



Heres a comparison with the other cubes that still have the stock fans:



Only problem is that it sounds like a Boeing 747 but i would rather be deaf than having burnt VRMs.

weird...the cooler one is higher amps...that doesn't make sense...usually the cooler the vrm the lower the amps...at least that is what i see on my jup units...oh well good job on lowering the temps Smiley
member
Activity: 61
Merit: 10
Good news guys.  I have reduced my VRM temps by 20 degrees C with my new delta fan:



Heres a comparison with the other cubes that still have the stock fans:



Only problem is that it sounds like a Boeing 747 but i would rather be deaf than having burnt VRMs.
hero member
Activity: 784
Merit: 504
Dream become broken often
Not true... Many of us mined around 50 coin per Jupiter... just so you know....

And it cost, what, 60 coins?    
... and Yifu Guo was no angel. There were a lucky few who got Avalons early, but make no mistake,
Avalon is just as guilty(or more) when it comes to orders which were delivered late.

Most of batch-1 that was a big winner. Batch-2 was a toss up. Everything after that was a disaster.

Jupiturds cost around 78!! meaning the ONLY way you could be happy with a 50 coin return is if....they were bought with STOLEN coins/coins bought with STOLEN bank details!!  (or you are retarded...)

Agree about Avalon Batch 1 - gutted i missed that by fear, uncertainty & doubt.

Batch 2 was pretty shit though all things considered - april delivery turned into mid august shipping etc...


I paid with PayPal.... Idiot; at a time I had a single coin I mined with GPU's.
and BTW, worked very hard to be able to get a few Saturns. I sold My fucking bad ass 2004 Subaru Outback, sold my entire comic book collection, and  several hundred orchids along with sacrificing "Good meals" for Top Ramen to be able to mine you Jackass. I endured every hardship there has been to deal with on those machines as well. They were shipped to the wrong place, had the wrong pool and password information programmed in, had bad URL's programmed into it, and had the "Dead Die problem to deal with, practically pioneering the wifi networking situation, discovering heat woke them up... lightning strikes (literally) and got slammed by both GOX and the CTSminer scam, and gave out about 5 grand in total reward money
You know what? I still turned a profit, so suck my chocolate salted balls and get a life.
I'd do it all again. I'm a die-hard tech junky addict that loves to tear things apart and make them better, but you wouldn't know about such things, because you are too busy having your head up your ass.
Besides, June 3rd, 2013, when ordering began, the price was 115 dollars, which makes a Jupiter 60 coin, not 78 again, idiocy
Your antics are futile.


damn phoenix...don't hold back...tell us how you really feel Smiley wow to still claim profit after all that is a very nice...now had you paid in btc, you'd be screw hehe...wish i had the money to pay in usd at the time and i'd be sitting pretty right now too...oh well what can ya do...gratz on the profit
hero member
Activity: 700
Merit: 504
Run a Bitcoin node.
I can see why jelin was so frustrated with the AX1200i. It won't even power one Neptune module plus controller... the Over Current Protection (OCP) kicks in and the PSU shuts down.


Just FYI,  the AX1200 (no i) works great,  if you have one of those around.  I think other folks have pushed the OCP way up on the iand had luck with that, also.
Thanks - I'll look for that.
hero member
Activity: 700
Merit: 504
Run a Bitcoin node.
I can see why jelin was so frustrated with the AX1200i. It won't even power one Neptune module plus controller... the Over Current Protection (OCP) kicks in and the PSU shuts down.

As annoying as this is, it is a safety feature as it would normally be unusual to draw that much current through a single PCIe cable. So i blame KnC for not putting two PCIe jacks on each module, rather than Corsair for their smart PSU. Bitmain put two PCIe jacks on each S3 blade, but it turns out only one is needed. That's a forward thinking design (in case the BM1382 28nm chip needed the power) - KnC take note.

Corsair link software lets you increase the OCP limit which fixes the problem... BUT the link cable is designed to plug into a motherboard USB pin header, not into a USB port (which makes sense as the PSU is designed to be inside a PC). Damn, I've only got a laptop.

I've got 2-to-1 PCIe cables ordered from Cablez, hopefully they will spread the Amps between two ports on the PSU and get around the OCP.

I'm also trying to source an EVGA 1300W superNOVA G2, but no luck yet finding one in New Zealand.

So my Neptune has been delivered, visually looks ok, but I can't power it on yet.
Does anyone have another PSU recommendation?  At this stage I can't source EVGA 1300W superNOVA G2 in New Zealand.
hero member
Activity: 784
Merit: 504
Dream become broken often
Yes!
The idea was if we get the big ass cooling tower, bar, and blanket out of the way, we may be able to cool those VRM's a bit better.... what ya think?

I agree.. Since they sheered the aluminum the blankets are going to vary based on the unit, some will work much better than others.. I would like to try replacing the current setup with the same type of bracket on the old Neptunes, possibly some small VRM heatsinks (with 48A the heat dissipation is pretty insane) and a higher CFM fan.. I suspect that will be the easiest option that provides plenty of heat dissipation..

I'm most interested in swapping out the VRMs for the 50A version, I suspect they can really be pushed if that jump is taken... While on the Nov we can drive the VRMs much higher they still don't hit the 48A hard shutdown, in the Neptune the stock config is within a few percent of 48A..
Ah yes.  The old Jupiter bar would be great, but that would also mean losing the box totally because the holes are on the wrong sides (in previous posts). No problem there really.
The ericcsons they are using in Neptune are auto-shutting down at 48a?... and pushed within a couple amps?. Damn.
Theres gotta be a cheaper way than buying 8 new vrms for every pcb....
Hmm...  well...  we will soon see exactly what those limits are if Tolip has anything to say about it...hehe

I'm not sure how hard the limit is enforced but per the datasheet for the VRM it cannot be raised past 48A and the VRM is supposed to shutdown at that number..

I hope it's soft enough that they allow maybe 50A before a hard kill but some testing will tell..

check out the thread pictures of your mining rigs...guy put a neppy board in that novac stuff...with volts -0,073 and 500mhz clock vrms were at max 52a-50a and a cool 60c Smiley but he said if he pushed farther it would shut down Sad so idk if much above 500mhz is possible with these vrms...good luck guys
legendary
Activity: 1428
Merit: 1000
https://www.bitworks.io
Yes!
The idea was if we get the big ass cooling tower, bar, and blanket out of the way, we may be able to cool those VRM's a bit better.... what ya think?

I agree.. Since they sheered the aluminum the blankets are going to vary based on the unit, some will work much better than others.. I would like to try replacing the current setup with the same type of bracket on the old Neptunes, possibly some small VRM heatsinks (with 48A the heat dissipation is pretty insane) and a higher CFM fan.. I suspect that will be the easiest option that provides plenty of heat dissipation..

I'm most interested in swapping out the VRMs for the 50A version, I suspect they can really be pushed if that jump is taken... While on the Nov we can drive the VRMs much higher they still don't hit the 48A hard shutdown, in the Neptune the stock config is within a few percent of 48A..
Ah yes.  The old Jupiter bar would be great, but that would also mean losing the box totally because the holes are on the wrong sides (in previous posts). No problem there really.
The ericcsons they are using in Neptune are auto-shutting down at 48a?... and pushed within a couple amps?. Damn.
Theres gotta be a cheaper way than buying 8 new vrms for every pcb....
Hmm...  well...  we will soon see exactly what those limits are if Tolip has anything to say about it...hehe

I'm not sure how hard the limit is enforced but per the datasheet for the VRM it cannot be raised past 48A and the VRM is supposed to shutdown at that number..

I hope it's soft enough that they allow maybe 50A before a hard kill but some testing will tell..
legendary
Activity: 938
Merit: 1000
LIR DEV
Yes!
The idea was if we get the big ass cooling tower, bar, and blanket out of the way, we may be able to cool those VRM's a bit better.... what ya think?

I agree.. Since they sheered the aluminum the blankets are going to vary based on the unit, some will work much better than others.. I would like to try replacing the current setup with the same type of bracket on the old Neptunes, possibly some small VRM heatsinks (with 48A the heat dissipation is pretty insane) and a higher CFM fan.. I suspect that will be the easiest option that provides plenty of heat dissipation..

I'm most interested in swapping out the VRMs for the 50A version, I suspect they can really be pushed if that jump is taken... While on the Nov we can drive the VRMs much higher they still don't hit the 48A hard shutdown, in the Neptune the stock config is within a few percent of 48A..
Ah yes.  The old Jupiter bar would be great, but that would also mean losing the box totally because the holes are on the wrong sides (in previous posts). No problem there really.
The ericcsons they are using in Neptune are auto-shutting down at 48a?... and pushed within a couple amps?. Damn.
Theres gotta be a cheaper way than buying 8 new vrms for every pcb....
Hmm...  well...  we will soon see exactly what those limits are if Tolip has anything to say about it...hehe
legendary
Activity: 1428
Merit: 1000
https://www.bitworks.io
Yes!
The idea was if we get the big ass cooling tower, bar, and blanket out of the way, we may be able to cool those VRM's a bit better.... what ya think?

I agree.. Since they sheered the aluminum the blankets are going to vary based on the unit, some will work much better than others.. I would like to try replacing the current setup with the same type of bracket on the old Neptunes, possibly some small VRM heatsinks (with 48A the heat dissipation is pretty insane) and a higher CFM fan.. I suspect that will be the easiest option that provides plenty of heat dissipation..

I'm most interested in swapping out the VRMs for the 50A version, I suspect they can really be pushed if that jump is taken... While on the Nov we can drive the VRMs much higher they still don't hit the 48A hard shutdown, in the Neptune the stock config is within a few percent of 48A..
legendary
Activity: 938
Merit: 1000
LIR DEV
Good to see people getting into the Neptunes just like the Jupiter..

Although I think watercooling is great the way the Neptune is put together and the heat one must dissipate means an i30 with a decent fan provides plenty of heat dissipation.. The VRM and temps on the bottom of the board are two of the most concerning and really the place one needs to focus to keep things under control.

I have built over 30 1Th/s 4 board Jupiters so far (6 boards all run 1.5Th/s) and raising the boards a little and swapping out the fan for a fairly low pressure high CFM fan (keeps the noise down with a low pressure fan) is enough to keep air moving quick enough to keep them where they need to be..

On the Neptune side I haven't put my thermal camera on the boards yet since I don't have one but based on the layout it's a similar problem..

The biggest challenge I've seen is the VRM shutdown, with a hardcoded 48A shutdown I don't know how to push past that but I have replaced VRMs on the Jupiters so I might do it on the Neptune and see what happens, if I get one at some point..

Yes!
The idea was if we get the big ass cooling tower, bar, and blanket out of the way, we may be able to cool those VRM's and the entire board a bit better.... what ya think?
legendary
Activity: 1428
Merit: 1000
https://www.bitworks.io
Good to see people getting into the Neptunes just like the Jupiter..

Although I think watercooling is great the way the Neptune is put together and the heat one must dissipate means an i30 with a decent fan provides plenty of heat dissipation.. The VRM and temps on the bottom of the board are two of the most concerning and really the place one needs to focus to keep things under control.

I have built over 30 1Th/s 4 board Jupiters so far (6 boards all run 1.5Th/s) and raising the boards a little and swapping out the fan for a fairly low pressure high CFM fan (keeps the noise down with a low pressure fan) is enough to keep air moving quick enough to keep them where they need to be..

On the Neptune side I haven't put my thermal camera on the boards yet since I don't have one but based on the layout it's a similar problem..

The biggest challenge I've seen is the VRM shutdown, with a hardcoded 48A shutdown I don't know how to push past that but I have replaced VRMs on the Jupiters so I might do it on the Neptune and see what happens, if I get one at some point..
legendary
Activity: 938
Merit: 1000
LIR DEV
Whilst on the subject of waterblocks.

My water loop is running Smiley
(not installed yet)
No leaks!
Might burp it once more but then it's time to take the plunge.

My goal is <15 minute pit stop per cube.


YMMV
Smiley

Pics of 4kW single rail, water cooled nep to follow in a few hours if all goes well.

You are the man! Can't wait to see what you're up to hehe
any firmware tweaks you recommend are always welcome too!
sr. member
Activity: 386
Merit: 250
Whilst on the subject of waterblocks.

My water loop is running Smiley
(not installed yet)
No leaks!
Might burp it once more but then it's time to take the plunge.

My goal is <15 minute pit stop per cube.


YMMV
Smiley

Pics of 4kW single rail, water cooled nep to follow in a few hours if all goes well.
legendary
Activity: 938
Merit: 1000
LIR DEV
It didnt take Elenelen long to ramp right up to 500mhz with simple vrm heatsinks. Imagine what we could do if that cooling tower, thermal blanket, and that stupid bar were all gone, and replaced with an H2O cooler. the vrms, and any heatsink you put on them would then have excellent exposure to airflow.

If someone can mount an Corsair H80 on the chip and mount a push/pull fan configuration right on the side of the box you can literally bring the size of that box down to a small shoe box size. I've been using that Corsair H80 on my overclocked Intel 3930K processor for a while now and works great. It is a completely sealed system and keeps the processor cool at all times. If there is a will, there would be a way. The ASIC chips are larger in size, so making it work would take a good DIY'er.



 Do you know how much heat the H80 can remove?  I think those ASICs are about three times the power draw of your 3930k... it probably wont be enough.

That depends on the speed of your fan, and ambient temps; but it can dissipate up to about 2500 watts according to a Corsair support specialist, but he couldn't give me solid figures although I think that's a lofty estimate we don't need nearly that.
theres the H-100 as well, but were only talking about a single asic here... about 150-200 watts max per asic (300 watts total draw for the entire cube according to KNC) found here: https://www.kncminer.com/news/news-96 The asic alone may be half that maximum(an estimation)
It can be done with H60's.     How many watts of you think that cooling tower is capable of? It's already pushing it's spec limit in warmer climates.
Surely even an H60 will be far superior to that bulky archaic arctic tower heatsink.

 Whoa! 2500 watts would suffice Wink

even if he had said 2 to 500   its better imho
legendary
Activity: 3808
Merit: 7912
It didnt take Elenelen long to ramp right up to 500mhz with simple vrm heatsinks. Imagine what we could do if that cooling tower, thermal blanket, and that stupid bar were all gone, and replaced with an H2O cooler. the vrms, and any heatsink you put on them would then have excellent exposure to airflow.

If someone can mount an Corsair H80 on the chip and mount a push/pull fan configuration right on the side of the box you can literally bring the size of that box down to a small shoe box size. I've been using that Corsair H80 on my overclocked Intel 3930K processor for a while now and works great. It is a completely sealed system and keeps the processor cool at all times. If there is a will, there would be a way. The ASIC chips are larger in size, so making it work would take a good DIY'er.



 Do you know how much heat the H80 can remove?  I think those ASICs are about three times the power draw of your 3930k... it probably wont be enough.

depends on the speed of your fan, and ambient temps; but it can dissapate up to about 2500 watts according to a Corsair support specialist, but he couldn't give me solid figures
theres the H-100 as well, but were only talking about a single asic here... about 400 watts max per asic(300 according to KNC)
it can be done with H60's.     How many watts of you think that cooling tower is capable of? It's already pushed it's spec limit.
Surely even an H60 will be superior to that bulky archaic arctic tower heatsink.

 Whoa! 2500 watts would suffice Wink
Jump to: