ANTMINER C1 Discussion and Support Thread

[tbolt]

I have not measured the core voltage directly with a DVM.

Are you sure you are looking in the correct location?
Do you have a schematic of the board or are you just using the ASIC device data sheet?

After you Modify voltage, you do a Save&Apply, then need to Power off and Restart the miner, right?

[notlist3d]

Did last attempt to make my spare C1 a full unit.   I pasted the back side of the board's.  It was a LOT of thermal paste even when I  made it spread out where it's a thin coat.   

Did 4 blades and at then end seemed like 1 works perfect on blade, others all depend.  So basically I have a parts unit for my other C1 if it ever craps out.   I will not be touching this one for a while.

[RichBC]

@tbolt Looking at your results I feel the Core voltage must be changing as I cannot see that anything else could give such a big change in power consumption.

My measurements all relate to S3 & S3+ which I think have similar hashing boards to the C2. There is BTW at least as much uncertainty in the S3 threads as to weather it is possible to reduce the core voltage with the software. I have the schematic for the S1, have not seen one for the S3 but they are similar. I am definitely measuring at the correct place, the output of the Buck Converter after the large inductor and even with the correct firmware, Bitmain disabled voltage adjustment in the latest, it does not change. Yes to Save & Apply and power off & restart.

Do you have a picture of the hash board in your C1 showing the buck converters to compare to the S3?

Thanks

[tbolt]

@tbolt Looking at your results I feel the Core voltage must be changing as I cannot see that anything else could give such a big change in power consumption.

My measurements all relate to S3 & S3+ which I think have similar hashing boards to the C2. There is BTW at least as much uncertainty in the S3 threads as to weather it is possible to reduce the core voltage with the software. I have the schematic for the S1, have not seen one for the S3 but they are similar. I am definitely measuring at the correct place, the output of the Buck Converter after the large inductor and even with the correct firmware, Bitmain disabled voltage adjustment in the latest, it does not change. Yes to Save & Apply and power off & restart.

Do you have a picture of the hash board in your C1 showing the buck converters to compare to the S3?

Thanks

@RichBC

These are the only images that I have of the C1 hash board.

[RichBC]

@tbolt  Thanks for the pictures of the C2 Hash Boards.  They show the familiar TPS53355 Buck Converters used in the S1, S3 & S3+ Miners. Sort of what I expected to see but the issue is that these set the voltage with a potential divider, as adjusted by the pencil mod etc, and without some additional circuitry of which there is not sign, they are not programmable by the software. 

I assume these are not pictures of your actual C1 so there is still a slim possibility that there is a different version of the hash board that crops up in some units? However for the moment I remain confused...


Rich

[notlist3d]

@tbolt  Thanks for the pictures of the C2 Hash Boards.  They show the familiar TPS53355 Buck Converters used in the S1, S3 & S3+ Miners. Sort of what I expected to see but the issue is that these set the voltage with a potential divider, as adjusted by the pencil mod etc, and without some additional circuitry of which there is not sign, they are not programmable by the software. 

I assume these are not pictures of your actual C1 so there is still a slim possibility that there is a different version of the hash board that crops up in some units? However for the moment I remain confused...


Rich

Those are stock photos.  But from spending a few hours re-pasting I can tell you that is what all 4 look like. 

But also add the water cooling blocks, and equipment.  They fit 4 blades in the case with using water cooling.

[dunand]

C2???

[notlist3d]

C2???

I'm assuming that part was a typo.  I don't think they will be making a C2.

But I could be wrong or out of the loop.

[RichBC]

Typo / wishful thinking.   Water cooled BM1384 based, 4 boards, 2TH/s, optimised for best efficiency at 0.3J/GH  

ich

[notlist3d]

Typo / wishful thinking.   Water cooled BM1384 based, 4 boards, 2TH/s, optimised for best efficiency at 0.3J/W  

ich

I honestly think syscooling killed it.  It was extra expense for water cooling vs air.   Was I neat idea thats why i got 2 of them.  One is still working, other is parts ATM.

The first pumps were horrible and so were the wire harness.  Again all syscooling not bitmain.  It's sad if they would have picked a different cooling company I think C1 would have had much better start, and possibly a second one in it's line.  What kind of water cooling company does not know do not dry test water pumps.

[RichBC]

Silly thing is as we move into an era, for those of us on normally priced electricity, where efficiency and J/KH is king, undervolting and underclocking is the order of the Day and you barely need a fan let alone water cooling.

Rich

[Sweminer777]

Typo / wishful thinking.   Water cooled BM1384 based, 4 boards, 2TH/s, optimised for best efficiency at 0.3J/W  

ich

I honestly think syscooling killed it.  It was extra expense for water cooling vs air.   Was I neat idea thats why i got 2 of them.  One is still working, other is parts ATM.

The first pumps were horrible and so were the wire harness.  Again all syscooling not bitmain.  It's sad if they would have picked a different cooling company I think C1 would have had much better start, and possibly a second one in it's line.  What kind of water cooling company does not know do not dry test water pumps.

My radiator whent out, had to purchase a EK 360XT, pump seems to dying any time soon, got a replacement aswell Phobya, but yes Syscooling has horrrible water cooling.

I did not want to make any invesments but i did now that the syscooling system wouldnt last atleast now i have my running overclocked at 50c on a closed envirment, pulling 1,1thz., still waiting for winter might put the machine on the balconi it gets -20c here for 7 months will see how it pulls trough

[dogie]

@tbolt Looking at your results I feel the Core voltage must be changing as I cannot see that anything else could give such a big change in power consumption.

My measurements all relate to S3 & S3+ which I think have similar hashing boards to the C2. There is BTW at least as much uncertainty in the S3 threads as to weather it is possible to reduce the core voltage with the software. I have the schematic for the S1, have not seen one for the S3 but they are similar. I am definitely measuring at the correct place, the output of the Buck Converter after the large inductor and even with the correct firmware, Bitmain disabled voltage adjustment in the latest, it does not change. Yes to Save & Apply and power off & restart.

Do you have a picture of the hash board in your C1 showing the buck converters to compare to the S3?

Thanks

@RichBC

These are the only images that I have of the C1 hash board.

https://bitcointalk.org/index.php?topic=846095 for PCB pictures of the original version.

[Ecnad]

Running my C1 at 150M Frequency and 0680 Voltage for the summer. Getting 608GHs average and temps steady at 32.

Is there any more efficient but cool settings folks have been using?

[mavericklm]

using mains water! i can get under 30c the temps.
atm i got 54c, i'll go take a shower and lower the temps to ~45...

[lovenlifelarge]

Setting 1

FREQ : 256.25

VOLT : 0665

SPEED = ?GH @ WATT = ?W

------------------------------------------------------------------------------------------------------------------
Setting 2

FREQ : 268

VOLT : 0680

SPEED = ?GH @ WATT = ?W

------------------------------------------------------------------------------------------------------------------
Setting 3

FREQ : 200

VOLT : 0675

SPEED = 803GH @ WATT = 530W
-----------------------------------------------------------------------------------------------------------------
Setting 4

FREQ : 250

VOLT : 0760

SPEED = 1008GH @ VOLT = 790W

-------------------------------------------------------------------------------------------------------------------

Can someone with a kill o watt please help me fill in the missing blanks.

Can anyone else please confirm any if not all settings above work for u & are stable??

If settings are not stable please let me know so this table can be adjusted...

& votes for best gh per watt that are stable for u so that others can use this as a guide...

[BlInK311]

Running my C1 at 150M Frequency and 0680 Voltage for the summer. Getting 608GHs average and temps steady at 32.

Is there any more efficient but cool settings folks have been using?

How many watts do these settings pull from the wall?  My summer settings have been 200M freq 0680v giving me 801GH at 535watts (~0.68w/GH).

[lovenlifelarge]

Running my C1 at 150M Frequency and 0680 Voltage for the summer. Getting 608GHs average and temps steady at 32.

Is there any more efficient but cool settings folks have been using?

How many watts do these settings pull from the wall?  My summer settings have been 200M freq 0680v giving me 801GH at 535watts (~0.68w/GH).

I know im knit picking here but u can drop the volt down another 5 to 0675...

Every bit counts...

[RichBC]

I know im knit picking here but u can drop the volt down another 5 to 0675...

Every bit counts...

From the work I have done undervolting an S3, Same BM1382 Chip as the C1. I have gone down to 0.67V (0670) with a small HW error rate of .0048%, so you should be fine at 0675

Rich

[BlInK311]

I know im knit picking here but u can drop the volt down another 5 to 0675...

Every bit counts...

From the work I have done undervolting an S3, Same BM1382 Chip as the C1. I have gone down to 0.67V (0670) with a small HW error rate of .0048%, so you should be fine at 0675

Rich

Thanks for the suggestion.  turning down to 0675 saved a few watts. 
I also tested dropping freq down to 150M/0675V instead of 200M/0675V and it actually gave me worse efficiency.  settings back to 200M/0675V