Bitmain's Released Antminer S9, World's First 16nm Miner Ready to Order

[HagssFIN]

please help ASAP
I replaced the controler on S9 which I bought with bitmain. The hashing boards are working I checked them on another S9. When I powered it up it shows the hashing boards in status, first minute or so there are all 0s but also zero hashing speed, after couple minutes all of them are showing xxx. I tried 600 and 500 freq but still the same. What am I missing ?

update:
is there a chance that the miner isn't working with a faulty fan? even if I disconnect it?

Yes. Replace the faulty fan.

shipping from bitmain costs too much, any alternatives from ebay/amazon?

This is a good one, I used to have AvalonMiner 6.0 equipped with this.
Sanyo Denki SAN Ace 120 3.0 Amp 12 Volt 120mm PWM Fan 224 CFM
https://www.amazon.com/Sanyo-Denki-SAN-Ace-120mm/dp/B00QWB1X3O/ref=sr_1_1?ie=UTF8&qid=1488887739&sr=8-1&keywords=san+ace+120

This will git s9 I guess but isn't the cable too short for rare fan ?

There are different size of extension cables available
for example https://www.amazon.com/extension-cable-Black-Sleeving-FC44PWM-12BKS/dp/B0055OLY88

take measurements and see what fits the best

Thank you for your help, you saved me so much time!

No problem, happy to help  

If you have few miners I would give advice to have at least one spare fan all the time.
That way there is only little down time because of faulty fans.  

And this is of course for the 120mm fans. I have those but no R4 fan because that is god damn hard to find  

[1714936161]

1714936161

[1714936161]

1714936161

[numnutz2009]

before you go replacing the fan with a different model you might want to check the power the current fan uses and how many watts the replacement uses. if the replacement uses too much ur gonna damage the controller. it may work for a bit of time but the extra wear on components will take its toll eventually and they will fail. are you willing to take that chance? that fan you posted uses almost 40 watts of power....im not sure what the stock fan is rated at.

[mettalmag]

before you go replacing the fan with a different model you might want to check the power the current fan uses and how many watts the replacement uses. if the replacement uses too much ur gonna damage the controller. it may work for a bit of time but the extra wear on components will take its toll eventually and they will fail. are you willing to take that chance? that fan you posted uses almost 40 watts of power....im not sure what the stock fan is rated at.

anyone has any info how much does the stock fan use ?
I will able to check at place after couple days only

[VRobb]

The stock Delta 12038 6000rpm fan pulls around 36W when at full speed.

[mettalmag]

The stock Delta 12038 6000rpm fan pulls around 36W when at full speed.

I actually never seen more than 5800rpm even if on static 100% speed, so I think even +4w is ok

[HagssFIN]

before you go replacing the fan with a different model you might want to check the power the current fan uses and how many watts the replacement uses. if the replacement uses too much ur gonna damage the controller. it may work for a bit of time but the extra wear on components will take its toll eventually and they will fail. are you willing to take that chance? that fan you posted uses almost 40 watts of power....im not sure what the stock fan is rated at.

Don't worry, I know what I'm doing.

And as said the current is about the same as with the stock fan.

[solteykr]

before you go replacing the fan with a different model you might want to check the power the current fan uses and how many watts the replacement uses. if the replacement uses too much ur gonna damage the controller. it may work for a bit of time but the extra wear on components will take its toll eventually and they will fail. are you willing to take that chance? that fan you posted uses almost 40 watts of power....im not sure what the stock fan is rated at.

The actual model of the fan is 'Delta QFR1212GHE" Its DC12v @ 2.70Ampere. This is the PUSH fan. located in front with your LAN cable. I have a broken S9 that's why I saw it's model inside.

Isn't it better to use a bigger Ampere fan so that its lifetime is stretched longer? I mean the controller gives the same volt and amp when powering the fan, meaning the 3amp fan can handle a full power from the controller and still has extra .30ampere (3amp minus 2.70amp). Am I right? I also want to confirm this... Anyone with electronics knowledge please. I want to buy a aluminum framed 12v 4amp fan it will really look bad ass. If i can use it..

Here is my fan FAN DATASHEET

[NotFuzzyWarm]

sn't it better to use a bigger Ampere fan so that its lifetime is stretched longer? I mean the controller gives the same amp when powering the fan,

Interesting reasoning there.....

How much power a fan draws has nothing to do with its lifetime. A fan drawing more power either means that it is moving more air against high static pressure or it has crappy motor/bearing.

The controller is supplying the same voltage to the fan and the fan will then draw much current as it needs. Question is, how much current can the circuit traces running between the PCIe plug and fan connector handle? Ans: not much. They are thin and anything over 4 amps would be really pushing it.

Now since the fans are 4-wire PWM all the power functions related to the fan spinning at x speed are inside the fan, all the controller itself provides is a low power pulse train (on the blue wire) telling how fast the fan needs to spin so those +12v and return traces (red & black wires on fan) should be the limiter.

[solteykr]

sn't it better to use a bigger Ampere fan so that its lifetime is stretched longer? I mean the controller gives the same amp when powering the fan,

Interesting reasoning there.....

How much power a fan draws has nothing to do with its lifetime. A fan drawing more power either means that it is moving more air against high static pressure or it has crappy motor/bearing.

The controller is supplying the same voltage to the fan and the fan will then draw much current as it needs. Question is, how much current can the circuit traces running between the PCIe plug and fan connector handle? Ans: not much. They are thin and anything over 4 amps would be really pushing it.

Now since the fans are 4-wire PWM all the power functions related to the fan spinning at x speed are inside the fan, all the controller itself provides is a low power pulse train (on the blue wire) telling how fast the fan needs to spin so those +12v and return traces (red & black wires on fan) should be the limiter.

yet another great info. Thank you sir.

Question sir, if example I used the 4amp fan. does the fan "ASK" and "DRAW" or "PULL" (DEMAND) the 4amp from the controller?
or the controller still controls the voltage and amp?

Thanks again for your reply.

[NotFuzzyWarm]

<snip>

Question sir, if example I used the 4amp fan. does the fan "ASK" and "DRAW" or "PULL" (DEMAND) the 4amp from the controller?
or the controller still controls the voltage and amp?
Thanks again for your reply.

The power the fan uses does not come from the controller itself. The controller does not directly control voltage/amps. In other words, no active devices switching power are involved.

The power DOES flow through traces (wiring) on the controller board from the PCIe plug to the red and black wire connection to the fans. Those are what may fry if overloaded. In theory there is no reason one can't directly wire the red and black fan wires to a Molex or whatever to bypass those thin connections.

As I said ^^ with 4-wire PWM fans all the controller provides is a speed command to the fan via what is usually a blue wire. Circuitry in the fan reads that signal and internally handles all 12v power switching to the fan motor. Same circuits also provide a speed feedback signal (yellow wire) back to the controller to tell it how fast the fan is running.

Only with a 3-wire (power and feedback only) fan is a controller actually having to handle power loads.

[solteykr]

<snip>

Question sir, if example I used the 4amp fan. does the fan "ASK" and "DRAW" or "PULL" (DEMAND) the 4amp from the controller?
or the controller still controls the voltage and amp?
Thanks again for your reply.

The power the fan uses does not come from the controller itself. The controller does not directly control voltage/amps. In other words, no active devices switching power are involved.

The power DOES flow through traces (wiring) on the controller board from the PCIe plug to the red and black wire connection to the fans. Those are what may fry if overloaded. In theory there is no reason one can't directly wire the red and black fan wires to a Molex or whatever to bypass those thin connections.

As I said ^^ with 4-wire PWM fans all the controller provides is a speed command to the fan via what is usually a blue wire. Circuitry in the fan reads that signal and internally handles all 12v power switching to the fan motor. Same circuits also provide a speed feedback signal (yellow wire) back to the controller to tell it how fast the fan is running.

Only with a 3-wire (power and feedback only) fan is a controller actually having to handle power loads.

So I can still use the 4amp fan if I really want to. just need to power it via other source or directly to the PSU. and still the controller has the ability to adjust the speed. hehe.. You are awesome sir!

[bitsink]

before you go replacing the fan with a different model you might want to check the power the current fan uses and how many watts the replacement uses. if the replacement uses too much ur gonna damage the controller. it may work for a bit of time but the extra wear on components will take its toll eventually and they will fail. are you willing to take that chance? that fan you posted uses almost 40 watts of power....im not sure what the stock fan is rated at.

The actual model of the fan is 'Delta QFR1212GHE" Its DC12v @ 2.70Ampere. This is the PUSH fan. located in front with your LAN cable. I have a broken S9 that's why I saw it's model inside.

Isn't it better to use a bigger Ampere fan so that its lifetime is stretched longer? I mean the controller gives the same volt and amp when powering the fan, meaning the 3amp fan can handle a full power from the controller and still has extra .30ampere (3amp minus 2.70amp). Am I right? I also want to confirm this... Anyone with electronics knowledge please. I want to buy a aluminum framed 12v 4amp fan it will really look bad ass. If i can use it..

Here is my fan FAN DATASHEET

Normally a consumer "pulls" the amount of current it "needs" (given its internal resistance), though there are exceptions, where a supply will only "give" a max current (usually by changing the voltage). V=IR (for a supply V is usually fixed and the R is given by the applyance, then I is derived from the combo).

[numnutz2009]

before you go replacing the fan with a different model you might want to check the power the current fan uses and how many watts the replacement uses. if the replacement uses too much ur gonna damage the controller. it may work for a bit of time but the extra wear on components will take its toll eventually and they will fail. are you willing to take that chance? that fan you posted uses almost 40 watts of power....im not sure what the stock fan is rated at.

The actual model of the fan is 'Delta QFR1212GHE" Its DC12v @ 2.70Ampere. This is the PUSH fan. located in front with your LAN cable. I have a broken S9 that's why I saw it's model inside.

Isn't it better to use a bigger Ampere fan so that its lifetime is stretched longer? I mean the controller gives the same volt and amp when powering the fan, meaning the 3amp fan can handle a full power from the controller and still has extra .30ampere (3amp minus 2.70amp). Am I right? I also want to confirm this... Anyone with electronics knowledge please. I want to buy a aluminum framed 12v 4amp fan it will really look bad ass. If i can use it..

Here is my fan FAN DATASHEET

Normally a consumer "pulls" the amount of current it "needs" (given its internal resistance), though there are exceptions, where a supply will only "give" a max current (usually by changing the voltage). V=IR (for a supply V is usually fixed and the R is given by the applyance, then I is derived from the combo).

these are PWM fans so the voltage will stay the same at 12v plus or minus 10%. 3 pin fans are controlled via the voltage being pushed to the fan. if the controller can handle 3a max that doesn't mean its using 3a all the time. the usage depends on the fans speed and how efficient that fan is at moving the high volume of air through the limited and restricted space...basically like the guy above said the static pressure. pushing air through a hollow tube is much easier than pushing it through a tube with obstructions which is what the s9 is. You should also note that just because the fan spins faster or has a higher CFM doesnt mean its a better fan. it can use more power due to an inefficient design and also if the fan isnt designed for a high static pressure it can move all the air in the world and it still wouldnt be able to work with the s9's design. So there is alot more to fan swap outs than just voltage amperage and cfm.

NotFuzzyWarm  - yessir...i dont think the tiny traces can handle much more power than they are handling now but honestly i wouldnt want to test it anyway. chances are the traces are just the right size for the fans being used now and since mining is a numbers game im positive bitmain is trying to cut costs wherever possible even if that means making traces less able to handle bigger loads. smaller traces + less material = some savings that add up based on the number of boards they have produced. you can do what you mentioned and unhook the red and black wire and power them on directly from the 12v supply. You can also control the fan speed of ann the s9's if you unhook the blue wire and connect that to an external pwm controller. it has to be a 4 pin controller though not a 3 pin and it can only be the blue wire...the other wires stay in the molex connected to the controller. I posted this info in a previous post but if you think its a good idea so you can run the miner in your living room think again....it raises the temps so you need super cooling to run them at lower speeds and on new models you cant control the freq at all so its really only usable on pre auto tune firmware unless you have beastly cooling setup.

[VTdude12]

Bitcointalk world!  Need some help understanding my burning question about S9 temperatures (pun intended).  Done some exploring on the board and only found pieces on info here & there.  Any info / insight would be greatly appreciated.

I have built and air cooled box for my 3 S9s.  Going to be doing some minor tweaks to move evenly distribute cool air, but here are my current miner stats:

                Chip1    Chip2
Chain 6 -      73        85
Chain 7 -      81        89
Chain 8 -      78        92

                Chip1    Chip2
Chain 6 -      79        86
Chain 7 -      76        84
Chain 8 -      76        84

                Chip1    Chip2
Chain 6 -      67        84
Chain 7 -      64        73
Chain 8 -      61        75

Should I be worried about the temps on miners 1 & 2?

Also, when looking at the kernel log it seems that 2 of my miners have more than one temp monitor.  Looks like they are on chip62 and chip32 - what should I be looking for here.  In the case of miner #1 above, it looks like what is displayed on the miner status page is chip62.  And temps on chip62 always seem to be higher than chip32.

Kernel log from miner #1

do read temp on Chain[5]
Chain[5] Chip[62] TempTypeID=55 middle offset=24
Chain[5] Chip[62] local Temp=76
Chain[5] Chip[62] middle Temp=85
Chain[5] Chip[32] TempTypeID=55 middle offset=23
Chain[5] Chip[32] local Temp=58
Chain[5] Chip[32] middle Temp=73
Done read temp on Chain[5]
do read temp on Chain[6]
Chain[6] Chip[62] TempTypeID=55 middle offset=26
Chain[6] Chip[62] local Temp=78
Chain[6] Chip[62] middle Temp=89
Chain[6] Chip[32] TempTypeID=55 middle offset=24
Chain[6] Chip[32] local Temp=60
Chain[6] Chip[32] middle Temp=81
Done read temp on Chain[6]
do read temp on Chain[7]
Chain[7] Chip[62] TempTypeID=55 middle offset=26
Chain[7] Chip[62] local Temp=80
Chain[7] Chip[62] middle Temp=93
Chain[7] Chip[32] TempTypeID=55 middle offset=24
Chain[7] Chip[32] local Temp=56
Chain[7] Chip[32] middle Temp=77
Done read temp on Chain[7]

Thanks!!!

[bitsink]

I am interested in someone.
for sale!!
Antminer S9 16nm Asic Bitcoin Miner (14.0 TH/s)
A total of 3 units
AUCTION 


Tomasminer
Regards

Are you selling or buying?

Scammer???

[bitsink]

Those are in ok limits but your ambient temp is freakin' high! 

Underclocking and/or Undervolting may help. Has anybody worked out how to achieve this with the "autotune" firmware? It seems that every chip gets its own custom frequency setting now. If the lowest frequency from the search file is used per board that would be a good start?

Bump... Has anyone done some experimentation with underclocking these latest batch of miners?

[NotFuzzyWarm]

Those are in ok limits but your ambient temp is freakin' high! 

Underclocking and/or Undervolting may help. Has anybody worked out how to achieve this with the "autotune" firmware? It seems that every chip gets its own custom frequency setting now. If the lowest frequency from the search file is used per board that would be a good start?

Bump... Has anyone done some experimentation with underclocking these latest batch of miners?

Unless you are able to completely rewrite the Firmware to bypass all of the checking performed it cannot be done. Period.

[bitsink]

Those are in ok limits but your ambient temp is freakin' high! 

Underclocking and/or Undervolting may help. Has anybody worked out how to achieve this with the "autotune" firmware? It seems that every chip gets its own custom frequency setting now. If the lowest frequency from the search file is used per board that would be a good start?

Bump... Has anyone done some experimentation with underclocking these latest batch of miners?

Unless you are able to completely rewrite the Firmware to bypass all of the checking performed it cannot be done. Period.

The 1h+ "checking" is only done in the beginning after a hard reset, then this info is stored somewhere and used to initialize the miner. If this info can be changed it should be possible the reduce the frequency and underclock the chips. Has really nobody had a go at this yet?

[bitsink]

I have a problem where the bmminer is not starting automatically.

/tmp/search contains:

Code:

waiting for receive_func to exit!
waiting for pic heart to exit!
Start bmminer ...

but bmminer does not start. When I start bmminer manually it starts hashing away and all is well.

Code:

/usr/bin/bmminer --version-file /usr/bin/compile_time --api-listen --default-config /config/bmminer.conf

Any ideas what could be the problem?

[NotFuzzyWarm]

The 1h+ "checking" is only done in the beginning after a hard reset, then this info is stored somewhere and used to initialize the miner. If this info can be changed it should be possible the reduce the frequency and underclock the chips. Has really nobody had a go at this yet?

Not that I know of.

If you can, try and capture the boot sequence.

A couple years back a Dr of CompSci friend of mine looked at I think s5 or s7 code. Firmware and a boot image are stored in NV memory that is part of the  RasPi/BB/ and I assume Xylinx CPU (the chip is actually 3D with both DRAM and NV on bottom layer).

The boot loader lstarts from there and makes a VDISK of the image in memory and that is what the miner runs off of. Of course is also where the Factory reset image is kept. Is why editing conf files do not stick through a power cycle - the VDISK and your changes goes bye-bye.

Betcha Bitmain still does that. Question is, how to access that normally hidden part of storage.