Bitcoin Forum
April 23, 2018, 12:36:07 PM *
News: Latest stable version of Bitcoin Core: 0.16.0  [Torrent]. (New!)
 
   Home   Help Search Donate Login Register  
Pages: 1 2 [All]
  Print  
Author Topic: Antminer cooler in a box - Antminer s9 bitcoin miner cooler box DIY 2018  (Read 744 times)
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 23, 2018, 03:17:04 AM
Merited by frodocooper (2), OgNasty (1), vapourminer (1)
 #1

Hey guys,

I wanted to create a clean silencer box like you have seen on youtube but I didnt see any good step by step videos, so I made one!


https://www.youtube.com/watch?v=HhNS2ElN5Mk




Let me know if you guys have any suggestions about making a better box?



Or even how to do it for multiple units!?





Tell me what you guys think!













In the DIY antminer s9 cooler box and silencer video will help you guys silence and cool your own antminers.

In this antminer setup we will talk about the tools needed to make the antminer cooler box and a step by step method of how to construct an antminer cooling box.

6 inch diameter hole drill to cut the air input and air output ducting for the antminer cooler.
1.25 inch diameter hole drill to cut the psu and ethernet cable hole on the side of the antminer s9 cooler.

Get as much 6" INSULATED ducting as needed to cool your antminer s9 bitcoin miner.

48 QT cooler to put the antminer inside. The antminer s9 should have more than enough room in this size cooler.

After cutting the holes on the antminer cooler, you can sand down the insides of the holes to give the ducting a smoother surface to fit through on the antminer cooler.

MAKE SURE TO WIPE DOWN YOUR NEW ANTMINER S9 COOLER after you make the holes to remove an excess dust or particles from cutting the holes. You do not want that going into your brand new antminer s9.

You can also use a fan to blow away any excess debris left on the cooler, use whatever you like just make sure you clean the cooler for the antminer well!

After you have finished cutting all the holes for the cooler, you can get a piece of styrofoam to put the antminer s9 bitcoin miner to rest on. The reason we do this is to align the antminer perfectly in the center to connect the hoses to each side of your antminer.

I am using a APW3++ power supply unit from bitmain but I have seen others using the EVGA power supply for this setup as well.

You guys can hear in the video that clearly the sound from the antminer is silenced. Its does not go away but it does tone down the noise level of these antminers. I put a decibel meter next to the antminer so you can see the actual readings of the antminers decibel.

 

I hope this help you guys with you diy cooling hacks for the antminer.
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1524486967
Hero Member
*
Offline Offline

Posts: 1524486967

View Profile Personal Message (Offline)

Ignore
1524486967
Reply with quote  #2

1524486967
Report to moderator
1524486967
Hero Member
*
Offline Offline

Posts: 1524486967

View Profile Personal Message (Offline)

Ignore
1524486967
Reply with quote  #2

1524486967
Report to moderator
Druiz287
Jr. Member
*
Offline Offline

Activity: 61
Merit: 8


View Profile
January 23, 2018, 04:17:34 AM
 #2

Hey!  First, thank you for sharing your techniques with us.  The data you provided on wind, sound, and temperature levels was great.  Do you know what temperature your Antminer is reading by logging in?  I have a silencer box thing going on but its running pretty hot for what it used to run at.
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 23, 2018, 11:46:34 AM
 #3

You are very welcome! That is why I am doing this is to spread the knowledge. Yes when I logged into the GUI my readings dropped about 2 degrees which is not bad when silencing this sucker!!!!

If you have any other ideas let me know, lets get more data on these!
ThugB33
Jr. Member
*
Offline Offline

Activity: 46
Merit: 0


View Profile
January 23, 2018, 01:34:47 PM
 #4

I've made it much cheaper... Ikea 5$ table ... tuned it upside down. Made the walls and top cover from of the sound dumping and isolating material, cut the holes ... use the duct tape or gorilla tape, remember to use only aluminium stretching duct pipes and just connect 125mm duct silencer straight to miners intake and outtake and v'uola ... cost around 90$ for the whole set.... there you go 35db at full power! 
CrypotoCreanis
Member
**
Offline Offline

Activity: 129
Merit: 10

Don’t do a screen name after 4 glasses of Chianti


View Profile
January 24, 2018, 02:35:15 AM
 #5

I've made it much cheaper... Ikea 5$ table ... tuned it upside down. Made the walls and top cover from of the sound dumping and isolating material, cut the holes ... use the duct tape or gorilla tape, remember to use only aluminium stretching duct pipes and just connect 125mm duct silencer straight to miners intake and outtake and v'uola ... cost around 90$ for the whole set.... there you go 35db at full power! 

pics?

I am working on a box 24"x24" by 16" high that will fit 3
Druiz287
Jr. Member
*
Offline Offline

Activity: 61
Merit: 8


View Profile
January 24, 2018, 03:46:41 AM
 #6

That is a great design idea!  I think its awesome that we are all trying to solve this problem of sound while maintaining the temperature and air flow at an acceptable level.  I would love to see pictures of the IKEA table design as well.
HoleShot
Member
**
Offline Offline

Activity: 70
Merit: 28


View Profile
January 24, 2018, 09:47:52 PM
 #7

I've made it much cheaper... Ikea 5$ table ... tuned it upside down. Made the walls and top cover from of the sound dumping and isolating material, cut the holes ... use the duct tape or gorilla tape, remember to use only aluminium stretching duct pipes and just connect 125mm duct silencer straight to miners intake and outtake and v'uola ... cost around 90$ for the whole set.... there you go 35db at full power! 

Very nice. Please share pics
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 25, 2018, 05:40:51 AM
 #8

Thank you! Yes lets see the pics of the IKEA design!

Remember the more information we provide to each other to help solve the problems  the quicker and better job we will do. We are like a collective mining pool!


Anymore suggestions?Huh? And cant wait to see your IKEA design!
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 25, 2018, 06:41:05 AM
 #9

Buy the Cooler directly? From where?

I agree 100% I get headaches as well from these damn things.
ThugB33
Jr. Member
*
Offline Offline

Activity: 46
Merit: 0


View Profile
January 25, 2018, 11:52:14 AM
 #10

Guys,

I hope this helps!





I didn't waste so much efforts on this as the devices come and go... but I have also one another project in mind mind ... I will post more soon. Also If you can I suggest to use those fans... they are real "blowers"

https://www.ebay.com/itm/PC-Cooling-Fan-12cm-12V-DC-2-4A-200CFM-Brushless-Motor-4Pin-Connector-FFC1212DE/142096267077?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m2749.l2649

 

jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 25, 2018, 02:15:58 PM
 #11

ThugB,

GET THAT PSU UNIT OUT OF YOUR BOXX!!!!!!!!!

I honestly think that IS NOT the way you should be cooling that thing.


 1. Never put your PSU in the same box
 2. your insulation is leaky
 3. MAKE A VENT FOR YOUR PSU!!!


What are the chip temps ThugB when running this thing?


love the input hopefully my suggestions can help.
ThugB33
Jr. Member
*
Offline Offline

Activity: 46
Merit: 0


View Profile
January 25, 2018, 02:31:27 PM
 #12

ThugB,

GET THAT PSU UNIT OUT OF YOUR BOXX!!!!!!!!!

I honestly think that IS NOT the way you should be cooling that thing.


 1. Never put your PSU in the same box
 2. your insulation is leaky
 3. MAKE A VENT FOR YOUR PSU!!!


What are the chip temps ThugB when running this thing?


love the input hopefully my suggestions can help.

Yo, No worries!

There's a second white duct taking care of that as the air inside the box is 21°C. That's why I have a BT amp reader in there which is connected to a smart e-socket so it will shut off power if temps goes over 40°C.

The temps really depend on the season. At summer time chips get around 82°C and on winter time it's around 69°C and that's with new control board. With old control board the temps were around 7-9 degrees warmer (Bitmain's quaility Sad ).

jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 25, 2018, 02:35:32 PM
 #13

Gotcha Nice!!!

ThugB you should think of a way to scale it! Keep up the good work.
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 25, 2018, 03:02:41 PM
 #14

ThugB, I was looking at your images again and can I suggest that you change your ducting to insulated ducting so you don't loose that heat in the transfer process.
Bennyb1
Newbie
*
Offline Offline

Activity: 13
Merit: 0


View Profile
January 25, 2018, 06:35:02 PM
 #15

Hey thanks for sharing, this looks like a great cost effective way to muffle a miner.

My question, and this is for any who has experience with miners, is static electricity a problem with a plastic cooler? Or styrofoam? Using a cooler without the plastic shell would be cheaper but would expose the miner to all that styrofoam.

It seems that it could be a problem with an enclosed space made of a material that collects static. I would hope not and that anyone who chimes in can attest to running a miner in a cooler for months with no issue. Then again caution is a good thing considering how much these things cost.
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 25, 2018, 06:44:25 PM
 #16

That is a great thought Benny, I want to do more research into that topic.
ThugB33
Jr. Member
*
Offline Offline

Activity: 46
Merit: 0


View Profile
January 25, 2018, 07:36:58 PM
 #17

ThugB, I was looking at your images again and can I suggest that you change your ducting to insulated ducting so you don't loose that heat in the transfer process.


Yo,

I've tested all kind of ducts as I've been growing some stuff  Roll Eyes Over the years I've come to a result that alu-duc is the best for the moving air and as this unit need a lot of air I've came to a conclusion that this is a good option. But in your option you could also use the fanduct adapters on the miner and drill a bigger hole for a dumpers installation (the longer the dumpers the quieter the sound, the dumper must be at least 125mm wide) In that way you could make it much more quieter I think.

I've jut made a project on dumping two A3's Cheesy that's double "barrel" shotgun as I call it Cheesy

jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 25, 2018, 09:01:56 PM
 #18

Great info bro! That is exactly what I am doing for my exhausting system. I am going from the 5 inch, to 6 inch to 12 inch ducting hooked up to a 12 inch inline fan.

Do you know the CFM on that alu-duc?
Sandal_Hat
Full Member
***
Offline Offline

Activity: 308
Merit: 112


View Profile
January 25, 2018, 11:46:49 PM
 #19

It is still noisy and your temps are bound to be higher in that box. U should run them as cool as possible to ensure they last. When u put a few of them together, the sound will get stronger on its own. Alot of it are low to medium frequency sounds which are harder to block and u can hesr them from some distance. U actually need soft 10-13cm of soft 100-200kg density material to really absorb the sound.

Selling at 20% of coupon price aka 80% off.
Selling 2 bitmain 125 USD coupons, 18 bitmain 100 USD coupon, 1 bitmain 50 USD coupon

PM me
Raymond_B
Member
**
Offline Offline

Activity: 126
Merit: 21

Just Getting Started...


View Profile WWW
January 26, 2018, 12:03:02 AM
 #20

Hey thanks for sharing, this looks like a great cost effective way to muffle a miner.

My question, and this is for any who has experience with miners, is static electricity a problem with a plastic cooler? Or styrofoam? Using a cooler without the plastic shell would be cheaper but would expose the miner to all that styrofoam.

It seems that it could be a problem with an enclosed space made of a material that collects static. I would hope not and that anyone who chimes in can attest to running a miner in a cooler for months with no issue. Then again caution is a good thing considering how much these things cost.

This would be humidity dependent too. If you run them in a really low humidity environment then some grounding would be in order. And just like hardware techs, if you're worried about static electricity you should always wear a proper grounding strap when handling the equipment.

Friends don't let friends use 110V
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 26, 2018, 03:06:15 AM
 #21

I grounded every rack that touch any of my units, I agree that is the only way to go.
HoleShot
Member
**
Offline Offline

Activity: 70
Merit: 28


View Profile
January 26, 2018, 05:08:44 AM
Merited by frodocooper (1)
 #22

Hey thanks for sharing, this looks like a great cost effective way to muffle a miner.

My question, and this is for any who has experience with miners, is static electricity a problem with a plastic cooler? Or styrofoam? Using a cooler without the plastic shell would be cheaper but would expose the miner to all that styrofoam.

It seems that it could be a problem with an enclosed space made of a material that collects static. I would hope not and that anyone who chimes in can attest to running a miner in a cooler for months with no issue. Then again caution is a good thing considering how much these things cost.

I have not tried that but...

I do not think the miners are sensitive to ESD, electrostatic discharge. ESD boards are normally red in color, these are green. If they were ESD sensitive, Bitmain doesn't seem to care. The metal casing is not bonded to ground but instead is isolated. If you ever touched the casing and a board at the same time, you violated ESD procedure. If you wear a grounding strap, where will you clip it to? Really, the miner is in violation as it is. I'm sure the fan blades have a static charge on them and they are mounted much closer to the boards than your cooler will be. If you're really worried despite my observations, just line the cooler in foil and tie the foil to your power supplies casing. It is actually grounded.
HoleShot
Member
**
Offline Offline

Activity: 70
Merit: 28


View Profile
January 26, 2018, 05:31:00 AM
Merited by vapourminer (1), achow101 (1), frodocooper (1)
 #23

I grounded every rack that touch any of my units, I agree that is the only way to go.

May I suggest you check a few things to avoid a safety issue.

It is critical that grounding conductor lengths from the single neutral to earthing bond are very nearly the same to everything that is grounded within reach.

All the power supplies are grounded via the power cord. If you run a different ground conductor to the rack they are sitting on, you should ensure length of that conductor is very nearly the same, within 4', but as close as practicable, as the electrical system ground.

The reason for this is that during surge events, there are massive potential rises which are very short in duration. Grounding systems are designed to ensure the voltage rises and falls at the same time on different devices. This way there is never a difference in potential so current cannot flow between the devices. Just as birds can safely land on 12000V power lines, high voltage cannot hurt you but a potential difference certainly can.

I prefer to isolate my racks from my miners and power supplies with an insulating material between them. I then bond the center of the rack to the electrical ground conductor via the power system ground conductor. I don't like to set the power supply on a metal rack as that creates a high impedance ground loop which will have unpredictable results during a surge event.



jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 26, 2018, 11:49:00 AM
Merited by vapourminer (1), frodocooper (1)
 #24

Holeshot,

Love the info bro. Quality information.

Here is my current setup for cooling the antminer s9 farm and power protecting it from surges that can hit my antminers.




Here is my special insulation separation that I did on my farm. I am also in the process of adding more insulation to divide each unit. What I am doing is I am making a cooling pocket for each unit so it will not allow radiant heat to spread to the antminer s9 that is above, bellow or next to its counterpart.




This is the grounding that I did on each rack. The grounding wire is running directly to my power build out that is passing through a surge protector as well.



This surge protector is my GUARD against anything and everything that hits my panel.


let me know your thoughts and very open to suggestions as I am still in the design process.


Sandal_Hat
Full Member
***
Offline Offline

Activity: 308
Merit: 112


View Profile
January 27, 2018, 04:34:10 AM
 #25

Holeshot,

Love the info bro. Quality information.

Here is my current setup for cooling the antminer s9 farm and power protecting it from surges that can hit my antminers...

Nice looking. Is it an issue that the antminer and psu are touching each other? Or not an issue?



(Moderator's note: This post was edited by frodocooper to trim jabby's quote.)

Selling at 20% of coupon price aka 80% off.
Selling 2 bitmain 125 USD coupons, 18 bitmain 100 USD coupon, 1 bitmain 50 USD coupon

PM me
HoleShot
Member
**
Offline Offline

Activity: 70
Merit: 28


View Profile
January 27, 2018, 07:35:00 AM
Merited by frodocooper (3), vapourminer (1), CrypotoCreanis (1)
 #26

You have done some good research and your layout is certainly clean.

I can help you skip a ton of work here.

If you feel the casing of an operating miner I think you'll find it is at or very nearly at ambient temperature. The same is true for the power supply casing. This observation eliminates any concern that heat transfer could occur by radiation or conduction. Almost all the heat in the miner is transferred into the air stream by conduction due to the temperature delta. The power supply is not a concern as the heat in that unit is Input power - Output power, so about 150 Watts. I did't measure because it is minimal. The heat in the power supply is transmitted by the same method.

The best way to deal with the heat is to ensure there is no pathway allowing the output air to return to the input air. A complete physical barrier is the only way to make that happen. Many assume the hot air will go up or you can use a fan to throw it over there. While most the air will obey your desire enough heat will transfer from the out to the in to measure the problem. Keep in mind that air molecules bounce all around willy-nilly.

If your air pathways on the in and out are not restrictive (very large) and fully isolated, you will not need additional fans. However, if and only if, room fans driving the input and/or output are directly coupled then the fan energy consumed by the tandem fans can be negated as all the fans can slow down due to the push-pull cooperation of the tandem fan system. Assumes all are VFD and well controlled. Sorry to veer off subject.

Grounding. You have done so much more than most miners.

It's important to understand how a surge suppressor works. The voltage specs shown on the label indicate the clamp voltage rating of the bank of MOVs inside the unit. This is the voltage which the MOV essentially shorts the circuit. Surge voltages come in many types but I'll speak typically. The high voltage surge is very high frequency and very short in duration. The MOVs close to limit the high voltage, high frequency, short duration energy from passing through. The rub is that MOVs are fast, but not that fast. All surge suppressors have a let-through voltage based upon the input voltage of the surge. A typical system has one surge suppressor at the main service entrance and the another on each sub panel. This way there are two in series for all circuits.

The reason is that input power line transient surges often exceed the ability of the first surge suppressor to limit the surge voltage enough. If the surge voltage is 10kV the first one may limit if to 2000V then the second will limit that 2000V to the 1200V L to L as shown on the label.

There is likely another MOV in each power supply. It is here the remainder of the surge energy is shunted. The final let through is probably around 600V but with very little energy. The components after the MOV are specified to withstand the voltage and energy expected. I haven't opened up a Bitmain power supply to see, but that is typical design.

Your shelf grounding system is not adequate and presents a safety hazard for personnel. I hate to sound critical because I can see your effort here is outstanding, and you got close. All grounding system connections must be very low impedance due to the high frequency of transient voltage surges. Any impedance will change the time of the voltage rise and fall relative to the devices that are properly connected.

I spent a few years correcting radio sites that had been destroyed by surges, hundreds of sites. 90% of them were caused by poor connections. Most of them started out good but then rotted in the weather or due to dissimilar metals but a fair bit were never correct to begin with.

It looks like there is not good way to ground these shelves. I like the kind with flat surfaces where I can grind off the paint and fasten a two hole compression lug to it with a thin layer of No-Ox. I can also weld all the legs and shelves together. All metal pieces touching your shelves should be electrically bonded with very low impedance connections. I do not trust mechanical connections for grounding at all. The connections you have here will definitely be high impedance at surge frequencies. One problem is the flat surface of the clamp to the round surface of the wire type shelf. I'm not sure from the pic but it looks like the screw is driven in to clamp onto the stranded conductor. If true, that is designed for solid conductor only. If the screw moves the clamp mechanism to squeeze the conductor then that connection would be okay so long as torque is verified and the connection is inspected for corrosion periodically. Corrosion could be avoided since it is indoors with proper application of No-Ox. (directions must be followed exactly).  The legs appear to not be bonded to the shelves.

The shelving appears to be chromium plated, the ground clamp galvanized steel, the ground conductor copper. I'm sure you see the trouble here. No-Ox is required between dissimilar metals. I always use silicone-bronze fasteners because it is non-reactive with other metals. High-grade stainless will get you there too but you must ensure the nickel content is high enough.
 
All server cabinets/shelving units should be grounded with a 2AWG minimum conductor, it is not okay to daisy chain shelving units.

If you want to do it the right way, bond all legs to all shelves by welding (preferred) or with bonding jumpers using two hole compression lugs (definitely the hard way). Use a single connection to the shelving unit 2AWG then tie it to a 2/0 ground conductor spine using compression H-taps.

This is just the tip. There are many, many details which must all be correct to avoid surge caused destruction. However, if you focus on each connection and ensure it is perfectly done, you're probably covered. Every conductor entering the building must be done well. Check the network entrance grounding to ensure they didn't use an inadequate clamp, especially outdoors.
Measure Ohms to ground with an earthing resistance meter, I use a clamp on unit. It should be <5 ohms but I spec. 2. Unless your soil is awesome, it will not be below 5 Ohms. I measure and look at the existing ground rod then calculate additional rods necessary to accomplish my spec. If the soil is sandy it could take 20 x 12' rods and an irrigation system, if the soil is very dark maybe 4. All ground connection out of door must be endothermic bonded.

Keep in mind that the power supply casing is bonded to the AC ground through the power cord. That casing is also bonded to the negative pole of the 12VDC supply line. However, the S9 casing is not bonded to ground or anything. It is fully isolated. So this means you either need to insulate with a high dielectric strength material or bond the power supply case and miner case. Really bad is how they sit on a shelf next to each other and touch just a little bit. This is a very high impedance connection which could even arc during a surge.




 
jabby
Member
**
Offline Offline

Activity: 86
Merit: 12


View Profile
January 28, 2018, 10:47:24 PM
 #27

Holeshot,

I am highly impressed with you knowledge and I thank you very much. I hope everyone get to see your posts on this thread because you are saying critical information and hitting solid points that help speed up the trial and error process exponentially with a much lower cost burn through rate. 

Thank you very much for the kind words as well. I am doing my very best to ensure a safe healthy setup for these machines and it makes me very happy to hear that you see the work I am putting into this to help others.

I am going to implement your information into my new design. Once I get it up and running I will share with pictures, please let me know any other valuable information you have!
HoleShot
Member
**
Offline Offline

Activity: 70
Merit: 28


View Profile
January 28, 2018, 11:16:02 PM
 #28

You are welcome and thank you.

It's always best to start with a drawing for your design. It the electrical schematic is correct, can the mechanical connection plan is good, you will never have a surge problem.

Pages: 1 2 [All]
  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!