Topic: Antminer S9 - Actual CFM at full fan RPM (Read 7362 times)
Awesome thank you for that info
So if I wanted to put in 24 S9's, would I need a intake fan that can supply a total of 24 (S9's)*300 (CFM each) = 7200 CFM? Or should I go a little higher?
Would the out take fan in the hot room need to be at the same level of the intake fan?
Would the out take fan in the hot room need to be at the same level of the intake fan?
Actually, I'd focus on the exhaust fan and perhaps oversize it a bit to handle hot days, but no need to be excessive about it.
Getting rid of the hot air is the primary goal. Then all you need to do is supply sufficient reasonably filtered (say Mirv 3) intake air. You can do that by providing approximately 1 square foot of intake per miner or by using fans. Of course, you save power by just having intake space. Given a sufficient intake opening, the fans on the miners will be happy to pump the air into your hot aisle.
So guys, still didn't get an answer to the topic yet. I'm aware of the specs of the stock fans that come with the S9, but I am hoping to find out what is the pressure differential across the fan so I can calculate the actual CFM through the S9 considering pressure drop across heat sinks etc?
I calculated my air needs based on a combined total of 300 cfm/s9. So far, so good (at least as far as air flow is concerned). I'm running about 120 S9s using that calc.
It should be noted that 841s draw less air.
I mine at home, tried all kinds of fans. It's really a waste of time, resources and downtime of your unit to install and test. The only thing close was the ultra kraze fan which still barely moved enough air for an s7.
Tried the Notcrula's or what ever their brand name is for these machines its junk, for your PC their awesome. If you can down clock your unit significantly you can use these fans but even at full speed they wont move much air through the heat sinks of an asic. Tenting is not a good idea with s9 but maybe for gpu farm.
I would down clock and drop fan speed to reduce db's. There should be write ups on how to do both for auto tune s9's. Good luck.
Tried the Notcrula's or what ever their brand name is for these machines its junk, for your PC their awesome. If you can down clock your unit significantly you can use these fans but even at full speed they wont move much air through the heat sinks of an asic. Tenting is not a good idea with s9 but maybe for gpu farm.
I would down clock and drop fan speed to reduce db's. There should be write ups on how to do both for auto tune s9's. Good luck.
@bpg2001bpg Any update on your project ?
I've seen this product which is quite similar but was unable to get any feedback on it : https://www.silent-miner.eu/store/Silent-Miner-p103601773
I've seen this product which is quite similar but was unable to get any feedback on it : https://www.silent-miner.eu/store/Silent-Miner-p103601773
I didn't do the research. I bought an industrial miner. It's too loud (I think the pitch of the fan is annoying) to run in my home. I still want to fix it and I'm hoping for some help. I found this forum looking for some data on static pressure.
My current plan is to use a grow tent blower to replace the fans. If the previous post of @VentMine is accurate, there is 18mm or 0.7 WG of static pressure. This amazon blower can move 190 CFM at 0.7 WG and operates at 33dB(A)
https://www.amazon.com/dp/B01M7PWPPA/ref=sspa_dk_detail_6?psc=1&pd_rd_i=B01M7PWPPA&pd_rd_wg=KSpw7&pd_rd_r=782MA6JN488PG4QG8NMB&pd_rd_w=wH5V2
My plan is to use a blower with a fan-b-gone:
https://www.ecwid.com/store/fan-b-gone/
I understand that if the blower goes out, the miner will burn up. At this point I am okay with that. Maybe in the future I can find a way to monitor the fan speed and shut down the miner if it goes out.
If this plan is even sound(criticism welcome), I have a few other blower options that are less expensive, smaller, quieter, or blowers that are even more powerful, but louder.
Thoughts?
My current plan is to use a grow tent blower to replace the fans. If the previous post of @VentMine is accurate, there is 18mm or 0.7 WG of static pressure. This amazon blower can move 190 CFM at 0.7 WG and operates at 33dB(A)
https://www.amazon.com/dp/B01M7PWPPA/ref=sspa_dk_detail_6?psc=1&pd_rd_i=B01M7PWPPA&pd_rd_wg=KSpw7&pd_rd_r=782MA6JN488PG4QG8NMB&pd_rd_w=wH5V2
My plan is to use a blower with a fan-b-gone:
https://www.ecwid.com/store/fan-b-gone/
I understand that if the blower goes out, the miner will burn up. At this point I am okay with that. Maybe in the future I can find a way to monitor the fan speed and shut down the miner if it goes out.
If this plan is even sound(criticism welcome), I have a few other blower options that are less expensive, smaller, quieter, or blowers that are even more powerful, but louder.
Thoughts?
Fully agree with Fanatic26 - these units were never designed for home use. That said, I had (
running in my basement heating my house for over a year. If you have an understanding family, its not that bad. I did use vent filters to quiet down the whinny ones. Fortunately our TV room was on the 2nd floor.
Oh, regarding CFM through the unit, eg. actual airflow from front to back with the 2 fans combined, I had previously read on the forum that at top speed the fans move about 300 CFM total.
running in my basement heating my house for over a year. If you have an understanding family, its not that bad. I did use vent filters to quiet down the whinny ones. Fortunately our TV room was on the 2nd floor.Oh, regarding CFM through the unit, eg. actual airflow from front to back with the 2 fans combined, I had previously read on the forum that at top speed the fans move about 300 CFM total.
Good idea, maybe I can finally contribute something useful to the community hehe. Will update with results 
+VentMine Any progress on measuring the pressure drop/differential?
No I never bothered to do it lol. I found out the fan CFM was 220 / S9, made some assumptions and built my farm out on the assumptions :S
Good idea, maybe I can finally contribute something useful to the community hehe. Will update with results
+VentMine Any progress on measuring the pressure drop/differential?
...making the air "cold" is not as important as creating enough air flow. The ideal setup for S9 is to have a negative pressure room for the exhaust and a positive pressure room for the intake.. [] Typically miners are happy at up to 80 degrees if there is sufficient air flow. Making a room cold doesn't do anything for you if you can't allow the machines to breath and isolate them from the hot exhaust air.
This is a very good and important point. I was wondering about the 25-40˚C operating temperature specified in the S9 official manual. Saying 'Typically miners are happy at up to 80 degrees' do you mean they operate at the maximal hashrate and then over 80˚F the hashrate starts decreasing?
Quote
so what about with this cooling warehouse system? (internal temperature is 0°C)
I realize this thread is a little old, but just wanted to make a point that making the air "cold" is not as important as creating enough air flow. The ideal setup for S9 is to have a negative pressure room for the exhaust and a positive pressure room for the intake, even in the dead heat of Florida in the summer, this concept works very well at room temperature believe it or not, as long as ambient is not too hot. Typically miners are happy at up to 80 degrees if there is sufficient air flow. Making a room cold doesn't do anything for you if you can't allow the machines to breath and isolate them from the hot exhaust air. How you achieve this is irrelevant, there are 100 ways to do it.
so what about with this cooling warehouse system? (internal temperature is 0°C)
Evaporator Cooling Room
Evaporator Cooling Room
Interesting picture... It looks like industrial style air handlers in a dedicated mining space. That probably cost a lot of money to install and requires the dedicated space. Most home miners are not going to have a room like that available (especially if it requires approval by the spouse).
I guess it is really a case of accepting that these miners are industrial - trying to change them to fit the needs of a home miner is a mistake. But if you have a room like that and the power to run to it, then go for it! (Just my humble opinion!)
As a newbie to home mining I acknowledge I made a mistake in buying the S9. I read over and over that is was loud and designed for data centers but I did not listen.
I decided after trying to quiet it down to sell it and fall back on three Avalon 741s with the voltage offset configured to -2. These are still industrial miners but with that setting the fans are tolerable. In addition the Canaan hardware seems to be much more reliable. The Avalons have a better heatsink arrangement which definitely cools better than the Antminers.
Trying to change the fans on any of these miners is A BAD IDEA. These high CFM and high static pressure fans are actually very high quality and specifically chosen for this cooling job. They are loud because they are pushing a lot of air. No PC style or quiet consumer fan is gong to work on these and keep things cool enough.
I decided after trying to quiet it down to sell it and fall back on three Avalon 741s with the voltage offset configured to -2. These are still industrial miners but with that setting the fans are tolerable. In addition the Canaan hardware seems to be much more reliable. The Avalons have a better heatsink arrangement which definitely cools better than the Antminers.
Trying to change the fans on any of these miners is A BAD IDEA. These high CFM and high static pressure fans are actually very high quality and specifically chosen for this cooling job. They are loud because they are pushing a lot of air. No PC style or quiet consumer fan is gong to work on these and keep things cool enough.
so what about with this cooling warehouse system? (internal temperature is 0°C)
Evaporator Cooling Room
As a newbie to home mining I acknowledge I made a mistake in buying the S9. I read over and over that is was loud and designed for data centers but I did not listen.
I decided after trying to quiet it down to sell it and fall back on three Avalon 741s with the voltage offset configured to -2. These are still industrial miners but with that setting the fans are tolerable. In addition the Canaan hardware seems to be much more reliable. The Avalons have a better heatsink arrangement which definitely cools better than the Antminers.
Trying to change the fans on any of these miners is A BAD IDEA. These high CFM and high static pressure fans are actually very high quality and specifically chosen for this cooling job. They are loud because they are pushing a lot of air. No PC style or quiet consumer fan is gong to work on these and keep things cool enough.
I decided after trying to quiet it down to sell it and fall back on three Avalon 741s with the voltage offset configured to -2. These are still industrial miners but with that setting the fans are tolerable. In addition the Canaan hardware seems to be much more reliable. The Avalons have a better heatsink arrangement which definitely cools better than the Antminers.
Trying to change the fans on any of these miners is A BAD IDEA. These high CFM and high static pressure fans are actually very high quality and specifically chosen for this cooling job. They are loud because they are pushing a lot of air. No PC style or quiet consumer fan is gong to work on these and keep things cool enough.
I think you guys just need to accept the fact that you bought an INDUSTRIAL miner that makes lots of noise because its not meant to be ran at home. Every day people are posting on how to make the miner quieter. ITS NOT MEANT TO BE QUIET. Putting weaker, quieter fans on it will just cause you problems. Just accept that you did not do the proper research before buying it and now you have to deal with the consequences.
Also comparing an s5 to an s9 makes no sense. With each generation of chips the heat load increases by a ton so what works for an s5 has zero bearing on what will work for an s9.
Also comparing an s5 to an s9 makes no sense. With each generation of chips the heat load increases by a ton so what works for an s5 has zero bearing on what will work for an s9.
Hi guys,
i read about that the noctua NF-F12 are too weak for the S9. I have some on my S5 and the work like a charm!
So I try to figure out how I could get this S9 more silent.
I am thinking about taking 2 noctua nf-a14 industrialppc-3000 pwm and a fan adapter that downsizes from 140 to 120mm.
The nf-a14 industrialppc-3000 pwm has 269 CFM, so it should blow more air at 3000rpm as the original at 6000rpm with 210cfm.
What are you thinking bout this guys?
mine on!
i read about that the noctua NF-F12 are too weak for the S9. I have some on my S5 and the work like a charm!
So I try to figure out how I could get this S9 more silent.
I am thinking about taking 2 noctua nf-a14 industrialppc-3000 pwm and a fan adapter that downsizes from 140 to 120mm.
The nf-a14 industrialppc-3000 pwm has 269 CFM, so it should blow more air at 3000rpm as the original at 6000rpm with 210cfm.
What are you thinking bout this guys?
mine on!
Noctua nf-a14 industrialppc-3000 pwm how can be fit with screw hole on S9 you think? There is something have any trick for it? Noctua nf-a14 and Noctua nf-a12 not the same? I thought the only matter is the size between with a12 - a14 models.
Hi guys,
i read about that the noctua NF-F12 are too weak for the S9. I have some on my S5 and the work like a charm!
So I try to figure out how I could get this S9 more silent.
I am thinking about taking 2 noctua nf-a14 industrialppc-3000 pwm and a fan adapter that downsizes from 140 to 120mm.
The nf-a14 industrialppc-3000 pwm has 269 CFM, so it should blow more air at 3000rpm as the original at 6000rpm with 210cfm.
What are you thinking bout this guys?
mine on!
i read about that the noctua NF-F12 are too weak for the S9. I have some on my S5 and the work like a charm!
So I try to figure out how I could get this S9 more silent.
I am thinking about taking 2 noctua nf-a14 industrialppc-3000 pwm and a fan adapter that downsizes from 140 to 120mm.
The nf-a14 industrialppc-3000 pwm has 269 CFM, so it should blow more air at 3000rpm as the original at 6000rpm with 210cfm.
What are you thinking bout this guys?
mine on!
@voodooman you alredy try it? its working that NF-A14 industrial PPC?
i think is 269 m3/h wich is 158 CFM.
Maybe you should put some new cooling paste onto your chips.
This dropped ym temps by 10°!
If your S5 still has its original colling paste on it, i think you should change it.
Cause after 2 or 3 years the paste isn't working well anymore.
Costs a bit of time and maybe 10-20€ for cooling paste but i think it's worth it.
And don't test the nf12 on the S9. They are way too weak for the S9 and will burn it!!
This dropped ym temps by 10°!
If your S5 still has its original colling paste on it, i think you should change it.
Cause after 2 or 3 years the paste isn't working well anymore.
Costs a bit of time and maybe 10-20€ for cooling paste but i think it's worth it.
And don't test the nf12 on the S9. They are way too weak for the S9 and will burn it!!
i read about that the noctua NF-F12 are too weak for the S9. I have some on my S5 and the work like a charm!
Not tested this(Noctua NF-F12 iPPC-3000 PWM) on S9 but I'm not happy with them on S5:https://i.imgur.com/WZEnmIC.png
Quite high temperature on low frequency, continuous work on 2880 rpm, no magic. Maybe I'll test them with one big fan before go back to the original fans.
Hi guys,
i read about that the noctua NF-F12 are too weak for the S9. I have some on my S5 and the work like a charm!
So I try to figure out how I could get this S9 more silent.
I am thinking about taking 2 noctua nf-a14 industrialppc-3000 pwm and a fan adapter that downsizes from 140 to 120mm.
The nf-a14 industrialppc-3000 pwm has 269 CFM, so it should blow more air at 3000rpm as the original at 6000rpm with 210cfm.
What are you thinking bout this guys?
mine on!
i read about that the noctua NF-F12 are too weak for the S9. I have some on my S5 and the work like a charm!
So I try to figure out how I could get this S9 more silent.
I am thinking about taking 2 noctua nf-a14 industrialppc-3000 pwm and a fan adapter that downsizes from 140 to 120mm.
The nf-a14 industrialppc-3000 pwm has 269 CFM, so it should blow more air at 3000rpm as the original at 6000rpm with 210cfm.
What are you thinking bout this guys?
mine on!
The 6000 RPM delta fan and the Ultra Sleek Vortex 14 are in totally different classes, it won't work. Delta is doing 210+ CFM at max speed, the vortex only 98CFM. There is almost nothing comparable to Delta fans on these things.
There's a fan called the Scythe Ultra Kaze that runs at 3000 RPM, pushes a ton of air and is slightly quieter, but it only works if u have the non-autotune S9 and can downclock the frequency a bit.
There's a fan called the Scythe Ultra Kaze that runs at 3000 RPM, pushes a ton of air and is slightly quieter, but it only works if u have the non-autotune S9 and can downclock the frequency a bit.
If I may, the Ultra Kaze is good - but - Enermax Twister Storm is better @ 3500 RPM for +150 CFM / +11 Static Pressure. It's working great on my Spondoolies SP20E. I've also, in order, a possible new king of the road (non delta / low noise), the Silverstone FHP141-VF @ 2000 RPM for +170 CFM. It's a 140mm mounting as a 120mm, one catch - the Static Pressure seems to be around 4...
Good idea, maybe I can finally contribute something useful to the community hehe. Will update with results
So guys, still didn't get an answer to the topic yet. I'm aware of the specs of the stock fans that come with the S9, but I am hoping to find out what is the pressure differential across the fan so I can calculate the actual CFM through the S9 considering pressure drop across heat sinks etc?
Probably because no one has measured the pressure differential. So -- get some clear flex PVC tubing and make your own manometer (google it -- very very simple) , poke a hole for access to the space between the fan and heatsinks and ya got yer answer.
So guys, still didn't get an answer to the topic yet. I'm aware of the specs of the stock fans that come with the S9, but I am hoping to find out what is the pressure differential across the fan so I can calculate the actual CFM through the S9 considering pressure drop across heat sinks etc?
YS Tech has fans that can argue with most Deltas, but none to my knowlage that can argue with the FFB series or it's derivatives like the QFR series - the YS Tech lines are designed to be competition for the Delta AFB "low backpressure design" series and it's derivatives.
I'm not aware of anyone else that even gets CLOSE to 200cfm at the kind of backpressure the FFB/QFR/etc lines are designed for in a 120mm format.
I'm not aware of anyone else that even gets CLOSE to 200cfm at the kind of backpressure the FFB/QFR/etc lines are designed for in a 120mm format.
I can't find any seller of those fans, can any of you please provide a link on eBay or Amazon?
I'm not aware of anyone else that even gets CLOSE to 200cfm at the kind of backpressure the FFB/QFR/etc lines are designed for in a 120mm format.
That's pretty cool, didn't realize these fans were such boss
The 6000 RPM delta fan and the Ultra Sleek Vortex 14 are in totally different classes, it won't work. Delta is doing 210+ CFM at max speed, the vortex only 98CFM. There is almost nothing comparable to Delta fans on these things.
There's a fan called the Scythe Ultra Kaze that runs at 3000 RPM, pushes a ton of air and is slightly quieter, but it only works if u have the non-autotune S9 and can downclock the frequency a bit.
There's a fan called the Scythe Ultra Kaze that runs at 3000 RPM, pushes a ton of air and is slightly quieter, but it only works if u have the non-autotune S9 and can downclock the frequency a bit.
YS Tech has fans that can argue with most Deltas, but none to my knowlage that can argue with the FFB series or it's derivatives like the QFR series - the YS Tech lines are designed to be competition for the Delta AFB "low backpressure design" series and it's derivatives.
I'm not aware of anyone else that even gets CLOSE to 200cfm at the kind of backpressure the FFB/QFR/etc lines are designed for in a 120mm format.
MrKaizer I'm wondering where you found this information, just for my reference? Seems like it is reasonable, based on the fan chart for the model QFR1212GHE-SP01 that comes with the S9, that CFM would correspond to a differential pressure of 18 mm of h20 or so, and falls right in the middle of the chart for the model.
Sorry for my mistake, that information is from S7, no S9.
For the 6000rpm fan is around 210 CFM.
The 6000 RPM delta fan and the Ultra Sleek Vortex 14 are in totally different classes, it won't work. Delta is doing 210+ CFM at max speed, the vortex only 98CFM. There is almost nothing comparable to Delta fans on these things.
There's a fan called the Scythe Ultra Kaze that runs at 3000 RPM, pushes a ton of air and is slightly quieter, but it only works if u have the non-autotune S9 and can downclock the frequency a bit.
There's a fan called the Scythe Ultra Kaze that runs at 3000 RPM, pushes a ton of air and is slightly quieter, but it only works if u have the non-autotune S9 and can downclock the frequency a bit.
Does anyone know what the difference in pressure is (inches or mm h20) across the Antminer S9 when running at full rpm?
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
It's around 188 m³/h.
I thinking about use some Prolimatech Ultra Sleek Vortex 140mm, I know that its 166 m³/h, but I'm talking about ~20 dBa at full speed instead of ~75!
MrKaizer I'm wondering where you found this information, just for my reference? Seems like it is reasonable, based on the fan chart for the model QFR1212GHE-SP01 that comes with the S9, that CFM would correspond to a differential pressure of 18 mm of h20 or so, and falls right in the middle of the chart for the model.
Does anyone know what the difference in pressure is (inches or mm h20) across the Antminer S9 when running at full rpm?
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
Why? the firmware will insist you run a fan at 4000 or 4500 rpm if it is auto tune. So it will stay noisy on auto tune or just shut off.
I'm sizing some exhaust fans and was thinking to size them such that they can handle the maximum expected CFM of my farm. For example, if I have 40 x S9's and the max CFM per S9 is 180 CFM (for example), then I would want to ensure my exhaust fans can handle 40x180 = 7200 CFM. I figure this is a simpler way to size the fans rather than try to calculate via heat load and ambient temperature info, which may have a degree of error. Does this make sense or maybe you know of a better way?
Does anyone know what the difference in pressure is (inches or mm h20) across the Antminer S9 when running at full rpm?
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
It's around 188 m³/h.
I thinking about use some Prolimatech Ultra Sleek Vortex 140mm, I know that its 166 m³/h, but I'm talking about ~20 dBa at full speed instead of ~75!
Does anyone know what the difference in pressure is (inches or mm h20) across the Antminer S9 when running at full rpm?
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
Why? the firmware will insist you run a fan at 4000 or 4500 rpm if it is auto tune. So it will stay noisy on auto tune or just shut off.
Does anyone know what the difference in pressure is (inches or mm h20) across the Antminer S9 when running at full rpm?
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
I found the specs of the fans (QFR1212GHE-SP01) and the CFM vs deltaP and am trying to calculate the actual CFM through the miner when running at full RPM.
THX
Jump to: