Topic: Suck or blow when adding a large external fan to a miner? (Read 148 times)
Actually the dev fee for S9 is still 2%, so if you want to see it that way, its 0%.
Thanks for that hint, that would be the easiest way to go I suppose.
I've just been wondering why Braiins offers free software AND reduced mining fees if using this software. How does that work as a business model (ducking for cover
)? I installed BraiinsOS on SD card and I think the penny has dropped:
It costs 2% to use Braiins pool when using other software;
it costs 2.5% to use BraiinsOS and then the Braiins pool is free, means 2.5% fee in total, i.e. an extra 0.5% fee for Braiins compared to not using their software.
The installation was very easy (for someone who has played around with miners for a few weeks) and now the S9J is running without noisy stock fans, just the Phresh fan sucking.
Well the unfortunate thing is that with factory fw it has to go very high and you cannot control it. It would be best if it didn't need to reach 130 to stop...
The fan sense is a double edge sword. Sometimes a fan fails but its still registering RPM (can't turn on its own but is being pushed/pulled by the others), and you will notice higher temps, and sometimes lower RPM from that fan.
The fan sense is a double edge sword. Sometimes a fan fails but its still registering RPM (can't turn on its own but is being pushed/pulled by the others), and you will notice higher temps, and sometimes lower RPM from that fan.
If the fan fails its basically the same as when one (or more) of the small fans fail. The fw should stop hashing when it overheats, in Braiins OS controlled by the Dangerous value, which you can lower even more.
I tested that and it works.
When turning off all air flow to the miner, the chip temp goes up to 130C and it stops mining, then tries to restart a couple of times, then blinks the red LED and runs the fans on high without further mining. After power cycling, all is back to normal.
When disconnecting a stock fan, it stops mining right away.
...
...
if you use braiins you can not worry about the fans at all
...
...
...
if you use braiins you can not worry about the fans at all
...
...
Thanks for that hint, that would be the easiest way to go I suppose.
I've just been wondering why Braiins offers free software AND reduced mining fees if using this software. How does that work as a business model (ducking for cover
)?
If the fan fails its basically the same as when one (or more) of the small fans fail. The fw should stop hashing when it overheats, in Braiins OS controlled by the Dangerous value, which you can lower even more.
The theory is that sucking is better because you don't add the heat of the fan itself. It also tends to attract more dust because its negative pressure for the miner.
So i guess it depends in your actual setup. If its well filtered with noise insulated ducts that reach the outside on both ends, i would go with suck. But if its something like no inlet duct or scooping air from the inside, maybe blow is better.
Obligatory reference:
With the likes of the infinity 8" you don't need any other fans at all. 800cfm is twice the required volume so it should be fine at 50% or less.
So i guess it depends in your actual setup. If its well filtered with noise insulated ducts that reach the outside on both ends, i would go with suck. But if its something like no inlet duct or scooping air from the inside, maybe blow is better.
Obligatory reference:
With the likes of the infinity 8" you don't need any other fans at all. 800cfm is twice the required volume so it should be fine at 50% or less.
let me find the kit I used.
https://www.ebay.com/itm/304119146264
if you use braiins you can not worry about the fans at all
https://www.ebay.com/itm/285066554125. I used the 8 inch fan as it is a bit cheaper than the infinity and does 735 cfm more than enough
https://www.ebay.com/itm/203764567923 spoofers are cheap if you do not want braiins
there is a down side to doing this if the 8 inch fan fails the miner will overheat
The theory is that sucking is better because you don't add the heat of the fan itself. It also tends to attract more dust because its negative pressure for the miner.
So i guess it depends in your actual setup. If its well filtered with noise insulated ducts that reach the outside on both ends, i would go with suck. But if its something like no inlet duct or scooping air from the inside, maybe blow is better.
Obligatory reference:
With the likes of the infinity 8" you don't need any other fans at all. 800cfm is twice the required volume so it should be fine at 50% or less.
PS: I don't like leaving the small fans because they can be forced to generate electricity by the external fan.
So i guess it depends in your actual setup. If its well filtered with noise insulated ducts that reach the outside on both ends, i would go with suck. But if its something like no inlet duct or scooping air from the inside, maybe blow is better.
Obligatory reference:
With the likes of the infinity 8" you don't need any other fans at all. 800cfm is twice the required volume so it should be fine at 50% or less.
PS: I don't like leaving the small fans because they can be forced to generate electricity by the external fan.
you don't stack a fan onto the stock fans. the stock fans will restrict the new fan if they are slower than the new fan.
and the new fan will restrict the stock fans if it is weaker.
if you have a s19 there are kits to fully remove and bypass the stock fans and have a more powerful new one.
Turns out you are spot-on.and the new fan will restrict the stock fans if it is weaker.
if you have a s19 there are kits to fully remove and bypass the stock fans and have a more powerful new one.
The powerful Phresh fan actually increases the noise levels by spinning up the stock fans to very high speeds. And they seem to excite each other with annoying interference resonance loops.
Would it be possible to 'spoof' the controlboard with some small PWM fans that can spin up to say 6000rpm, just until I can get some of the spoofing boards?
I suppose I could also leave the stock fans connected to the the control board (but not connected to the miner housing) for now. Just need to fasten them so they don't take off at bootup!
Re: Sucking or blowing: Sucking appears much more convenient so far, for 2 reasons:
1) the dangerous intake side of the Phresh fan gets covered with the adapter, less risk of fingers ending up in the fan blades
2) Exhaust hose is easily attached to the Phresh fan to direct the warm air to where it's needed, therefore no 2nd adapter required.
you don't stack a fan onto the stock fans. the stock fans will restrict the new fan if they are slower than the new fan.
and the new fan will restrict the stock fans if it is weaker.
if you have a s19 there are kits to fully remove and bypass the stock fans and have a more powerful new one.
and the new fan will restrict the stock fans if it is weaker.
if you have a s19 there are kits to fully remove and bypass the stock fans and have a more powerful new one.
You are correct, but I don't know if the air flow restriction and resulting increased energy consumption is significant, considering the overall energy consumption by the miner.
And then there's the waiting time to get the fan spoofing device to where I am. Definitely maybe later!
I want to keep the stock fans in place to avert any issues when / if the external fan fails, or something in my experimental setup goes wrong.
I will also quite likely remove the external fan repeatedly in order to experiment / develop quieter cooling setups with other miners.
Regarding 3): Collapsing tubing: I tested it, the tubing tries to shrink in lengths and will tear itself off any insecure attachment, but the lumen does not collapse. One less worry.
you don't stack a fan onto the stock fans. the stock fans will restrict the new fan if they are slower than the new fan.
and the new fan will restrict the stock fans if it is weaker.
if you have a s19 there are kits to fully remove and bypass the stock fans and have a more powerful new one.
and the new fan will restrict the stock fans if it is weaker.
if you have a s19 there are kits to fully remove and bypass the stock fans and have a more powerful new one.
I assume most miners have fans on each end, one side blowing in and the other side sucking out.
When using an external (much more powerful and more quiet) fan, what considerations could guide the decision to have it blow into or suck out of the miner?
My intention is to leave the stock fans on the miner and then blow (or suck) increasing amounts of additional air through it until the stock fans spin at their minimum rpm, or until the external fan is louder than the stock fans, whatever comes first.
I got one of these fans, it feels like it can produce at least several times the air flow that my S9J miner requires: (Don't click the link unnecessarily, bloated advertising!): https://www.hyper-phresh.com/hyperfan
These are the specs in a nutshell:
It comes with a speed controller and I can print any adaptor needed to make it fit to a miner.
Specific questions that crossed my mind are these:
1) The external fan might project more noise on the outlet than the inlet side. I tested it, the noise projection seems the same in either direction. Means that from a noise reduction perspective, it does not matter which side of the fan is 'open to the environment' or maybe connected to the noise reducing ducting.
2) The fan's ability to create suction might be different from it's ability to create pressure. To be tested, not sure how, yet. I'll try to make the 3D printed adaptor so it will fit either side equally, so I can use it both ways.
3) The tubing (Seahawk acoustic duct 200mm https://www.benchmarkbioponics.com.au/products/acoustic-polyester-ducting-100mm-x-5m ) might collapse when the fan sucks through it. Not an issue if the fan is connected directly to the miner, but might be an issue if there is a length of ducting between miner and the sucking fan. It could even be an issue if blowing into the ducting before going into the miner (i.e. it might leak or pop open), but I expect that to be much less likely than collapsing when sucked on.
Any suggestions to help understand the potential differences between blowing and sucking would be much appreciated!
Also, any suggestions for how to objectively test the difference once I have printed an adaptor.
When using an external (much more powerful and more quiet) fan, what considerations could guide the decision to have it blow into or suck out of the miner?
My intention is to leave the stock fans on the miner and then blow (or suck) increasing amounts of additional air through it until the stock fans spin at their minimum rpm, or until the external fan is louder than the stock fans, whatever comes first.
I got one of these fans, it feels like it can produce at least several times the air flow that my S9J miner requires: (Don't click the link unnecessarily, bloated advertising!): https://www.hyper-phresh.com/hyperfan
These are the specs in a nutshell:
It comes with a speed controller and I can print any adaptor needed to make it fit to a miner.Specific questions that crossed my mind are these:
1) The external fan might project more noise on the outlet than the inlet side. I tested it, the noise projection seems the same in either direction. Means that from a noise reduction perspective, it does not matter which side of the fan is 'open to the environment' or maybe connected to the noise reducing ducting.
2) The fan's ability to create suction might be different from it's ability to create pressure. To be tested, not sure how, yet. I'll try to make the 3D printed adaptor so it will fit either side equally, so I can use it both ways.
3) The tubing (Seahawk acoustic duct 200mm https://www.benchmarkbioponics.com.au/products/acoustic-polyester-ducting-100mm-x-5m ) might collapse when the fan sucks through it. Not an issue if the fan is connected directly to the miner, but might be an issue if there is a length of ducting between miner and the sucking fan. It could even be an issue if blowing into the ducting before going into the miner (i.e. it might leak or pop open), but I expect that to be much less likely than collapsing when sucked on.
Any suggestions to help understand the potential differences between blowing and sucking would be much appreciated!
Also, any suggestions for how to objectively test the difference once I have printed an adaptor.
Jump to: