Fan ohms and fan emulators / spoofers

[Nomadix]

Am trying to figure out the ohms setting for the fans that are on a Antminer S19, testing out the fan simulators I see that the fan speed readings are off in the 8k range as opposed to the norm defined by bitmain.

As the simulators I have have an adjustable pot that I can adjust the resistance levels would like to know if anyone has the data on this.

Thanks

[1715273775]

1715273775

[1715273775]

1715273775

[BitMaxz]

What do you plan and why do you want to use the fan simulator for s19?
Are you planning to run it with immersion cooling?

If yes then you can just switch to other firmware like Braiins OS to bypass the fan you can check the link below for the download link and FAQs

- https://braiins.com/os/plus

[NotFuzzyWarm]

Fans and fan emulators do not use resistance (ohms) for control. The are changing the duty cycle of a pulse train to report and set speed. I believe most use 1kHz with a 10% = slow speed to 100% = full speed duty cycle.

[Nomadix]

What do you plan and why do you want to use the fan simulator for s19?
Are you planning to run it with immersion cooling?

If yes then you can just switch to other firmware like Braiins OS to bypass the fan you can check the link below for the download link and FAQs

- https://braiins.com/os/plus

You are correct it is for immersion mining, I noticed that some of the spoofers I have are working and some are not, but what I noticed the most were the fan speeds rpm off from the manufacturer speeds. I was already looking at the asic.to firmware as a consideration, but thank you for the headsup on brains will look more into it. If you do have some personnel experience on the brains software how much power savings can one achieve with base clock speeds after auto tuning ?

[Nomadix]

Fans and fan emulators do not use resistance (ohms) for control. The are changing the duty cycle of a pulse train to report and set speed. I believe most use 1kHz with a 10% = slow speed to 100% = full speed duty cycle.

Yes the spoofer comes with 1kHZ pot that can be adjusted, I was trying to figure out if the ohms would be basis to check on the multimeter in reference to the fan ohms, still trying to figure out the correct reference point to adjust the spoofer pot to to get the right simulated fan speeds, as it seems that is the problem with fan failure readings when checkin asic logs.

[NotFuzzyWarm]

The resistance that the pot is set to is pretty meaningless without knowing what the rest of the spoofer circuit is using. The pot is part of a timing circuit and without knowing what that circuit is...

What you need to measure is the duty cycle and frequency of the pulse train being sent. Best way is with either a multi-meter that has those functions, use an oscilloscope, or -- measure the average DC voltage you see between the yellow tach wire and the black wire (common). You can assume that 100% = 12v. That will at least let you calculate the duty cycle different fans speeds give but you still need to find out the baseline frequency, as I mentioned before, 1kHz is common.

[Nomadix]

The resistance that the pot is set to is pretty meaningless without knowing what the rest of the spoofer circuit is using. The pot is part of a timing circuit and without knowing what that circuit is...

What you need to measure is the duty cycle and frequency of the pulse train being sent. Best way is with either a multi-meter that has those functions, use an oscilloscope, or -- measure the average DC voltage you see between the yellow tach wire and the black wire (common). You can assume that 100% = 12v. That will at least let you calculate the duty cycle different fans speeds give but you still need to find out the baseline frequency, as I mentioned before, 1kHz is common.

Thank you for the heads up on this, much appreciated

[NotFuzzyWarm]

Or just google 'fan simulator circuit' or 'fake fan signal' to get prints or information about it. One nice discussion is here

[mikeywith]

You are correct it is for immersion mining, I noticed that some of the spoofers I have are working and some are not, but what I noticed the most were the fan speeds rpm off from the manufacturer speeds. I was already looking at the asic.to firmware as a consideration, but thank you for the headsup on brains will look more into it. If you do have some personnel experience on the brains software how much power savings can one achieve with base clock speeds after auto tuning ?

I have no experience with immersion cooling but I have tested both firmware in the past, and I personally think Vnish (asic.to / awesomeminer) is far better, although,Vnish is a modified bitmain firmware (originally Cgminer) while BOs+ is written from scratch and thus does not violate any license, but then "copyrights" aside, I would personally use Vnish over BOs.

But then again, everyone is different, you can try both and see what works best for you, the power saving on both firmware versions is pretty good compared to the stock firmware, it's good enough to justify the fees especially if your power rate is high.

[BitMaxz]

If you do have some personnel experience on the brains software how much power savings can one achieve with base clock speeds after auto tuning ?

I do not have an s19 unit so I can't give you how much power you can save and Autotuning on BraiinsOS you can manually set how much power limit that your PSU can provide on 3 hashboards.

Braiins OS will automatically find the optimal settings for your unit(Of every ASIC chips) and I think if the miner is cooler then it will give you more boost on hashrate and I think also can save more power based on my s9 miner.

[mikeywith]

Braiins OS will automatically find the optimal settings for your unit(Of every ASIC chips) and I think if the miner is cooler then it will give you more boost on hashrate and I think also can save more power based on my s9 miner.

I am sure you know this, but I would like to clarify to OP that a hashrate boost means a sacrifice of efficiency, when we say custom firmware is more efficient than the stock we are talking about (when comparing the same hashrate), in other words, if the stock firmware gives you 100th and 3100w, you could expect the custom firmware to give you 100th and 2900w or maybe lower, so you go down from 31w per th to 29w per th.

Now if you underclock the miner using custom firmware, you get even a better efficiency at the cost of losing hashrate, so maybe you can achieve something like 70th at 1600w, or 23w/th, which is a lot more efficient than the stock.

However, if you were to go to 120th overclock, then the efficiency of this setup will be less than the stock, you will also need to replace the PSU, if 130-140th is your target, then a 6kw is needed, pushing towards 160-180 will require an 8kw PSU, overclocking with the stock PSU is going to pretty limited.

[Nomadix]

Or just google 'fan simulator circuit' or 'fake fan signal' to get prints or information about it. One nice discussion is here

Thank you for the info and links.. 

[Nomadix]

Thx mikeywith and BitMaxz for the info and as per my objective with the possible firmware integration, I am only considering in increasing the overall efficiency of the asics, whilst I have alot of experience on gpu mining am still working my way with the asics setup.

Asics are all in a immersion tank environment, I will be testing my 2 old s9's in a smaller tank initially and if all checks out then will apply the rest on the main units.

[mikeywith]

Thx mikeywith and BitMaxz for the info and as per my objective with the possible firmware integration, I am only considering in increasing the overall efficiency of the asics, whilst I have alot of experience on gpu mining am still working my way with the asics setup.

Asics are all in a immersion tank environment, I will be testing my 2 old s9's in a smaller tank initially and if all checks out then will apply the rest on the main units.

You are welcome, please keep this topic updated, it would be a great addition to the forum given the little info related to immersion cooling, also some pictures and more details about your setup will be apppricated.

I gave you some merit points to help you rank up so you can post with less resitctions.

[philipma1957]

op Keep posting I would love to read about your results.

[Nomadix]

Thx mikeywith and BitMaxz for the info and as per my objective with the possible firmware integration, I am only considering in increasing the overall efficiency of the asics, whilst I have alot of experience on gpu mining am still working my way with the asics setup.

Asics are all in a immersion tank environment, I will be testing my 2 old s9's in a smaller tank initially and if all checks out then will apply the rest on the main units.

You are welcome, please keep this topic updated, it would be a great addition to the forum given the little info related to immersion cooling, also some pictures and more details about your setup will be apppricated.

I gave you some merit points to help you rank up so you can post with less resitctions.

Thank you for the firmware upgrade.... will be posting shortly...

[Nomadix]

S9's for testing
https://postimg.cc/zLHhQxGQ

Testing tank got it from a friend that had no use for it as asics were overheating so doing some mods on it.
https://postlmg.cc/QK3c1csy

Immersion fluid
https://postimg.cc/HjnrvSv9

Arduino mini project to set immersion fluid safety trigger email above immersion fluid temps.
https://postimg.cc/F1yBqZf1

Original immersion tank pump being replaced with a home hot water circulator
https://postimg.cc/56HFyPw2

Original Immersion tank radiator vs new radiator
https://postimg.cc/Whf6fF7h

Main tank db prior to organization, 3 phase via single phase breakers rated at 20amps each for 6 asics
https://postimg.cc/YL3K2QtH

The main tank is aluminum as after testing found that the overall cooling with an alu tanked helped with dissipating the heat faster on the tank shell being in a cooled down room by hvac.

Will post pics of the main tank asics when I get some more pics.

The pics posted above is for a smaller setup to do some testing with units close at hand then apply the benefits to the main tank and units.

[mikeywith]

Main tank db prior to organization, 3 phase via single phase breakers rated at 20amps each for 6 asics

How are you getting 220v out of 3 phase? phase and neutral 220-240v phase to phase 380-400v?

Also, I think you mean a total of 2 gears per phase, so 10 amps for each miner, that will work just fine with S9s, but not with S19, you will need the whole 20amps breaker for one gear, in fact, if you push it towards the 6kw figure, even the 20 amp won't be enough, hope you are keeping this info in mind.

Great effort by the way, gave you more merit points for the efforts, I don't fully recall how the ranking system works, but I believe at this stage, you can post images directly on the forum by coping the .extension link and using the image icon above, here is an example of your breakers image

[Nomadix]

Main tank db prior to organization, 3 phase via single phase breakers rated at 20amps each for 6 asics

How are you getting 220v out of 3 phase? phase and neutral 220-240v phase to phase 380-400v?

Also, I think you mean a total of 2 gears per phase, so 10 amps for each miner, that will work just fine with S9s, but not with S19, you will need the whole 20amps breaker for one gear, in fact, if you push it towards the 6kw figure, even the 20 amp won't be enough, hope you are keeping this info in mind.

Great effort by the way, gave you more merit points for the efforts, I don't fully recall how the ranking system works, but I believe at this stage, you can post images directly on the forum by coping the .extension link and using the image icon above, here is an example of your breakers image

https://i.postimg.cc/66qyBjfL/IMG-20220513-WA0002-Small.jpg

Thank you for the upgrade... the local 3 phase is 415v.

Local electricity is UK std, the lowest breakers avail in the market start at 10amps. The s9's will be assigned 10amp breakers, the s19's are assigned 20amp breakers each.

Yes I checked on the image posting prior to posting I believe it is on ranking will keep climbing the ladder.. ;-)

[Nomadix]

Managed to get some time for a pic..

https://i.postimg.cc/BnRYM2Cb/asicbath.jpg

Very simple immersion tank setup works well...