Lowering the noise level of the Antminer S5.
I've been running miners in a 1 bedroom condo, that I share with my girlfriend. I've placed them on top of our refrigerator, in the center of the condo. She works at home during the day, so it's important that they are not too loud.
When I hooked up the S5, I heard the loud whining sound that some people had posted about and knew I had to do something about it. I already had ordered 2 Delta fans off of Amazon, but they had yet to arrive. The were $22 each and are, Delta afb1212she-pwm 120mm x 38mm 4pin pwm+tac sensor extreme hi-speed 3700 rpm 151cfm case fan.
I had seen a post from ckolivas, where they found that the flow divider was the cause for much of the noise and found great improvement in using S3 fans instead of S5 fans. I removed the pull fans from 2 of my S3s and used them to replace the noisy fans on my 2 S5s. So now I have 2 S3s and 2 S5s all with only 1 S3 fan each in the push position. The S3 fan wire connector was not long enough, so I had to cut the end off of the S5 fan, strip the wire ends, and insert them into the S3 fan connector to make it work.
Results on the S3 are, the push fan runs a little faster, but overall quitter, and the machines are only 1 or 2 degrees warmer.
On the S5s, I down-clocked to 300mhz to keep the temps at or below 60C. The fans ran at about 3200RPM but were so much quitter that the stock fans. I was able to run the 2 S3s and the 2 S5s with a noise level that is bearable. Oh and as others have said, the pull fans don't seem to accomplish much as far as cooling, but definitely drive up the noise level.
When the delta fans got here, I tried them on the S5s instead of the S3 fans. The connector is 4 pin and the fans do self speed adjust, but I had to make a slight mod to the connector end to make it fit just right.
Results, 1 of the above delta fans on a S5 is slightly quitter than the S3 fan on the S5. It cools better and can run at a lower fan speed 2880RPM @ 300mhz, compared to 3200RPM @ 300mhz. You can use this delta fan to go back up to the 350mhz stock frequency and still stay at or below 60C in a room temperature room, with about the same noise as the S3 fan @ 300mhz.
So far I'm happy with the following setup.
1 - S5 with 1 of the above Delta fan in push position running at 300hmz mining at approx. 980-990gh/s
1 - S5 with 1 of the above Delta fan in push position running at 300hmz mining at approx. 980-990gh/s
1 - S3 with 1 S3 fan in push position running at 250mhz at approx. 500gh/s
1 - S3 with 1 S3 fan in push position running at 212.5mhz at approx. 425gh/s
1. see highlighted. You meant that you clip off one or both of the rails? 2. Did you try delta at 350 mhz on S5? If you did, I am not sure what the result was as far as dB is concerned (there are free android and iphone apps that measure this). I am already at 1000GH (337.5 mhz) with quiet (46 dB) Noctua/Scythe combo, so for me it wouldn't make sense to try Delta unless i can go to 350mhz or above. |
|
|
|
Anybody knows what kind of fan Sp20 has? 38mmX120X120mm or something else? I can look up myself, of course, but don't want to stop the hashing.
Thinking of putting either delta or Scythe ultra kaze in place of the stock fan; i know someone even put gentle typhoon, but i need to support just 1200gh/600-650w.
>snip image< I've used a Scythe ultra kaze but they are only 130CFM which is fine if you are underclocking. Any of the Scythe high CFM fans will work, the black high speed GT's or the even scythe kaze jyuni 1900rpm But yes its a 38x120mm fan but you can use a 25mm fan if you want. I've currently got the Ultra Kaze on one of my SP20's However as its a propietary connector to the daughterboard in the SP20, you either have to cut the connector off the stock fan and wire it up to the new one, or you have to connect it to the PSU directly like I did: >snip image< Thanks. Did it help? I am not sure what % of the sound (72db at fan 20 for the stock fan) is caused by the fan itself and how much is caused by moving the air through the metallic grille/sieve. The stock fan is supposed to be just 61dB, but SP20 produces way more than that (up to 72dB close by). |
|
|
|
|
Anybody knows what kind of fan Sp20 has? 38mmX120X120mm or something else? I can look up myself, of course, but don't want to stop the hashing.
Thinking of putting either delta or Scythe ultra kaze in place of the stock fan; i know someone even put gentle typhoon, but i need to support just 1200gh/600-650w.
|
|
|
|
put nicehash first on your list use a high price in the password spot on your pool section use p=0.0199 you will only run if price is over 0.0199 that is a very good price to mine at if price is lower you drop to your second choice on your pool list. you could use this one for second: https://www.f2pool.com/ this is paying 0.01097974 BTC / Thash/s that is about 0.2194 a day for 20 th or use my favorite : https://mining.bitcoinaffiliatenetwork.com/index.php this pays more then any other pool but can be slow to pay. Right. I added westhash (as I'm in the US) as my first pool, and add a high price per TH in my password so I don't miss opportunities to mine at a higher market rate. This way when crazy things like Payoin come along and boost the market price to 500+% your miner switches automatically, even if you are asleep! My regular primary pool goes in the second pool slot, and is used until my price threshold is exceeded at westhash. I hope this helps! yeah, i had that, but then miner periodically connected to westhash even when it was in third position (maybe it was on S5, not Sp20) and the whole setup got frozen once every 24 hr for some reason. Once i removed the westhash, everything was back to normal. Maybe just a coincidence. |
|
|
|
Just received a couple of Gentle Typhoon AP-31 (5400RPM) for testing, this is a high-air-flow fan for 120x25 size, it performs almost identical to Delta AFC1212DE at 20% speed
I've got the 4250rpm version on my SP20 for a few weeks now. Also tried the Scythe Kaze Ultra, which works quite well, but is also hard to find. Also from the same range the scythe kaze jyuni 1900rpm I've also tried, at 110CFM None let you run at the same temps as the stock fan, all are significantly quieter even at full speed. With a slight underclock they are fine. I got Scythe and would like to try. How did you connect it-to internal connector or to PSU through adapter? Thanks i don't think that it has PWM anyway, so it does not matter, but Scythe is just 46dB (3000rpm, 133 CFM). Surely, it beats the SP20 stock fan by being quieter. i am running SP20 at ~650W/1200GH anyway, so less sound would be great. |
|
|
|
The first half of 2015 could be very similar to 2012 and then we can expect a decent rise in the second half of 2015. We're definitely going to see some major turbulence in the world markets come 3rd and 4th quarter of 2015.
Buckle up.
In this scenario, VIX would be much higher and late 2015 options premium would be very high, but they are not. |
|
|
|
|
The only point of this is to plug his exchange where you could have 10-fold leverage which is equal to financial suicide.
Here is my analysis:
Bitcoin will either be worth a lot of money in 10 years or it will be worth close to nothing. Choose your scenario.
EVERYTHING is pump and dump: dollar,euro, gold, silver, stocks, bonds, currencies, real estate.
Oil dump, euro dump and dollar rise are the examples.
Conclusion: choose your "speculative" poison, but never tell me that your choice is rational while another person's is not.
|
|
|
|
Stock S5 is 74 db, its a lot! You need to change the fan to drop the sound. Delta seems to be the best one!  Holy cow, one fan makes that much noise. I was a little interested in picking one of these up, but a constant noise that high is a problem. Technology will take care of the electrical problem (how much they hog up), but what's going to take care of cooling these things silently. Whoever makes something like that will have a winner! the solution or solutions were posted on multiple threads, if you want to spend money on one or two extra fans (anywhere between $20 to $40). I went with Noctua 3000(push)/Scythe ultra kaze (pull)-good for at least 1100GH at ~58-59C and ONLY ~46dB;powering Scythe through molex directly off the PSU-piece of cake because it comes with an adapter; Noctua connector fits into the same slot as the stock (you have to clip one rail off the connector because it gets in the way). Phil likes two deltas because he found a good deal (two for $20 now, but you have to use an external additional power supply with voltage regulation or it will be louder than I like PLUS these have an unconventional connector). |
|
|
|
ahh the silverstones don't cool as well as the delta's. but they are pretty good. How were your deltas when powered from the miner or directly off the PSU at 12V as far as temp and noise is concerned? |
|
|
|
Did anybody tried to run SP30 more energy efficiently? If yes, care to share the settings and what the energy consumption/hashing speed will result?
I have my units in hosting, so cannot play with settings as easily as with SP20 at home.
A few that I tried, the green was the optimal until the recent price drop, now they are all offline, UK electric rates kill any chance of profit   Thanks, this will give me some options. |
|
|
|
I know this post is old, I am new to overlocking too.
I just picked up 4 of these miners very loud, so will be getting replacement fans.
I have free electricity, does anyone have a guide to how to best overclock these miners? Do I just turn the frequency up? What is the safest level to do so?
Thank you.
Basically, yes, in advanced settings as long as temperatures are at or below 60C (or <65C if you want to push it). Typically, replacement fans result in downclock, not overclock, but you save your eardrums-a pretty good tradeoff, in my opinion. |
|
|
|
|
Did anybody tried to run SP30 more energy efficiently? If yes, care to share the settings and what the energy consumption/hashing speed will result?
I have my units in hosting, so cannot play with settings as easily as with SP20 at home.
|
|
|
|
1. One relatively major currency had dropped more than 50% in a few months (rub), resulting in "evaporation" in ~$1 tril in GDP. 2. Another major currency changed hands 20% higher as a result of one (albeit major) announcement. I am talking about CHF/EUR of course. major brokers STOPPED trading the pair (maybe temporarily). http://forexmagnates.com/eurchf-crash-leads-brokers-to-halt-trading-in-chf-pairs/3. A major commodity (oil) lost 60% of its value in a few months without any recession being apparent. What does it tell you? To me it tells that a significant % of value of the above is determined by speculation and is NOT an intrinsic property. I fail to see how bitcoin is any different. At least we know the emission rate until 2140. |
|
|
|
I think that this is a slight exaggeration: maybe 10 fold first (for 48hr), then 3-4x for three days or so and back close to par on saturday (end of the POW period). Maybe it was 30X for a few hours, i did not get that. However, as a result, we empowered garza and his paycoin nonsense AND bitcoin declined more then 50%. this is how the timeline looks like. i know that this does not prove a casual link, but it is at least possible. If you think that this was not in any way influenced by paycoin-I say-we don't know it yet. It certainly gave garza ammunition to tout paycoin and argue against bitcoin. |
|
|
|
Just a quick update: it's about -1.985 now, while the all-time-low was -2.23. Still not in uncharted area. To break the ATL, the VWAP has to go below $173 today.
(This is based on bitfinex data as bitcoincharts.com has stopped reporting bitstamp data since the hack)
according to the OP formula calc, it went to -2.38 (new record) when BTC was $155 briefly. Right now -2.11 (at BTC=$203). I will wait for the double dip and if we will not go below $155, it might be a time for a longterm buy. Or not. Please note that the analysis is based on daily VWAP, not instantaneous minimum/maximum price. What price would a +1.5 be? it depends how fast it will go up, but if 2011-2013 was to repeat itself, which I don't expect, AND this trend sustaining itself (a major doubt as well), then probably at least roughly X100-125 from the bottom (see previous lows and highs) or to ~$19,000-20,000. I am not sure that we won't double dip to establish the bottom within <=1mo, but btc does not have to follow the trend as depicted here, but I wish that it will. |
|
|
|
Nothing goes up or down like BTC!
Loving it...never a dull moment around here these days
Except swiss frank and the euro, LOL. |
|
|
|
So I would like to ask.... Who is getting ready or who has turned their miner(s) off yet? SP30/31/35 specifically?
Curious to find out...
Thanks
Didn't quite turn mine off since they are all three all ROIED but I did underclock them to use 50% power SP10 or SP3x? I don't see how a SP3x could ROI yet even if you had free electricity and sold at the best times. My SP30 have only paid off around 25% of its value and I was one of the first to get them in the august GB.. To answer the question, I paid for six months hosting so it doesn't make sense to turn them off before that time has expired. p.s I have never mined BTC directly I alaways point to westhash and if I see a coin is raising the price of hash rentals I turn off westhash and mine the coin directly. Its the smartest thing to do. I missed the first two days of paycoin actually I could have made a ton more if I had been paying attention to what westhash was sending me for my rentals that Saturday and sunday. please elaborate a bit, i totally missed this westhash/nicehash/paycoin opportunity people just gave their hashing to paycoin by pointing miners to nicehash or westhash. some opportunity! |
|
|
|
typically you have cable going to the cable box and not the cat5 ethernet, then you have cable box connected via ethernet cable to the WAN port on the router, but router typically have more than one LAN port. Your PC could be connected to one LAN port and miner (or switch) could be connected to another LAN port. Not sure re splitter. |
|
|
|
Just a quick update: it's about -1.985 now, while the all-time-low was -2.23. Still not in uncharted area. To break the ATL, the VWAP has to go below $173 today.
(This is based on bitfinex data as bitcoincharts.com has stopped reporting bitstamp data since the hack)
according to the OP formula calc, it went to -2.38 (new record) when BTC was $155 briefly. Right now -2.11 (at BTC=$203). I will wait for the double dip and if we will not go below $155, it might be a time for a longterm buy. Or not. |
|
|
|
|
Show Posts
