Author

Topic: ANTMINER S5: 1155GH(+OverClock Potential), In Stock $0.319/GH & 0.51W/GH - page 233. (Read 451048 times)

legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
The test unit also seems to cut out at 9.45V @ 200MHz even if I start at a higher voltage, and 9.75V @ 250MHz.

If anyone else has had any luck starting these at 10V or under I'd be interested in hearing it.

Interesting. How did you decide to use those frequency numbers? I would imagine that at a lower voltage it may be necessary to reduce the frequency at a lower rate.

I just made a matrix with frequency in 12.5Mhz steps and voltage in 0.5v steps, and measured voltage, current and hash rate.

can you dial down volts as the gear is running?

if you can try start at  freq 150 using  10.5 volts and dial down to 9.75 then 9.5 then 9.25 then 9 
legendary
Activity: 1456
Merit: 1000
I have been looking at miner link on the S5.   I could use a little help if possible.

Looking at it is the UID is your Bitmain email on account?

Also on Server Addr what should it be set to?

Thanks for any help.
legendary
Activity: 1274
Merit: 1004
The test unit also seems to cut out at 9.45V @ 200MHz even if I start at a higher voltage, and 9.75V @ 250MHz.

If anyone else has had any luck starting these at 10V or under I'd be interested in hearing it.

Interesting. How did you decide to use those frequency numbers? I would imagine that at a lower voltage it may be necessary to reduce the frequency at a lower rate.

I just made a matrix with frequency in 12.5Mhz steps and voltage in 0.5v steps, and measured voltage, current and hash rate.
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
The test unit also seems to cut out at 9.45V @ 200MHz even if I start at a higher voltage, and 9.75V @ 250MHz.

If anyone else has had any luck starting these at 10V or under I'd be interested in hearing it.

Interesting. How did you decide to use those frequency numbers? I would imagine that at a lower voltage it may be necessary to reduce the frequency at a lower rate.
hero member
Activity: 818
Merit: 508
I keep reading that these machines run warmer.  What is acceptable?  I currently have my miners in my basement and one of the windows is cracked down there.  Temp is around 67 degrees.  Since nobody spends time down there it's not an issue. 
sr. member
Activity: 338
Merit: 250
Here's my experience with the 2 S5's (batch 2) I received yesterday.  Got it unpacked and hooked up and turned on I thought "Oh, hell no!"  No way those fans can make this a home miner unless you're talking about keeping them in the Silence of the Lambs basement.  Switched to the Silverstone 141 fans on performance mode (1,920RPM is a loud woosh over the scream of the OEM fan) and put them in the back in a pull setup.  Turned on and they quickly shot up to 79 degrees Celsius, tried at 312.5M and they stayed at 79.  Turned them off and moved the fans to the front (fits fine if you use a bootless network cable) in a push setup, turned them on and they started going up past 70 again, so set the miners at 312.5M and waited.  They settled at 70-74 degrees C according to GUI, in a room that's about 75-80 degree F with 0 HW errors.  I can handle that, each is hashing about (1,025 GH/s avg.) that's about what my C1 did (1,005 GH/s avg.).  Need to get some readings with the Kill-A-Watt today to see how much power it's drawing at 2 x 312.5M.
 

Here's a little update, so at night the room with little ventilation would get too warm (in the 90's F) and the miners would get too hot, so I added a pair of these fans  http://www.amazon.com/gp/product/B0078IWSBG/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1  that run at 1,320RPM  to the rear in a push pull setup and it drops the temps on the S5's down 4 degrees C.  Wanted to go with another pair of the Silverstone 141's but the case they're in won't allow due to space so these fit and work well with the 141's in the front and these in the back.  

Upping the freq. to 325M on both brings the Gh/s avg. to about 1,070 ea. (about 1,050 avg. on the pool site) and the wattage at the wall on 230v to 1,100w with temp. at about 69 C (in a room at about 83 F).
legendary
Activity: 1666
Merit: 1185
dogiecoin.com
Do you have a Canadian reseller with known stock?
Great price,  but importing one will kill me

Resellers are unofficial, as anyone can be a reseller. Best bet is a having a good google to see who is listing in CA, then posting here so other members can backtrack or vouch for their legitimacy.
legendary
Activity: 1666
Merit: 1185
dogiecoin.com
Well, I can see why Bitmain struck the verbiage from their post about running on 9V. This is not authoritative by any stretch, as it's done on a single unit, but undervolting doesn't seem that effective here. I'll have to look into if I can share numbers (it wasn't one I bought for myself or was given as a review, they were purchased by someone for the sake of me testing them for them) but so far I've only gotten the S5 stable at 312.5 at 11.0V, and not stable much under that. If I start at 11V I can lower it to 10.75@300MHz, but I can't get it to fire up from being off at 10.75V at any frequency.

I wonder if Bitmain added that term based on chip efficiency and voltage in testing, and then just multiplied the voltage by 14 for their figures. Hopefully it's just balancing protections added in FW that can be solved later.

The test unit also seems to cut out at 9.45V @ 200MHz even if I start at a higher voltage, and 9.75V @ 250MHz.

If anyone else has had any luck starting these at 10V or under I'd be interested in hearing it.

I've poked the right people who can answer this, they'll be along shortly [after weekend]
legendary
Activity: 1652
Merit: 1067
Christian Antkow
If you have an older router load it with DDWRT      you can plug as many antminers into that router as there is ports.
Set up DDRWRT as a client bridge.... It will connect to your router. basically you could have a small farm run as far as the signal goes and not have to run a cable.
If the signal is good this works. if not you can increase the transmit signal on the router..
If I am too far I get drop outs but this is how I connect my S1 in the garage right now
+1 for DDWRT. I use this on all of my Linksys routers as wireless bridges for all the miners running in various rooms and the garage.

http://www.dd-wrt.com/site/index
legendary
Activity: 1204
Merit: 1000
The test unit also seems to cut out at 9.45V @ 200MHz even if I start at a higher voltage, and 9.75V @ 250MHz.

If anyone else has had any luck starting these at 10V or under I'd be interested in hearing it.

This is a little disappointing.
newbie
Activity: 22
Merit: 0
If you have an older router load it with DDWRT      you can plug as many antminers into that router as there is ports.

Set up DDRWRT as a client bridge.... It will connect to your router. basically you could have a small farm run as far as the signal goes and not have to run a cable.

If the signal is good this works. if not you can increase the transmit signal on the router..

If I am too far I get drop outs but this is how I connect my S1 in the garage right now
legendary
Activity: 3892
Merit: 4331
Can I run these using wifi or does net a direct connection to the router?
You can't.

You can run it in a place physically separated from your main router using a bridge.
For example, using airport extreme as a router, then bridge with airport express in another room. In this setup miner would be connected to the bridge either directly or through the ethernet hub, but you are effectively mining wirelessly. The advantage-you can do it pretty much anywhere in the house/apartment.
legendary
Activity: 2030
Merit: 1076
BTCLife.global participant
Can I run these using wifi or does net a direct connection to the router?
You can't.
sr. member
Activity: 457
Merit: 250
legendary
Activity: 1582
Merit: 1019
011110000110110101110010
Can I run these using wifi or does net a direct connection to the router?
legendary
Activity: 1022
Merit: 1010
I would suggest that the first thing to do is by hearing protectors. Wink

Haha, well we have the replacement fans ready to install. And these will be in a server room kept nice and chilly.
U
We will be selling off some of our B7 / B8 S3+ units to free up some PSUs.

Perfect condition, housed in Server Room at 64 F (approx) Hepa Air Handler. Clean and Like New.

I apologize if this is off topic but if there is any interest just shoot me a PM.

Strato

I am curious though. If these are in a server room, why tinker with the fans at all? Is there a noise concern in the room? I thought the issue with the S5 fan was strictly noise related.

The server room does have a high noise floor to begin with. But I was concerned these might be over the top db wise.

Strato

Also from what I have noticed with servers is that noise affects the performance of reading from and writing to the mechanical hard drives (SSD's are obviously unaffected by noise vibration as they have no moving parts). The noisier the environment, the slower they get.

BTW, is anyone able to compare the noise to an S4 please? I have an S4 and had to find another location for it. I had my neighbour knocking on my door asking what the hell the noise was and saying that he could not sleep with it (we both are in separate houses and his bedroom is about 50 meters away from my house). I have seen reports of the S4 between 70 and 90 db but not sure how accurate they were. It would be great if someone could do a test using the same metering device at the same distances for testing both.

Interesting Ive never heard that. All system critical drives were upgraded to SSD long ago. The only spinning drives left would be the large raid units, but they're housed in 8 FT Rack Cabinets, and the drives are all enterprise grade.

But yea SSDs are the way to go. I recently upgraded my workstation RAID to 16 Crucial 960 GB SSDs in two groups of 8, each on RAID 0. Im getting 5.50 GB/s read write performance between the two, and 2 GB/s transfers on our 10GBe Dual Network Cards.

Amazing how fast things have gotten.

I know this seems off topic, but just responding to your comments on drives... And thinking how in terms of speed and how fast the technology is getting... Bitcoin seems to be the lone ranger going in the opposite direction.

But perhaps thats the point.

Wink

Strato
full member
Activity: 175
Merit: 100
@Ilan123, I put 2 delta fans on and it is rather quite now, running at 54-56c with an RPM at 2880 for both fans when I last looked. I don't have the P/N with me right now though. My SP20's are about 70db and the S5 about 60db with the new fans. It was louder than my other units before the change but I did not take a reading. Others have the stock db level from the S5 units here though. I will get particulars later like ambient room temp and the P/N for the fans later today.
Fan: Delta AFC1212DE, 12vdc, 1.6A, 3900 RPM, 192.96CFM, 51 dBA, 120mm x120mm x 38mm or equivalent. These came out of some old Dell desktops, Dell P/N: Y4574. Right now room temp 26.1C, blades at 57-55 with the fans at 3K and 3120.

Here is a video with the delta fans installed on the S5. I am using my SP20s for reference. This video was taken with my phone.http://vid1284.photobucket.com/albums/a568/rponce454/MINER%20SOUND-Small_zpsyi3zjhtk.mp4
legendary
Activity: 1274
Merit: 1004
The test unit also seems to cut out at 9.45V @ 200MHz even if I start at a higher voltage, and 9.75V @ 250MHz.

If anyone else has had any luck starting these at 10V or under I'd be interested in hearing it.
legendary
Activity: 1274
Merit: 1004
Well, I can see why Bitmain struck the verbiage from their post about running on 9V. This is not authoritative by any stretch, as it's done on a single unit, but undervolting doesn't seem that effective here. I'll have to look into if I can share numbers (it wasn't one I bought for myself or was given as a review, they were purchased by someone for the sake of me testing them for them) but so far I've only gotten the S5 stable at 312.5 at 11.0V, and not stable much under that. If I start at 11V I can lower it to 10.75@300MHz, but I can't get it to fire up from being off at 10.75V at any frequency.

I wonder if Bitmain added that term based on chip efficiency and voltage in testing, and then just multiplied the voltage by 14 for their figures. Hopefully it's just balancing protections added in FW that can be solved later.
Jump to: