Author

Topic: [Review/My Experience]With Bitmain's Second Hand (Used) Antminer S5 Batch 7 (Read 2436 times)

legendary
Activity: 1302
Merit: 1068
You state your best OC setting is 393.75, but the chart shows that 387.50 does more TH/s and less electricity used?, doesn't it?


They are 10 mins runs as labeled, so not super accurate. 393.75 does 1.3 after the hashrate average out properly. I have a single unit that does 400 for 1.32.

That review was based on my first S5, now i got a couple and i been running them for months. I use EVGA ATX PSU and at their voltage, 393.75 is the best. One of them takes 400 without a hitch.

They're possibly for sale, too.
sr. member
Activity: 338
Merit: 250
So two days ago, the Used Antminer S5 Batch 7 from Bitmain finally arrived. Me and several user were previously discussing the condition and value of buying used hardware for the full price. So this is for the few who asked question such as "Second Hand S5, Is it worth?" Wink

This will be a pseudo review on specifically the Used batch 7 Antminer S5 bought directly from Bitmain;

So the packaged arrived in the usual box it always does. I quickly opened it and found some very minimal packaging. No need to include a picture, it was simply 2 white foam square holding the extremities, with no other kind of padding or packaging whatsoever.

The unit was fairly dusty and there was clumps of dust in the heatsink. Hmm well, a bit of pressured air and as new. Or not quite, but i guess its not that bad.

2 Minutes later i finished placing it, plugging it and logging into the interface and replacing the Bitmain's worker info with Kanopool. Then flashed to Bitmain's latest firmware.

My first impression was Great, the open wrt Luci is very responsive (My S1 are hella slow).

I then first tested a few things, the 2 most important point i must bring;

-Fan control works, you set a %, no value seem to default to 80%~ Setting it to 65-70% made it at least quieter than my Zeus Thunder X3 which is honestly not that great. It makes the temp go up by 3c.
-The unit did not burn itself down when i starved it of an internet connection. Instead after 2-3 minutes, the unit's alarm start blaring. Then a minute after, and not before the hash boards powered down, then did the fan wind down.

Then i tested the clocks. On a EVGA G2 1000W, i did 10 minutes runs and filled a chart to get an idea;


During the tests, twice did the first hashboard not hash. (Guessed from its temp dropping to 38c)

After, i set it to run at Freq 375, i noticed twice that the first hashboard stop hashing and is picked back up one minute later. By looking at the graphs, this has happened 3 times in 24 hours. I tried 12 hours Kano pool and 12 hours Nicehash. Both gave me the same hashrate. 1.23THs Client side. 1.225THs pool side. Actual HWE is <0.0001%.

This seem to be similar to what some other users reported, while using the latest firmware. I don't really know what to think about this.

Pushing Limits;


After a week of running this baby, the best overclock value is 393.75. Up to that point, the unit conserve its J/GH efficiency, so i have to say the overclock potential and voltage control is excellent imo. It has the added bonus of having nearly no HWE% as well.

I also improved the temps by directing air from a box fan on and over the heatsink, but not force it down the heatsink, let the pressure blow over. Otherwise the temp would rise a bit instead of lower. I get 56 and 58C instead of 59 and 67 or such.

I am able to run the fan at 58% with temps at 61c~ at 30c~ ambiant. At this level the S5 is very similarly noisy as a Zeus. As in the dB is lower but the piercing high pitch noise meld with the lower tone of other ASIC.

You state your best OC setting is 393.75, but the chart shows that 387.50 does more TH/s and less electricity used?, doesn't it?
legendary
Activity: 1302
Merit: 1068
Yeah I think the s5 is normally shipped only with 2 white foam pieces..

 Of the 5 I've gotten, that's how 4 of them showed up. One had peanuts and some antistatic foam and was in a plastic bag, but that one I bought via Amazon used not from Bitmain.

 The "30 days starts when the units are shipped" part is actually illegal under US law - US code specifies that warrenttees start when the buyer recieves the item (I've BEEN to court once over that and WON, it's in the US Code somewhere) but for practical purposes it's not a significant issue given that Bitmain seems to get stuff out promptly once they claim it's shipped - 2 to 3 day difference that most folks would never notice.
 Just be sure if you DO have a warenttee claim that you get it filed promptly - once the claim is FILED there in no specific time limit on resolution.

 I suspect the coupons aren't going to be particularly large amount. I'd bet on $50 or less - which is still better than nothing when they FINALLY start selling S7s.



I won't spit on 50~$ and worse case scenario i can probably sell them to an user if i can't afford the next gen unit. Regardless. I look forward to it.

I'm thinking of opening a ticket with Bitmain about the instability of board one. Overall its not a big hash loss but if its precursor to more severe issue, then that would really bother me ^_^".

If its a stable issue that wont degrade, i don't mind too much. Its a big hash rate loss when it happen, but in the last 3 days it only happened once. It would definitively be a loss to ship it back or something no matter what.

Other than that, since my later tests that i didn't report, 393.75 is the best Freq. You can get 5 more GH~ at 400 but you start getting over <0.0001% errors and ultimately you get like 5 gh for 9watt~ of power.
legendary
Activity: 1498
Merit: 1030
Yeah I think the s5 is normally shipped only with 2 white foam pieces..

 Of the 5 I've gotten, that's how 4 of them showed up. One had peanuts and some antistatic foam and was in a plastic bag, but that one I bought via Amazon used not from Bitmain.

 The "30 days starts when the units are shipped" part is actually illegal under US law - US code specifies that warrenttees start when the buyer recieves the item (I've BEEN to court once over that and WON, it's in the US Code somewhere) but for practical purposes it's not a significant issue given that Bitmain seems to get stuff out promptly once they claim it's shipped - 2 to 3 day difference that most folks would never notice.
 Just be sure if you DO have a warenttee claim that you get it filed promptly - once the claim is FILED there in no specific time limit on resolution.

 I suspect the coupons aren't going to be particularly large amount. I'd bet on $50 or less - which is still better than nothing when they FINALLY start selling S7s.

legendary
Activity: 1302
Merit: 1068
I really care about baby penguins and polar bears.Bitmain is a good company.

About the used miners, every miner by every manufacturer are in use till they are delivered to the last user.30 day warranty is actually working and I feel comfortable to buy products from Bitmain.

Its not too bad its been running pretty well now. I did some longer tests pushing real OC on the machine till the machine no longer operate properly;

sr. member
Activity: 331
Merit: 250
I really care about baby penguins and polar bears.Bitmain is a good company.

About the used miners, every miner by every manufacturer are in use till they are delivered to the last user.30 day warranty is actually working and I feel comfortable to buy products from Bitmain.
legendary
Activity: 1302
Merit: 1068
No need to include a picture, it was simply 2 white foam square holding the extremities, with no other kind of padding or packaging whatsoever.
Is that a bad thing? You can work that out by asking the following questions:
1) When was the last time you heard of an S5 damaged in transit?
2) When was the last time you heard of a Tube / Prisma damaged in transit? [A miner that used copious packaging].
3) Do you like paying to ship unneeded packaging materials around the world?
4) Do you want all the baby penguins to die?

That's up to the next buyer to decide.

For me, personally the different units i have came in smaller/tigher boxes with the unit in an anti-static bag. Overall there's less packaging material. The sides are have a small layer of foam which serve arguable purpose beside balancing the pressure if the unit get thrown or dropped.

Doesn't sound particularly overkill to me and it is better packaging and less material weight, while the extremities get the same absorption rate.

Is that a bad thing? Probably not.
Was the packaging i received very barebone? Yes.
Would i have preferred a tighter and better packaging? Yes.

Yeah I think the s5 is normally shipped only with 2 white foam pieces.. well also with the bubble wrap bag but with these second hand units i can see why they dont include the bubble wrap.. probably just packaging them up on the fly really quick

Mhm. That is how it went. I do hope we get those coupons toward some other product down the road. It would make me fully satisfied, since ultimately i did pay full price for a slightly used unit, but used nontheless.
legendary
Activity: 1050
Merit: 1001
Yeah I think the s5 is normally shipped only with 2 white foam pieces.. well also with the bubble wrap bag but with these second hand units i can see why they dont include the bubble wrap.. probably just packaging them up on the fly really quick
legendary
Activity: 1666
Merit: 1185
dogiecoin.com
No need to include a picture, it was simply 2 white foam square holding the extremities, with no other kind of padding or packaging whatsoever.
Is that a bad thing? You can work that out by asking the following questions:
1) When was the last time you heard of an S5 damaged in transit?
2) When was the last time you heard of a Tube / Prisma damaged in transit? [A miner that used copious packaging].
3) Do you like paying to ship unneeded packaging materials around the world?
4) Do you want all the baby penguins to die?
legendary
Activity: 1302
Merit: 1068
So two days ago, the Used Antminer S5 Batch 7 from Bitmain finally arrived. Me and several user were previously discussing the condition and value of buying used hardware for the full price. So this is for the few who asked question such as "Second Hand S5, Is it worth?" Wink

This will be a pseudo review on specifically the Used batch 7 Antminer S5 bought directly from Bitmain;

So the packaged arrived in the usual box it always does. I quickly opened it and found some very minimal packaging. No need to include a picture, it was simply 2 white foam square holding the extremities, with no other kind of padding or packaging whatsoever.

The unit was fairly dusty and there was clumps of dust in the heatsink. Hmm well, a bit of pressured air and as new. Or not quite, but i guess its not that bad.

2 Minutes later i finished placing it, plugging it and logging into the interface and replacing the Bitmain's worker info with Kanopool. Then flashed to Bitmain's latest firmware.

My first impression was Great, the open wrt Luci is very responsive (My S1 are hella slow).

I then first tested a few things, the 2 most important point i must bring;

-Fan control works, you set a %, no value seem to default to 80%~ Setting it to 65-70% made it at least quieter than my Zeus Thunder X3 which is honestly not that great. It makes the temp go up by 3c.
-The unit did not burn itself down when i starved it of an internet connection. Instead after 2-3 minutes, the unit's alarm start blaring. Then a minute after, and not before the hash boards powered down, then did the fan wind down.

Then i tested the clocks. On a EVGA G2 1000W, i did 10 minutes runs and filled a chart to get an idea;


During the tests, twice did the first hashboard not hash. (Guessed from its temp dropping to 38c)

After, i set it to run at Freq 375, i noticed twice that the first hashboard stop hashing and is picked back up one minute later. By looking at the graphs, this has happened 3 times in 24 hours. I tried 12 hours Kano pool and 12 hours Nicehash. Both gave me the same hashrate. 1.23THs Client side. 1.225THs pool side. Actual HWE is <0.0001%.

This seem to be similar to what some other users reported, while using the latest firmware. I don't really know what to think about this.

Pushing Limits;


After a week of running this baby, the best overclock value is 393.75. Up to that point, the unit conserve its J/GH efficiency, so i have to say the overclock potential and voltage control is excellent imo. It has the added bonus of having nearly no HWE% as well.

I also improved the temps by directing air from a box fan on and over the heatsink, but not force it down the heatsink, let the pressure blow over. Otherwise the temp would rise a bit instead of lower. I get 56 and 58C instead of 59 and 67 or such.

I am able to run the fan at 58% with temps at 61c~ at 30c~ ambiant. At this level the S5 is very similarly noisy as a Zeus. As in the dB is lower but the piercing high pitch noise meld with the lower tone of other ASIC.
Jump to: