ANTMINER S3+ Discussion and Support Thread

[pekatete]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

[CHAOSiTEC]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

[pekatete]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

You've got both. Open them up and see the difference (the chips have numbers on them!) rather than waste every-body else's time with your circular question!

[chinesebob]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

You've got both. Open them up and see the difference (the chips have numbers on them!) rather than waste every-body else's time with your circular question!

As an outside observer I have to say, CHAOSiTECs question is and was very easy to understand.  pekatete didn't actually answer the question and then went straight to attacking CHAOSiTEC for no reason that I can see, and now seems to have digressed to trolling.

[CHAOSiTEC]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

You've got both. Open them up and see the difference (the chips have numbers on them!) rather than waste every-body else's time with your circular question!

im really starting to wonder who pissed you off today...

i am asking an engineering question, that i have been trying to figure out for a while, i have had both opened, but not spotted any visual differences, since it can be alot off different things, i am asking here, since, there might be someone who knows...

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

You've got both. Open them up and see the difference (the chips have numbers on them!) rather than waste every-body else's time with your circular question!

As an outside observer I have to say, CHAOSiTECs question is and was very easy to understand.  pekatete didn't actually answer the question and then went straight to attacking CHAOSiTEC for no reason that I can see, and now seems to have digressed to trolling.

[klondike_bar]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

Since pekatete seems inclined to be useless - here's my understanding:

I *beleive* the difference had something to do with the dc/dc regulators being either sub-par or of a lower rating in the s3. As a result, power to the chips was inconsistent at higher clockrates

[CHAOSiTEC]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

Since pekatete seems inclined to be useless - here's my understanding:

I *beleive* the difference had something to do with the dc/dc regulators being either sub-par or of a lower rating in the s3. As a result, power to the chips was inconsistent at higher clockrates

ahh, so in principle, i could change the dc/dc regulators and get the antminer to perform better

[pekatete]

you gave me possible solutions, but not what i asked for, that is the difference...

If you know what you are asking for, why ask in the first place?

omg.. really??

ok, asking everyone else ...

What was the engineering difference between S3 and S3+

Since pekatete seems inclined to be useless - here's my understanding:

I *beleive* the difference had something to do with the dc/dc regulators being either sub-par or of a lower rating in the s3. As a result, power to the chips was inconsistent at higher clockrates

No need to have a dig at me ... believe me. But that you did, and your suggestion of the dc/dc regulators, I suppose those should work well at lower freqs then .... and not cause the rig to drop off in speed .... or is that just an answer to provide room for name calling (or to make you look clever when you clearly are not)?

[pekatete]

As an outside observer I have to say, CHAOSiTECs question is and was very easy to understand.  pekatete didn't actually answer the question and then went straight to attacking CHAOSiTEC for no reason that I can see, and now seems to have digressed to trolling.

I can not see how that answers the question, so I'll add to it thus: If you have nothing to say, do not say it!

[chinesebob]

Interesting, so maybe the people messing with their voltage settings inadvertently stumbled on a workaround.  Also, maybe try a better regulated PSU, or run directly from a big SLA battery?

[CHAOSiTEC]

Interesting, so maybe the people messing with their voltage settings inadvertently stumbled on a workaround.  Also, maybe try a better regulated PSU, or run directly from a big SLA battery?

this one s3 is running with its own psu (rm850) which works fine on the other miners, also tried with delta 2000 same results

[chinesebob]

As an outside observer I have to say, CHAOSiTECs question is and was very easy to understand.  pekatete didn't actually answer the question and then went straight to attacking CHAOSiTEC for no reason that I can see, and now seems to have digressed to trolling.

I can not see how that answers the question, so I'll add to it thus: If you have nothing to say, do not say it!

I think you should take your own advice.  You're attacking people for no reason, I may be new to this forum but, am certainly not new to tech or bitcoin, you have no authority over me and thus cannot bark orders at me.  As usual there are a number of immature people on this forum that have a big head because they figured out something technical, as-if they've discovered something noone else could possibly do,  none of that gives you the right to yell at people you don't know.  Its highly likely that you and many of the people on this forum use some of the hardware and software that I've played a direct or indirect role in bringing to the world over the last couple decades, and I've probably held bitcoin for longer than most here but, regardless of the question, I'm not going to start attacking someone for not knowing something that I might already know.  I won't respond to you again.

[chinesebob]

Interesting, so maybe the people messing with their voltage settings inadvertently stumbled on a workaround.  Also, maybe try a better regulated PSU, or run directly from a big SLA battery?

this one s3 is running with its own psu (rm850) which works fine on the other miners, also tried with delta 2000 same results

If you do end up changing the regulators, please report the results as, I'm having the same issue with some of my S3s.

[pekatete]

As an outside observer I have to say, CHAOSiTECs question is and was very easy to understand.  pekatete didn't actually answer the question and then went straight to attacking CHAOSiTEC for no reason that I can see, and now seems to have digressed to trolling.

I can not see how that answers the question, so I'll add to it thus: If you have nothing to say, do not say it!

I think you should take your own advice.  You're attacking people for no reason, I may be new to this forum but, am certainly not new to tech or bitcoin, you have no authority over me and thus cannot bark orders at me.  As usual there are a number of immature people on this forum that have a big head because they figured out something technical, as-if they've discovered something noone else could possibly do,  none of that gives you the right to yell at people you don't know.  Its highly likely that you and many of the people on this forum use some of the hardware and software that I've played a direct or indirect role in bringing to the world over the last couple decades, and I've probably held bitcoin for longer than most here but, regardless of the question, I'm not going to start attacking someone for not knowing something that I might already know.  I won't respond to you again.

That is not an immature rant, it is just a rant. Grow up and snap out of your self importance!

[IITravel01]

not sure I trust the new firmware, all my GHS are all of a sudden down in the 100s per S3 three hours after the update

I woke up this morning an noticed on my pool 2 of the 3 S3's I firmware updated yesterday were only running around 50 Gh/s.  I went in and hit save and apply to the Advanced frequency controls and rebooted and they went back to normal.  Might have something to do if you over or underclocked the settings?

[Duce]

Interesting, so maybe the people messing with their voltage settings inadvertently stumbled on a workaround.  Also, maybe try a better regulated PSU, or run directly from a big SLA battery?

this one s3 is running with its own psu (rm850) which works fine on the other miners, also tried with delta 2000 same results

If you do end up changing the regulators, please report the results as, I'm having the same issue with some of my S3s.

I have been here since this discussion started as a B1/B2 owner and continue to read daily. A batch of regulator chips did not meet the specifications, though some did. Most of us in the beginning tried about all you could think of with various results. The voltage on some were lower, some did not like heat and others just did not want to work. What some of us also found was this degrade of speed. Various members came up with methods to restart the units at different intervals in an attempt to maintain speed, this did work to a degree. What others found was that if you just satisfy the chips your units came with you could get by with the speed they would produce that would run for weeks. Some of us ran units at lower speeds such as 212.5, at this frequency one my units will run non-stop with very low HW errors and stay within the BITMAIN spec of the advertised +/- speed. Other units could run higher than the S3 stock clock of 218.75.

As an engineer I understand the need to play around and try to fix the units to work faster, this is what we do. In some cases though it might not be worth the effort for the amount of return such as replacing chips. But again for some of us this is a hobby. I hope this sheds a little light on the subject, but from what I remember from this thread the difference between the S3 and S3+ seems to be a different batch of regulators.

[CHAOSiTEC]

Interesting, so maybe the people messing with their voltage settings inadvertently stumbled on a workaround.  Also, maybe try a better regulated PSU, or run directly from a big SLA battery?

this one s3 is running with its own psu (rm850) which works fine on the other miners, also tried with delta 2000 same results

If you do end up changing the regulators, please report the results as, I'm having the same issue with some of my S3s.

I have been here since this discussion started as a B1/B2 owner and continue to read daily. A batch of regulator chips did not meet the specifications, though some did. Most of us in the beginning tried about all you could think of with various results. The voltage on some were lower, some did not like heat and others just did not want to work. What some of us also found was this degrade of speed. Various members came up with methods to restart the units at different intervals in an attempt to maintain speed, this did work to a degree. What others found was that if you just satisfy the chips your units came with you could get by with the speed they would produce that would run for weeks. Some of us ran units at lower speeds such as 212.5, at this frequency one my units will run non-stop with very low HW errors and stay within the BITMAIN spec of the advertised +/- speed. Other units could run higher than the S3 stock clock of 218.75.

As an engineer I understand the need to play around and try to fix the units to work faster, this is what we do. In some cases though it might not be worth the effort for the amount of return such as replacing chips. But again for some of us this is a hobby. I hope this sheds a little light on the subject, but from what I remember from this thread the difference between the S3 and S3+ seems to be a different batch of regulators.

thanks, that was the information i was looking for, to me, its a hobby, same goes with undervolting, underclocking my antminer s1 blades  :-)

[SunnyIgor]

is it possible to have more than 3 pools on antminer s3?
and if so please show me how

anyone knows?

[chinesebob]

Interesting, so maybe the people messing with their voltage settings inadvertently stumbled on a workaround.  Also, maybe try a better regulated PSU, or run directly from a big SLA battery?

this one s3 is running with its own psu (rm850) which works fine on the other miners, also tried with delta 2000 same results

If you do end up changing the regulators, please report the results as, I'm having the same issue with some of my S3s.

I have been here since this discussion started as a B1/B2 owner and continue to read daily. A batch of regulator chips did not meet the specifications, though some did. Most of us in the beginning tried about all you could think of with various results. The voltage on some were lower, some did not like heat and others just did not want to work. What some of us also found was this degrade of speed. Various members came up with methods to restart the units at different intervals in an attempt to maintain speed, this did work to a degree. What others found was that if you just satisfy the chips your units came with you could get by with the speed they would produce that would run for weeks. Some of us ran units at lower speeds such as 212.5, at this frequency one my units will run non-stop with very low HW errors and stay within the BITMAIN spec of the advertised +/- speed. Other units could run higher than the S3 stock clock of 218.75.

As an engineer I understand the need to play around and try to fix the units to work faster, this is what we do. In some cases though it might not be worth the effort for the amount of return such as replacing chips. But again for some of us this is a hobby. I hope this sheds a little light on the subject, but from what I remember from this thread the difference between the S3 and S3+ seems to be a different batch of regulators.

thanks, that was the information i was looking for, to me, its a hobby, same goes with undervolting, underclocking my antminer s1 blades  :-)

The thing that doesn't make sense, with mine anyway, is that a simple restart of cgminer without rebooting or anything will instantly get the hash rate back up.  I would think if it were a heat issue they'd need a bit more time to cool down to get back to the initial performance level.

[Duce]

It never made sense, some problems seemed heat related and others didn't. There was a guy who had his units in a data center with the sides off and extra fans that had your same problem. Here is a quote from a PM he sent me July 27th:

 "Chip temps are 31 degrees, DC/DC modules are cool to the touch.
I do see 475+ for about 30-45 minutes on one unit but it tapers off below 430 after more than 10 hours.
Using server power supplies and a separate PWS for the fans. (They are 12V 2A each)"

I have one that did the same thing so I ended up just clocking it lower and it runs stable for weeks. Just need to find the sweet spot and let them run.