klondike_bar
Legendary
Offline
Activity: 2128
Merit: 1005
ASIC Wannabe
|
|
August 20, 2015, 07:49:16 PM |
|
so... x27 6pin pcie cables per miner.... im planning to use 3 hp 1200w power supply per miner... but.. 9 cables per power supply .. this seems hard to do... anyone knows where to purchase those 6 pin pcie cables awg 16.. ? cough. look at my sig
|
|
|
|
klondike_bar
Legendary
Offline
Activity: 2128
Merit: 1005
ASIC Wannabe
|
|
August 20, 2015, 07:53:19 PM |
|
ALL the 3 PCIE connectors are needed to be connected to power supply on each hash board .... since it is based on serial power solution and there is no DC/DC inside the miner.
Those two statements are unconnected and don't make sense as a justification. If each board can still only be powered by one PSU due to cross-loading, then each board's 3 PCI-E connectors are still connected. While we don't have any high res images to see if we can see tracks, each of the PCI-Es is still in-line and *appears* to be powering the same plain. That agrees with the cross-loading argument which would mean that each PCI-E is the same as the next, meaning the only limitation is maximum load and not because it doesn't have DC/DC. i tend to agree with this. if each board is drawing <400W you could provide that with only 2 cables. my guess is that it does two things: 1) prevents people from burning up cheap 18awg cables (that are limited around 150W) 2) slightly better balance of power distribution so the connector or copper traces dont fail
|
|
|
|
dogie
Legendary
Offline
Activity: 1666
Merit: 1185
dogiecoin.com
|
|
August 20, 2015, 08:34:53 PM |
|
ALL the 3 PCIE connectors are needed to be connected to power supply on each hash board .... since it is based on serial power solution and there is no DC/DC inside the miner.
Those two statements are unconnected and don't make sense as a justification. If each board can still only be powered by one PSU due to cross-loading, then each board's 3 PCI-E connectors are still connected. While we don't have any high res images to see if we can see tracks, each of the PCI-Es is still in-line and *appears* to be powering the same plain. That agrees with the cross-loading argument which would mean that each PCI-E is the same as the next, meaning the only limitation is maximum load and not because it doesn't have DC/DC. i tend to agree with this. if each board is drawing <400W you could provide that with only 2 cables. my guess is that it does two things: 1) prevents people from burning up cheap 18awg cables (that are limited around 150W) 2) slightly better balance of power distribution so the connector or copper traces dont fail I absolutely understand that the official line will always be use all 3, however the technical explanation as to why you must use all 3 didn't really make sense.
|
|
|
|
edonkey
Legendary
Offline
Activity: 1150
Merit: 1004
|
|
August 20, 2015, 09:42:43 PM |
|
tend to agree with this. if each board is drawing <400W you could provide that with only 2 cables. my guess is that it does two things: 1) prevents people from burning up cheap 18awg cables (that are limited around 150W) 2) slightly better balance of power distribution so the connector or copper traces dont fail
I absolutely understand that the official line will always be use all 3, however the technical explanation as to why you must use all 3 didn't really make sense. Maybe someone who has an S5+ can tone out each side of the power connector pins. I think it's that simple, right? If they're all connected with no resistance, then they're all going to the same place. That would mean the pragmatic answer is you can probably use 2 16 AWG cables.
|
Was I helpful? BTC: 3G1Ubof5u8K9iJkM8We2f3amYZgGVdvpHr
|
|
|
fullzero
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 20, 2015, 10:10:30 PM |
|
I have an S5+. I am running it on default freq with 3x evga 1300 PSU.
With the PSU's I am using; each board uses 373 watts and the controller uses 185 watts for a total of 3542 watts.
I don't think anyone would necessarily have a problem running off of only 2x connectors on each blade so long as the PSU and cables they are using are very high quality and gauge. However;
I also don't see the point in not connecting all three. The miner costs ~ 9.35BTC. Cables and connectors are nothing compared to this.
|
|
|
|
notlist3d
Legendary
Offline
Activity: 1456
Merit: 1000
|
|
August 21, 2015, 02:40:51 AM |
|
I have an S5+. I am running it on default freq with 3x evga 1300 PSU.
With the PSU's I am using; each board uses 373 watts and the controller uses 185 watts for a total of 3542 watts.
I don't think anyone would necessarily have a problem running off of only 2x connectors on each blade so long as the PSU and cables they are using are very high quality and gauge. However;
I also don't see the point in not connecting all three. The miner costs ~ 9.35BTC. Cables and connectors are nothing compared to this.
That is kinda my feeling. I can see why people with PSU's with less PCIe connectors want 2 per. But Bitmain made a big deal about 3. If you have a problem and it's obvious it only had 2 plugs I think they will void warranty. So that one extra plug on each to keep warranty valid is worth it to me.
|
|
|
|
sebdude420
|
|
August 21, 2015, 04:18:44 AM |
|
what an abomination.
|
OG Bitcoin Miner turned Proof of Stake Validator. Maxed out Raspberry Pi 4 8GB at 120$ a Day Revenue with ~15K XTZ Bonds in Summer of 2021. Looking at Proof of Stake systems all across the crypto ecosystem to expand operations.
|
|
|
mavericklm
|
|
August 21, 2015, 05:31:51 AM |
|
maybe they know how cheap the connectors are on the boards..... made in china...
|
|
|
|
notlist3d
Legendary
Offline
Activity: 1456
Merit: 1000
|
|
August 21, 2015, 05:37:23 AM |
|
what an abomination.
It is actually a very nice miner from my experience with it so far. It is not for everyone with the amount of electricity needed. And also you do need a lot of PCIe cables I don't deny this. But the recommended psu works great with them. So far 1 day and 7 hours running not a single issue and running up to spec. So actually looks good as far as miner stand point.
|
|
|
|
yslyung
Legendary
Offline
Activity: 1500
Merit: 1002
Mine Mine Mine
|
|
August 21, 2015, 09:24:54 AM |
|
what an abomination.
It is actually a very nice miner from my experience with it so far. It is not for everyone with the amount of electricity needed. And also you do need a lot of PCIe cables I don't deny this. But the recommended psu works great with them. So far 1 day and 7 hours running not a single issue and running up to spec. So actually looks good as far as miner stand point. wut's your ambient temp & miner chip temp ?
|
|
|
|
fullzero
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 21, 2015, 05:48:24 PM |
|
My highest temp is 44 C the ambient temperature is 25 C.
I have almost no errors. Pool reports slightly higher hashrates than the miner itself.
The 1x heatsink on each side of every chip really increases the heat dissipation. It also makes the design dependent on constant airflow.
After my warranty ends I will try using a fan controller and testing how low the fans can be set while keeping the chips at a reasonable temperature.
|
|
|
|
notlist3d
Legendary
Offline
Activity: 1456
Merit: 1000
|
|
August 21, 2015, 10:54:14 PM |
|
what an abomination.
It is actually a very nice miner from my experience with it so far. It is not for everyone with the amount of electricity needed. And also you do need a lot of PCIe cables I don't deny this. But the recommended psu works great with them. So far 1 day and 7 hours running not a single issue and running up to spec. So actually looks good as far as miner stand point. wut's your ambient temp & miner chip temp ? My ambient temp is around around 78F which is kinda mild. I think I have made it past most of the hotter day's in summer. I have a lot of fans in my main mining room to help pushing new air. One hashing blade is hitting 57c but most are hitting low 50's. Very low errors and preforming as spec says.
|
|
|
|
sloopy
|
|
August 23, 2015, 05:15:09 AM |
|
I previously checked on the site for S5+ manual and firmware, but they are present now.
The manual makes more sense regarding the control board power connection, kind of... It states to connect it to any power supply, but it is still worded oddly. I am going to assume it must not matter, but I am at a loss why they say anything about it other than to connect it to a power supply unless there is something they are trying to tell us.
If anyone has any ideas what they are truly getting at feel free to post up.
Regarding the control board connection the manual states: "...And be sure to connect the hash boards to the PSU first and then connect the controller to PSU" "Connect connector on control board to any power supply"
If you are connecting the controller to the same PSU as hash boards, why does it matter if it is first or last, it will get power at the same time... They have stated this several times, so there must be something there.
and just because I love to repeat things that make me contemplate the mysterious statements in life I will repeat this from the manual as well: "Higher input voltage will cause higher mining efficiency"
So how high can someone go?
I see improvements in this manual over previous generations. I appreciate there is a manual for this model. I hope we see some clarification on the points made because whatever issue (if there is one) they are getting at with the control board power confuses the hell out of me. I previously thought that maybe they wanted the hash boards powered up slightly before the control board, but the way the manual reads with connecting to any power supply that can't be it and besides you don't want to be fiddling around getting that powered on with any delay when that many chips are putting out heat -- less fans.
I have seen an unfinished heatsink. It isn't cut out all the way. I will get a picture "from the guy" and get it posted. Overall it looks like a good unit. IT needs to be tested on a couple of other pools and ck solo to collect more data by "the guy" before anything lengthy is posted.
|
Transaction fees go to the pools and the pools decide to pay them to the miners. Anything else, including off-chain solutions are stealing and not the way Bitcoin was intended to function. Make the block size set by the pool. Pool = miners and they get the choice.
|
|
|
tennozer
|
|
August 23, 2015, 06:30:06 AM |
|
My highest temp is 44 C the ambient temperature is 25 C.
I have almost no errors. Pool reports slightly higher hashrates than the miner itself.
The 1x heatsink on each side of every chip really increases the heat dissipation. It also makes the design dependent on constant airflow.
After my warranty ends I will try using a fan controller and testing how low the fans can be set while keeping the chips at a reasonable temperature.
Did you change any fans on miner and where do you keep your miner? Air circulation looks like enough
|
|
|
|
notlist3d
Legendary
Offline
Activity: 1456
Merit: 1000
|
|
August 23, 2015, 07:05:06 AM |
|
I previously checked on the site for S5+ manual and firmware, but they are present now.
The manual makes more sense regarding the control board power connection, kind of... It states to connect it to any power supply, but it is still worded oddly. I am going to assume it must not matter, but I am at a loss why they say anything about it other than to connect it to a power supply unless there is something they are trying to tell us.
If anyone has any ideas what they are truly getting at feel free to post up.
Regarding the control board connection the manual states: "...And be sure to connect the hash boards to the PSU first and then connect the controller to PSU" "Connect connector on control board to any power supply"
If you are connecting the controller to the same PSU as hash boards, why does it matter if it is first or last, it will get power at the same time... They have stated this several times, so there must be something there.
and just because I love to repeat things that make me contemplate the mysterious statements in life I will repeat this from the manual as well: "Higher input voltage will cause higher mining efficiency"
So how high can someone go?
I see improvements in this manual over previous generations. I appreciate there is a manual for this model. I hope we see some clarification on the points made because whatever issue (if there is one) they are getting at with the control board power confuses the hell out of me. I previously thought that maybe they wanted the hash boards powered up slightly before the control board, but the way the manual reads with connecting to any power supply that can't be it and besides you don't want to be fiddling around getting that powered on with any delay when that many chips are putting out heat -- less fans.
I have seen an unfinished heatsink. It isn't cut out all the way. I will get a picture "from the guy" and get it posted. Overall it looks like a good unit. IT needs to be tested on a couple of other pools and ck solo to collect more data by "the guy" before anything lengthy is posted.
As far as the power I think you are worrying a little to much on it. I think they designed controller unit to look for blades being powered on, so blades being powered on is a good thing. If for some reason it the controller is booted up first and didnt work just reboot it I have had 0 issues with mine. As far as power getting more efficient not sure. 220/240 is more efficient then 110/120. But I suspect almost all have it on 240 and not multiple 120's. But not sure how far they are thinking. What do you mean by you have seen a unfinished heatsink can you take a picture?
|
|
|
|
tonycamp
Member
Offline
Activity: 84
Merit: 10
|
|
August 23, 2015, 07:08:25 AM |
|
well if world was perfect into 4 months the roi will come and some eletrecity costs too
|
|
|
|
sloopy
|
|
August 23, 2015, 10:13:08 AM |
|
I previously checked on the site for S5+ manual and firmware, but they are present now.
The manual makes more sense regarding the control board power connection, kind of... It states to connect it to any power supply, but it is still worded oddly. I am going to assume it must not matter, but I am at a loss why they say anything about it other than to connect it to a power supply unless there is something they are trying to tell us.
If anyone has any ideas what they are truly getting at feel free to post up.
Regarding the control board connection the manual states: "...And be sure to connect the hash boards to the PSU first and then connect the controller to PSU" "Connect connector on control board to any power supply"
If you are connecting the controller to the same PSU as hash boards, why does it matter if it is first or last, it will get power at the same time... They have stated this several times, so there must be something there.
and just because I love to repeat things that make me contemplate the mysterious statements in life I will repeat this from the manual as well: "Higher input voltage will cause higher mining efficiency"
So how high can someone go?
I see improvements in this manual over previous generations. I appreciate there is a manual for this model. I hope we see some clarification on the points made because whatever issue (if there is one) they are getting at with the control board power confuses the hell out of me. I previously thought that maybe they wanted the hash boards powered up slightly before the control board, but the way the manual reads with connecting to any power supply that can't be it and besides you don't want to be fiddling around getting that powered on with any delay when that many chips are putting out heat -- less fans.
I have seen an unfinished heatsink. It isn't cut out all the way. I will get a picture "from the guy" and get it posted. Overall it looks like a good unit. IT needs to be tested on a couple of other pools and ck solo to collect more data by "the guy" before anything lengthy is posted.
As far as the power I think you are worrying a little to much on it. I think they designed controller unit to look for blades being powered on, so blades being powered on is a good thing. If for some reason it the controller is booted up first and didnt work just reboot it I have had 0 issues with mine. As far as power getting more efficient not sure. 220/240 is more efficient then 110/120. But I suspect almost all have it on 240 and not multiple 120's. But not sure how far they are thinking. What do you mean by you have seen a unfinished heatsink can you take a picture? Yeah, I have a pic and will put it up tomorrow... err later today. It is a heatsink which wasn't completed but it was installed anyway. I agree regarding the power. When they make such a big deal about something by putting a different blurb about it in every announcement, manual, etc then it would be great to have some clarification. I've e-mailed them to ask. Things like that make me want to know why. Part of playing with stuff like this for me. I've seen 0 problems other than a couple of minor things. As long as it hashes well there is certainly nothing to complain about. The speed and density are fantastic and it runs much cooler than an S4 in the same space and ambient temps.
|
Transaction fees go to the pools and the pools decide to pay them to the miners. Anything else, including off-chain solutions are stealing and not the way Bitcoin was intended to function. Make the block size set by the pool. Pool = miners and they get the choice.
|
|
|
edonkey
Legendary
Offline
Activity: 1150
Merit: 1004
|
|
August 24, 2015, 02:22:13 PM Last edit: August 24, 2015, 03:02:48 PM by edonkey |
|
EDIT: Looks like batch 2 is shipping.
I wanted one of these before the new chip announcement. While I'm still tempted (bird in the hand), I'm probably going to wait for the S7.
I take it back. This is probably a scam. See my next post on this thread.
|
Was I helpful? BTC: 3G1Ubof5u8K9iJkM8We2f3amYZgGVdvpHr
|
|
|
edonkey
Legendary
Offline
Activity: 1150
Merit: 1004
|
|
August 24, 2015, 03:00:36 PM Last edit: August 24, 2015, 03:49:03 PM by edonkey |
|
Last batch everyone: SCAM LINK REMOVED
Yeah, I saw that link this morning too, but now I think it's a scam. Note that the text of the link above says "bitmaintech.com", but the actual domain is "bitmaintech.co". Looks like "bitmaintech.co" was registered on Friday August 21st, 2015. The main bitmaintech.com site says nothing about batch 2. I first saw notice of the last batch on reddit, but the link to the article was also likely a scam. The news article link on reddit was "bitcoinmaqazine.com" (note the 'q'), not bitcoinmagazine.com. Be careful. This is probably a scam. The question is did dailey123 intentionally obscure the above link, or was it an innocent mistake?
|
Was I helpful? BTC: 3G1Ubof5u8K9iJkM8We2f3amYZgGVdvpHr
|
|
|
dusca
|
|
August 24, 2015, 03:11:33 PM Last edit: August 24, 2015, 03:25:19 PM by dusca |
|
|
|
|
|
|