Any chance of you getting 8 pin cables?
I dont see much point in an 8pin cable - no ASIC hardware uses them, and most GPUs use only 6pin anyhow. The extra 2 pins are both black GND wires and AFAIK the GPUs with 8pin sockets will run fine with only a 6pin cable The poster was no doubt referring to the 8-pin EPS +12v cables used on A1/A2 miners based on Innosilicon's boards eg. 'Dragons'. They are designed for supplemental power to CPU's with 4 +12v and 4 returns. http://www.moddiy.com/pages/Power-Supply-Connectors-%26-Pinouts.htmlThose are not the 6+2 PCIe GPU variants with the 2 extra returns in them. And yes, I'd be interested in a few of them as well... |
|
|
|
Skin effect has nothing to do with 50-60Hz power and is not a significant effect until several kHz.
As for capacity of stranded vs solid, depends on how it us um, gauged (pun intended). AWG when applied to electrical wire is not a hard number, there is a fair bit of tolerance in the unit when applied to stranded wire. Comparing capacity of wires exactly the same diameter in inches or mm and depending on strand thickness (changes fill factor) yes solid technically can carry more current from DC up to several kHz. That is why in many parts of the world and for larger cables the actual equivalent copper cross section of a wire or cable is used for ratings.
With reference to AC which is what we were discussing... Not sure where you have your degree from, but well, perhaps you need to edit all the books and websites that tell a different, lets call it story ... http://en.wikipedia.org/wiki/Skin_effectTo quote from that link "At 60 Hz in copper, the skin depth is about 8.5 mm." That is 0.335" or around the same as the diameter of 4ga. cable making it entirely a non-issue in this case which is my point. |
|
|
|
Believe what they wanted to say was 16 AWG wire, not 16 amp?
I believe the same. Don't think so: 10 AWG is generally thought of as 30 amp, 12 AWG as 20 amp, 14 AWG as 15amp, and 16 AWG as 10amp. Cassey This is true when dealing with solid single-conductor wires meant for permanent installation in walls and ceilings and whatnot. However most all wall-to-appliance power cables use multi-strand wire for flexibility, which muddies the water somewhat. A multi-strand "wire" cannot carry as much current as a solid single-conductor wire of the same gauge can. With AC that is an incredibly incorrect statement. Google AC skin effect. Skin effect has nothing to do with 50-60Hz power and is not a significant effect until several kHz. As for capacity of stranded vs solid, depends on how it us um, gauged (pun intended). AWG when applied to electrical wire is not a hard number, there is a fair bit of tolerance in the unit when applied to stranded wire. Comparing capacity of wires exactly the same diameter in inches or mm and depending on strand thickness (changes fill factor) yes solid technically can carry more current from DC up to several kHz. That is why in many parts of the world and for larger cables the actual equivalent copper cross section of a wire or cable is used for ratings. |
|
|
|
At 1m I get 72 on both my professional meter and my phone for a max RPM S4. If you're getting 90 then you're ramming it into the fan and the airstream which is wrong.
I'm about a 1.5 metres away. Iphone on my desk with mic facing the S4. I have 3 of them running at close to max fan rpm between 3960-4200. Also, I'm in a rather small home office, so might be the reason too. Can't turn the fan down unless I can get cool air. This is as cool as it can get indoors without the air conditional. I can't turn it on anyway, as it can't keep up even though it's 12000 BTU. Edit: Oh and with the huminity levels here in Singapore, evaporative cooling doesn't work. 12kBTU = 3.5kw heat load and is only barely good for 2 s4's. When it comes to cooling miners forget window A/C units. The math: BTU/hr=Watts x 3.4129 1 Tonne=12,000 BTU/hr |
|
|
|
ya know... this sounds a lot like how Bitmine.ch screwed with AMT on the first A1 attempts... Never told them how to save to the Dpot NVM so it always defaulted to mid-point (way too high). Think they just never write to that chains Dpot or hybrid regulator address to save start value? temps across the chains range from #1 @ 41c, 2 @ 42c, #3 @ 45c and chain 4 @ 44c Hmm. S4 #1 is set to 200mHz with Core set to 725. I'll look at work tomorrow to see what the other S4 is set to. Should be as-shipped. As to conflicts between SPI and i2c coms - they are supposed to co-exist... ja when things get really fast is best to interrupt the hash coms to set/query the regulators but... should only need to do that on power-up before the ASICS start hashing or if otherwise queried. still sounds like someone dropped the ball on setting a value to startup NVM memory vs the working segments. Just my 2-cents... EDIT: Just checked the other s4 and it is using same settings of 200mHz clock and 0725 for core voltage. That 290w power difference between the 2 units really bothers me...  |
|
|
|
For various reasons I've added a rather nice always online UPS http://www.cyberpowersystems.com/products/ups-systems/smart-app-ups/paragon-series/ol2200rtxl2u.html?selectedTabId=specifications&imageI=#tab-box to my 2nd s4 and found something rather interesting: My 1st s4 on it's own same UPS reports a very significantly higher power draw. Like 290Watts more. Both s4 are Batch-2, the first one possibly was made a week earlier earlier than #2 but both arrived within 1 day of each other (direct vs hosted shipping goof). The UPS's report #1 1,480 to 1,490w load #2 1,100 to 1,200w load. Both consistently reporting as 1.950-2.25THs by my pool. Anywho, in some earlier threads it was mentioned that on some Batch-2, one of the ASIC chains can't/doesn't change Vcore and pretty much runs near max safe voltage. Wonder if that is the case here? Certainly would explain the 290w higher load... I take it that during power-up when the Vcore regulators are either set from a pre-stored value in a Dpot or set by the I2C link one chain is not getting the message... Any way to fix that by Firmware? |
|
|
|
Gotta love natural Canadian cooling.  And it isn't even winter yet. If you're wondering, temps are showing 0 because S4's temp sensors don't register below 20c. Your HW Error is rather high.. and do you realise you have have a X on one of the chips? Hard reset for that. I had one and a soft reset just makes it worse  Talking about that... my chain 1 goes offline after about 1-2 days running, sometimes in a matter of few hours. GUI will be reporting as if nothing wrong. You'll only notice when your hash rate at pool end drops. You can also look into the casing via the fans, you should be able to see 4 red leds if all chains are running. Was thinking of taking the S4 to a host but with me constantly having to look to see if there is 4 red leds, not going to be possible. Plus having to hard reset when you get "Hugs & Kisses".. makes it even more impossible to have the S4 hosted. Could it be a firmware issue or hardware? Hmm wonder if that is what happened to me. After running flawlessly for 22 days on Sunday one of my S4's dropped to only 1.2-1.3ths, checked the gui, all chains reporting good and still have CK's fix applied so wasn't a power glitch causing a full reboot. Did a soft reboot, reapplied CK's tweak and came right back up to full speed. I do still suspect power so have an always on-line ups coming for it just to be sure. So far it is only the S2 and S4 Beaglebone based miners that have been needing the ups's, S1 & 3's seem to handle power dips just fine If it happens again I'll check the activity lights. |
|
|
|
What I was curious about is how long it takes for payment to be confirmed and they send an email to me? I already received an email telling me I have one hour to make my purchase before cancelation. I completed that purchase about 2 minutes after receiving that email.
Within a few hours, usually. Don't worry, as long as you sent the correct amount of BTC to the address they provided, you'll be fine. Okay, sweeeeeeeeeet Thanks for the reply so I won't be pulling my hair out. Thank you very much! It's a lot to send to someone in another country. David It sure is, and I remember that very same feeling the first time I made my first BITMAIN order, but in a few days you'll have a nice shiny S4 at your door and you won't think twice about ever buying from anyone but BITMAIN. For all the bitching I've done about not receiving my compensation yet, I can't deny that BITMAINTECH.COM (or any authorized reselllers, like SUSHI) is without a doubt the way to go. You will never want to waste your time or money buying rigs on ebay again, especially when 95% of the bitcoin rigs on ebay are way overpriced, and typically have always been used (sometimes even damaged or DOA). ditto. My first 4 Ants were bought on Amazon via Jones Gear back in March for minimal markup over direct from Bitmain but gave nice fuzzy warm feeling of Buyers Protection (um, seriously burned by Joshoua Zipkin / AMT Miners). However asy my mined BTC rose said what the heck and went direct. Now 24 various Ants of all models 'cept the USB toys later gotta say, Bitmain is the only way to go. Spondoolies also good but are hairdryer loud by nature. |
|
|
|
Might want to refrain from including the images in your quotes, these pages are getting stretched LOL. Yeah, that's the extension. Unfortunately I was wrong, that isn't a L6-30P connector. Looks like a L14-30 or something. Do they have a L6-30P version? That's what you need. You might want to take a look here: https://en.wikipedia.org/wiki/NEMA_connectorShould give you a good idea. So, the L14-30 that is 120/240 is not good enough? It has to be just 240? Not 120/240? Remember I'm posting from a cell LOL... The original socket you posted:  Should work. You'd only use 3 of the 4 pins, one ground two live. Just make sure the extension is L14-30P. 125/250 doesn't matter, it just means it can legally be wired for both. I'd try to find L6-30R + L6-30P if you can, though. Local Ace or True value hardware as well as Home Depot, Lowes should have them. From McMaster, these are very nice http://www.mcmaster.com/#7174k18/=ujkq67matching plug http://www.mcmaster.com/#7162k53/=ujkru9 |
|
|
|
Just want to update to say that Amt finally sent my other miner.... For some reason it doesn't seem to talk to the router. Issue with the controller board? What do you guys think?
Thanks,
Max
If it came with a network cable my 1st thought would be to get a different cable. The one that came with mine was intermittent depending on how far the plug was inserted: All the way in, no connection. Slightly in and enough to latch sometimes good. Any lights lit on the network socket (assuming the socket has lights)? Miner and router on the same address block (192.168.1.xxx)? No lights for the network socket on the router. Already tried a few cables. Can't ping the miner because the router isn't seeing it. Looking at mine, just to the right of the RasPi connections for the LAN and USB cables there is a slot where you should see some lights on the RasPi board, 2 are the usual yellow & green ones for for the network link and should also be a red one seen. Obviously since no link the yel & green will be out, is the red one lit? If it is not lit then the RasPi isn't running. Check to see if it has power and a SD card in it. |
|
|
|
That is why I was thinking the cloud hashing. Buys hashing power with dirty money, the cloud hashing spits a stream of clean money.
Still is not 'clean' though no doubt Josh would think it is  The 'dirty' transactions are funding the cloud mining. Any gains derived from it is considered as tainted as well. Mining output is just another link in the chain. And a nice repeatable patterned one at that. |
|
|
|
|
that was my 1st thought as well (mining). Hasn't KNC moved to cloud hashing and gotten out of hardware? I seem to recall most of their very unhappy latest rig customers getting cloud instead...
|
|
|
|
I don't believe you can do a hard enough reboot via SSH that will reawaken X'ed chips. PSU restart it is.
Well, that sucks... How come this software https://bitcointalk.org/index.php?topic=596178.0 can correct the X's for the S1, S2, and S3 from a reboot, but not S4? If the S1, S2, and S3 are able to correct the X's through SSH, why wouldn't something like that be available for the S4? Unless there's some hidden command we don't know about. Probably because unlike the earlier models the s4 has programmable Vcore regulators whereas the others uses fixed resistors to set it. Along with OC'ing it have you played with using a higher core voltage? Generally an ASICS voltage needs will track what freq you are running at - lower speed and you can reduce Vcore, higher - raise it. Vcore is set only when power is first turned on and not during any sort of soft(ware) boot. As said previously, a hard boot is killing power and restarting. Actually just to clarify, there are 3 different PCBs of S4s which have slightly different powering methods. Some have programmable Vcore, some don't. I'm not certain if the 'have to hard hard reset' applies to all three revisions. different control pcb's in them or different versions of the hashboards? The control boards should have fixed local regulators on them to create the 5v the Beaglebone needs along with whatever the ASIC interfaces need to be (3.3 & 1.1v?) Those voltages never need to change, when playin, er 'optimizing' just Vcore on the hashboards should be adjusted. Now as to resets... based on forensics I did on the disastrous 1st version AMT/Bitmine.ch A1 boards Vcore is set only when power is first applied to the hashboard. Each board has its own independent programmable regulator that come alive as soon as the main +12v is applied to the board. Its setting is stored internally to the regulator. Looking at the limited info on the Bitmain design and others rigs out there seems everyone does that part the same way. While the Beagle can modify it on-the-fly so one could test before saving the setting don't know Bitmain is allowing that. I'm sure the Beaglebone has a reset pin on it, someone just needs to read up on it to tell us where it is on the board... As to if that would also force a reset of the ASICS - no idea. |
|
|
|
I don't believe you can do a hard enough reboot via SSH that will reawaken X'ed chips. PSU restart it is.
Well, that sucks... How come this software https://bitcointalk.org/index.php?topic=596178.0 can correct the X's for the S1, S2, and S3 from a reboot, but not S4? If the S1, S2, and S3 are able to correct the X's through SSH, why wouldn't something like that be available for the S4? Unless there's some hidden command we don't know about. Probably because unlike the earlier models the s4 has programmable Vcore regulators whereas the others uses fixed resistors to set it. Along with OC'ing it have you played with using a higher core voltage? Generally an ASICS voltage needs will track what freq you are running at - lower speed and you can reduce Vcore, higher - raise it. Vcore is set only when power is first turned on and not during any sort of soft(ware) boot. As said previously, a hard boot is killing power and restarting. |
|
|
|
Just want to update to say that Amt finally sent my other miner.... For some reason it doesn't seem to talk to the router. Issue with the controller board? What do you guys think?
Thanks,
Max
If it came with a network cable my 1st thought would be to get a different cable. The one that came with mine was intermittent depending on how far the plug was inserted: All the way in, no connection. Slightly in and enough to latch sometimes good. Any lights lit on the network socket (assuming the socket has lights)? Miner and router on the same address block (192.168.1.xxx)? |
|
|
|
How do we do something like a power cycle to reset the O's and X's on the chips through SSH?
For example, sometimes I see X's on the chips and when I reboot through the GUI, it doesn't get corrected. It's only when I manually do a power cycle by powering it off and on from the power button is when all the chips turn back to O's on the status page.
The reason I am asking is because sometimes I get X's and i reboot it from the GUI remotely through a VPN and it doesn't get fixed. I don't want to have to keep going there and manually power cycle it every time because my S4 is located about an hour away and it's a hassle =(.
Any ideas?
I've spent some time thinking this through, and haven't come up with a great solution. "shutdown -h now" won't clear the x's, the only thing I've found is killing the power at the PSU and bringing it back up, which is difficult to pull off without being in the DC. Maybe someone know a trick we don't? As a general any ASIC mining chip can only be fully reset by cycling the power off/on. To remotely cycle power get something like this http://www.wifiplug.co.uk/ |
|
|
|
Is anyone else experiencing an issue using ck's s4 firmware where the S4 powers off after approximately 5 hours? I'm not even sure that the issue is a result of ck's firmware, but as i mentioned earlier in this thread I mine exclusively in the eligius pool, so if I want to achieve a hashrate > or = 2 th/s, it is necessary for me to use his firmware which seems to work excellently with Vardiff pools that don't let you set minimum worker difficulty...
Edit: I guess a better question might be- has anyone had a similar issue with their S4 powering off after approximately 4.5-5 hours (regardless of firmware)? If you have, what firmware are you using?
Is anyone using ck's s4 firmware for their S4 and has been able to run 24/7 without any issues?
Running perfect here using CK's tweaked firmware. My 2 s4's have been going 24x7 for the past 2 weeks with zero issues. Any fault lights lit by the plug on the psu? Using the right power cord that has at least 16ga wire in it? 14ga would be ideal. Per my UPS the s4 pulls 12.6A with 120v feed. A 'normal' cord is 18ga and definitely not a good idea as they are only rated to 8 amps and aside from getting rather warm will also produce a large voltage drop between the wall and the psu... |
|
|
|
A friend of mine kept having problems with his computer power supplies. He was using standard surge protector, but not a UPS. Eventually, after I talked him into it, he bought an APC UPS. And because of the voltage monitoring and protection, he found out that his voltage went up and down quite a bit. The UPS had to intervene often to normalize the voltage input. So, he found out the fluctuation in voltage was causing the problem. I'm not trying to say this is the reason for any of the reported problems on this forum. I'm only passing along information for future reference. Food for thought...
What you are saying is: Buy a good PSU!!! Right!? Well, that is definitely a good thing, but not what I was saying. The house wiring had a problem. But the problem did not trip a breaker or a surge protector. However, the problem was caught by a battery backup (UPS, Uninterruptible Power Supply). I'm pointing out one of the benefits of a UPS is to ensure the power line is "clean". It's too expensive to use UPS's for the miners, however I have all of my computers and networking equipment using UPS. A UPS (agreed) is overkill for a miner, but voltage protection for such devices is just as important. For my miners, I use a APC LE1200 voltage regulator. Most _decent_ UPSes have this capability built-in, but provides the voltage at a constant level while not being a UPS (with all the hassle of one). Clean power without all the UPS baggage? Check. Extra power plugs for multiple devices? Check. Offset plug for those fscking volt-on-a-ropes? Check. Does it work? Double Check. Good for all your electronics (TVs, computers, etc)? Check. http://www.amazon.com/gp/product/B00009RA60/ref=oh_aui_search_detailpage?ie=UTF8&psc=1Because not all of us have decent power to our homes. Hmm that looks interesting since the Amazon link does not give the important particulars (and I'm too lazy to look them up) do you know the switching time on voltage changes? I *do* use several large on-line UPS's from Cyberpower http://www.cyberpowersystems.com/products/ups-systems/smart-app-ups/paragon-series/ol2200rtxl2u.html for my s2's and s4's - the normal standby ones with AVR use relays to change the voltage taps which glitches the line long enough to kill an S2/4. |
|
|
|
|
Show Posts
