With that being said, I would request a separate thread on voltage regulation. I recall Phil had mentioned, think it was s15 thread Norstar Power Converter with Voltage Stabilizer and a Larson Buck/Boost Step-Down Transformer as options but have seen a follow up as far as functionality or durablity. Of course this sorta should have it's own thread but - in reality few folks use AC voltage regulators. That said, ANY buck-boost transformer that is designed & intended for industrial use (like the ones from Larson) should last for at least several decades even when ran at 100% load. It's just the nature of the beasts.  Consumer types such as this are utter crap! Long long a go I tried a few different types and despite being ran at about 50% of their 'rated 3kw' load all crispy-fried themselves in about 1 month. Upon doing forensics it became obvious why: a. the core is too small causing excessive eddy current loses, b. the wire used in them is far too thin leading to even more losses and the windings getting very hot. As for regulators: Biggest thing is that they MUST use fully electronic control for adjusting the output voltage (like the Norstar's) or drive a variac to do it. Lower cost tap-switching units cannot be used as they all momentarily drop to 0v out during the switching process. Ja it is only for a few ms but nonetheless it's a safe bet that the miner PSU will have a problem with it and probably trip a reset. The only place I use a regulator is at home where the line can have substantial sag during the summer. For that I use Cyberpower's line of dual-conversion UPS's that run AC > DC > DC > AC. Their AC power is always derived using a inverter that directly adjusts the voltage (and freq) so there is zero switching time for voltage adjustment or switching to battery when the power drops out. |
|
|
|
|
^^Perhaps.
However I also see a technical need for less THs per-miner and the resulting less power draw per-machine: namely what happens when a miner goes down.
The fact that 1 miner down can now mean up to what, around 75THs or more lost? On the PSU side, keeping the total output to say (with margin) 1,600w makes it easier and cheaper to make than one that has to pump out 3-4kw. Certainly easier to cool. Those multiple tiny fans scream....
To me, unless you are running several hundred PH or more, I'd much rather have say 20-30THs per miner drawing 1,500w or less. Sure, you need more of them but unless they are uber-reliable like Avalons have historically been then I'd much prefer a miner failure to take out a lower % of the total farm hash rate.
The side benefit is that it keeps power draw to something that a 'home' miner can easily provide if they decide to mine regardless of electric cost.
|
|
|
|
|
Should add that another reason for PCIe power plugs melting is folks using the same PSU's to power a few consecutive generations of miners. The connector socket pins are rated for surprisingly few mate/unmate cycles before the plating is scraped off so repeated plugging/unplugging of the connectors is very bad. Lose the plating > higher resistance > more heating > remaining tin plating begins to oxidize and eventually things avalanche to failure. So to be on the safe side, if the PCIe connectors on the PSU have seen a lot of use - then re-pin them.
Oh, and how few cycles you ask? I've seen manufacturer ratings for mating lifetimes as short as 10 cycles for el-cheapo socket pins. One does have to remember that these connectors were designed to be very seldomly plugged/unplugged as most of the time the connections are made only once in the lifetime of normal computer equipment.
|
|
|
|
a. Bitcoin can ONLY be mined using AISC-based hardware. Ref https://bitcointalk.org/index.php?topic=2959621.msg30383956#msg30383956 points 2 and 3 b. Only Altcoins can be mined with CPU's and this (BTC mining) was the wrong area of the Forum to discuss that. Use the Altcoin areas which is where the mods moved your post as I was replying. b. Mining (altcoins) using Azure or other cloud-based CPU platform is not a crime however, it most likely violates the Terms of Service and that is why you keep getting banned. c. I take it you have never bothered to ask MS Azure support folks why they are banning you? Pretty certain they will be more than happy tell you... |
|
|
|
Merit given for reporting back that you found the solution! |
|
|
|
|
@OP, who is "we" Kemosabe? Would be better if this topic was titled "Why some People Think Craig Wright is Satoshi and Why That Matters".
Personally I think CW is a self-promoting asshat that just like to make pointless noise (and certainly is not Satoshi).
|
|
|
|
 After finding our last block of 2019, here's to looking forward to another great year with more blocks awaiting us Mine on folks!
|
|
|
|
In the article about the 5nm dev tools, they say: TSMC says that when compared to N7 (1st Gen 7 nm, DUV-only), N5 technology will allow chip developers to shrink die area of their designs by ~45%, making transistor density ~1.8x higher. It will also increase frequency by 15% (at the same complexity and power) or reduce power consumption by 20% power reduction (at the same frequency and complexity).
Since they will be essentially using the same logical chip designs their current chips use (same complexity) it will be interesting to see what direction BM & Canaan take: Higher speed at same power or same speed at lower power. As for the cost of the 1st 5nm miners out-the-door: no doubt cost will be insane $$$$$$$$$$$$$$$$$$$ |
|
|
|
|
Depends on just *how* old those miners are. The C5 were used from the batch-1 up to maybe batch-5 or so. From batch-6 on up you should be able to use the newer xilynx boards. If the hash boards have the big power inductor on them by the PCIe plugs you should be good.
It's the batch-1 through 5 that might not work because the early batches do not have the on-board Vcore regulator for the chips. Those early miners just directly used the PSU to power the chips. The later miners have the Vcore regulator on each board that allows you to set the voltage fed to the chips. The probable problem would be how the firmware handles no regulators.
|
|
|
|
but to add insult I got a notice from the state of Illinois for 8% sales tax. Ja, that is definite gotcha with Formal Customs entries: CBP automatically notifies your state about your purchase and if you do not pay the sales tax (as you are supposed to with all out-of-state purchases depending on your states rules) the state will definitely eventually contact you about it. With informal entries the state usually does not know about the purchase but -- depending on how lucky you feel you still should declare the purchase(s) come tax time or face the chance the state might find out later and come after you for it. |
|
|
|
of course I gave them my TIN - I dont have issues with doing so. My curiosity is still about the Tariff - did it change? are they wrong for saying 2500?
and yes I called UPS and asked for this person directly - got them and provided the information.
AFAIK nothing changed. The under $800 is related to the Trump tax being added to possible duty charges. The $2500.49 is the threshold for requiring Formal entry (paperwork and supporting documents as well as a base line duty %). Under that is informal entry with no duty and very minimal paperwork. The section 301 they refer to is related to the Trump Tax in that certain electronic items like miners, TV's, computers, etc., are now required to pay the minimal duty even though the shipment value is under $2500. |
|
|
|
As the 1st one to receive the 0.1 BTC block finders bonus that Kano implemented, gotta say Faboo!  |
|
|
|
Hallelujah! Yep, was one of the remaining 841's I have running in a 63THs cluster in the farm at work   |
|
|
|
|
re: eth connection & GUI being flakey
If you mean sometimes you cannot log in or switching to Log confuses it, that started with the A10's which use Canaan's new controller chip & FMS.
Sometimes when logging in you will get 'Connection Refused'. Just keep reloading/retrying. Sometimes a fast double-click on Log In will do it as well.
For Connection Refused when trying to access the Log, just reload it or go back and try it again.
|
|
|
|
While I stopped running BM miners in favor of ones from Canaan a few years ago, I gotta say good job! This should make many many folks very happy |
|
|
|
Ideally anyone running production code involving computers that handle money (even if the code itself doesn’t), should review any libraries, fully understand what it is doing before importing them. I would also hope they wouldn’t use any code period that relies on downloading content from an unaffiliated third party as what these libraries were doing.
Spot on target. Far too many BTC devs seem to not care that in the end miner, proxy & pool code are FINANCIAL software and need to be vetted as such. People not only can but HAVE lost money due to bugs in it that testing would have easily caught. In other words: test the hell out of it BEFORE going live! Ya listening Slush/Braiins devs? |
|
|
|
@ Phil, bet all we get from them is the sound of crickets amidst the silence....  |
|
|
|
|
@KriptoKyng, once you have registered with the Kanopool Discord group, bookmark the page and use that to get to the group. Do not keep using the invite link as it is for 1st time use only and will not give you full access.
|
|
|
|
So i got my S17 5 days ago, it didnt instantly work, because i was getting "Hardware version: Socket connect failed: Connection refused" error, which resolved itself, after running for 5-10min, is this normal? Yes it is normal for ALL Antminers since the s9 batch-15 when BM started using an Autotune function for the miners to test themselves and decide how fast they can run. You will get that message for as long as autotune is running. Depending on a lot of things autotune can run for several min to over 45min before the miner starts hashing. The one good thing is that while they are testing themselves you can enter user & pool info and restart. Then once they start hashing they will be doing so pointed at your pool vs BM's. |
|
|
|
|
Show Posts
