but if they are an "Authorized distributer" shouldn't the same rules apply? If not why?
They may. But - last time I looked at BM.ch (end of March) they only had pre-order, no 'reservation queue'. Is why most of their customers are in the same boat as those here over refunds. And Toots is right. You can cancel from their new queue only because you had not given BM.ch yer money yet. |
|
|
|
AMT,
Still waiting for the missing backplane and cable that you promised to ship me a week ago.
It has been 3 weeks now since I got my miner.... it has not been runniing because of missing parts.
Are you going to compensate me for the 3 weeks that I am unable to mine?
It may be an ass-backwards way to get them but have you tried PM'ing Phin to get the parts from his miner? Considering he does not mine, might actually be useful for that. Well, I pinged Bitmine, and they claim that they are sending me the backplane. Faboo! As for the GPIO cable, it is a standard and common item through Amazon, Digikey, etc. |
|
|
|
AMT,
Still waiting for the missing backplane and cable that you promised to ship me a week ago.
It has been 3 weeks now since I got my miner.... it has not been runniing because of missing parts.
Are you going to compensate me for the 3 weeks that I am unable to mine?
It may be an ass-backwards way to get them but have you tried PM'ing Phin to get the parts from his miner? Considering he does not mine, might actually be useful for that. |
|
|
|
AMT deleted all my posts earlier.
These frauds like to manipulate their image on these boards. Oh well, not a surprise.
Good luck everyone.
Same here. Good luck to all Hmm, could it be because this section is for hardware issues? Aside from not yet getting it (hardware) that is. While I too am without a miner from them I fully support their periodic purging of trolls. Same rules apply here as in any Forum - stay more or less OT or Those that Watch from Above, Below, and In-between shall striketh thee. There are plenty of bitchin' threads here for all who need space to vent. |
|
|
|
AMT, since you are in design mode for the 2-3TH rigs, any particular reason you haven't thought about feeding in 24vdc (or 48v)? In the multi-kw range much much cheaper per/w and hugely lowers the currents to 1/2-1/4 (48v) that connectors & traces have to handle.
Safe bet the core supply buck converters would be happier about it as well.
We're headed in that direction like most of the industry. +10 again! There's hope for this industry yet! Guess folks had to get out of the 'super PC' average consumer mindset and treat these for what they are no matter what chipset is used: very hungry power devices that just so happen to do math for coin mining/ BTC exchange verification. |
|
|
|
AMT, since you are in design mode for the 2-3TH rigs, any particular reason you haven't thought about feeding in 24vdc (or 48v)? In the multi-kw range much much cheaper per/w and hugely lowers the currents to 1/2-1/4 (48v) that connectors & traces have to handle. Safe bet the core supply buck converters would be happier about it as well. Oh - needless to say but, fusing. Ifin' yer gonna have a single high-current source then per-NEC guidelines you MUST limit the current per feed conductor to whatever the wiring/connector is rated for! (part of what sets the ATX PSU design guidelines) Best is either a lot of in-line fuses or a fuse panel along these lines http://www.ebay.com/bhp/12v-fuse-panelA 200+ amp supply feeding a short at the end of say even 16ga wire triplet is gonna make a nice fire... |
|
|
|
+10 for ya then! ^ ^ Along those lines, design tip for AMT: if the watts per GH/s of the A1's stays the same for the 2-3 TH/s rigs... Then using off-the shelf PC psu's made primarily for gamers is certainly not going to cut it. At simple scaling we are looking at what, 3-4kw? Better plan on spec'ing only 208-240vac in. Yes they can provide x number of kw but - the loading from graphics cards (doing video, not CUDA or other such things) is highly dynamic ranging from not too bad, say 50% or less up to hitting max for maybe several minutes before dropping again.. The PSU's currently used are designed with that in mind. opieum2, if ya gt a chance it would be nice to have some actual voltage measurements close to the PCIe plugs to see if/how much this is currently a valid concern.. Not the case with miners. Once miners get their rear in gear and start hashing it is a continuous load. No loafing times. 24x7x365. Ya' know. Like servers? http://www.trcelectronics.com/Meanwell/hot-swappable-rcp2000.shtml1.2kw per module would be a good start. I've seen others that are 3kw/module @12vdc and Platinum rated (96% efficiency) ta boot. edit: found the link to the 3kw modules http://www.artesyn.com/power/power-supplies/websheet/529/DS3000TE |
|
|
|
|
I still say the best solution for AMT & Bitmine would be to just use a REAL 1.5-2kw +12v server supply like in the link I've given further back. Bolt PCIe pwr cables to the buss lugs and then just a simple tab or other small regulator fed from the 12v rail to give the 5v needed for the Pi & backplane. Whole PSU issue solved.
|
|
|
|
Just curious, does anyone actually have a verified working 1.2T miner yet?
Mostly. Mine mines at 1.2Ths for a while but I noticed some power sags. My guess is that the PSU is not up to snuff despite the wattage on it. Each rail is getting fed around 20A-25A which may not be enough. The PSUs that AMT recommended have the following +3.3V@25A,+5V@25A,+12V1@110A,+12V2@50A, -12V@0.5A,+5VSB@3.5A noting the 2 12V rails being at 110A and 12V2 at 50A. This should resolve the issue. It appears that the AMT systems are volted differently than the stock coincraft machines which I think was mentioned in this thread or another one. So it might make sense why it would work like this. But as far as reliable and consistent mining I have not experienced that quite yet. I want to test with AMT's recommended PSU's before saying its 2 faulty miners. I have been working at this trial and error. If the PSU's work then I will get a 24/48/72 hour test out with stats from that. I will have them tomorrow so I can start posting some preliminary results then. I figure if I use AMTs recommended setup then there is no excuse why it should not work. The chips are being detected and the hardware is showing up...its just not hashing due to the power issue. Casey also had a similar problem and just added more problem and it worked for him. But because poserko loshia is withholding information and does not want to share his vast knowledge with the world, it appears we will never see them fixed /sarcasm Query: With all these unknowns regarding PSU's, has anyone yet actually measured the supplied voltage at the cards power connectors (under load of course)? The ATX spec calls for +12vdc +/-5% at the load for each output. That equates to a range of 11.4-12.6vdc. If you have a recording DVM even better so it can record any sags. Frankly with my background +/-5% is damn sloppy but the spec is what it is.(most industrial supplies are at worst +/-1% and usually more like a 1/10th of that...) If a rail/cabling can't support the load voltage will dip well below that range. eg, I had a Corsair CX600M (modular cables) power an Ant. Total load 'only' 410-415w and yet at the connector voltage was down to 11.15vdc. The cable was noticeably warm as well from the power drop. The non-modular CX600's are giving 11.6 at the connector because they use separate cabling for PCIe power |
|
|
|
Having the damaged hardware though they do need to RMA it and provide you with a working replacement. That is just the law.
@loshia you ask for stats but you dont actually provide useful information. All you do is say get pool stats and then say its a scam. You are the epic definition of a troll. Go fuck yourself. And your response in the other thread is a joke. If you "could have told you that" why the fuck didnt you? Because you are a useless human being who takes pleasure in other peoples pain. So go fuck yourself. At least I am helping people out here. That will bear out over your useless comments. You have already been called out by more people than just me. The other people who have gripes against AMT are at least helpful in providing some advice towards users. sirmineslaot usually provides some advice instead of just spouting off useless crap like you. You are simply just a weasel of a human being. Clearly useless. But I figure trolling you might be better. ANTI-TROLL MODE ON!!!! Who agrees here that loshia has contributed nothing useful to this thread besides just talking smack? Anything technical to help the users? All he asks for are stats.
+10 ^ ^ Shhh... Be vewy vewy quiet... We're hunting twolls  |
|
|
|
For those in the US at best if you don't recoup the cost or get a replacement for the hardware it can be written off as a business tax expense. You can claim it as a loss. I had already consulted with a CPA on this and it is possible to claim as a business loss. It isn't any consolation but its better than nothing at this point. Whatever you spent will just lower your taxable income at that point.
and it will hopefully offset what the IRS is gonna want when our BTC are spent/cashed in... |
|
|
|
I ran into a similar issue. And yea it seems to be a power issue. Least it is on my 1.2Ths systems. I experience similar problems on both of them. But the drop in hashrate is not alot. Looking at maybe 10-20ghs drops (which might be alot but in light of a 1.2Ths system is not too bad and still within a good range).
im a little worse off than that i have 3 boards 2 doing 204-205ghs and one does 155ghs  COULD be a power problem. But I dont know. How much are you running on the circuit the miner sits on? And what else is plugged into it? Your PSU or AMT's? If yours, make sure that each PCIe cable you are using will put out what is needed. As shown with the TH PSU's, not all rails deliver the same current. Look up the specs on the rails of the PSU to make sure. Tried swapping the good/low hash cars & power feeds? If so does the low hash card do better if in 'good card' slot/fed from it's 'good' power? And vica-versa as well - does a good card drop off? Still, at least meets advertised spec + a bit more and is less hungry than 2 of my OC'd Ants (837w for 2 running 190ghs each). |
|
|
|
I'd check Mouser. They sell the connectors/pins so proly sell the extractor. On Monday I'll see if the extractor I have for AMP circular connectors works. Pins look the same size. If you have some brass tubing that just fits over the pins would do the same thing. Just need to push the fingers in. As for crimps, use the same good ol' AMP terminal crimper that is used for all the different DB-xx connector pins. Oh, the Formfactors link for all psu's is http://www.formfactors.org/formfactor.asp |
|
|
|
Welp, ifin yer gonna do that, here are the official design specs that all ATX PSU's are supposed to follow. We're talking a-z here... http://www.formfactors.org/developer/specs/ATX12V_PSDG_2_2_public_br2.pdfAlso something to keep in mind is that there is a limit to the power any single output is supposed to be allowed to provide. Quote: The maximum short-circuit energy in any output shall not exceed 240 VA, per IEC 60950 requirements. |
|
|
|
|
The size looks rather like the 20 pin AT connector. Hmm. Seen a lot of PSU mobo connectors that have a block of 4 that can be removed, wonder if those cables can be used...
|
|
|
|
|
opieum2 - any chance on finding what they are using the 24-pin mobo connector for? The using the +5v and/or +3.3v feeds? Do they at least look at the POWER_OK line?
I'm still thinking of using a server supply with POL regulators fed from the 12v buss down to +5 or whatever. Will need to chop PCIe extender cables to tie to the buss bars of course.
|
|
|
|
|
+2 to both opie & freddy ^^ I'll find a source for the pin extraction tool for the PCIe/CPU connectors. Would eliminate the whole splicing thing. Odds are Mouser since they carry the connectors & headers, Molex Mini-fit Jr series.
|
|
|
|
<snip>
Need to rename this thread: Official AMT Hobbyist and Craft Fair Thread - "Send us $3000 and we'll send you a fun electronics builder kit (tools not included)."
Wouldn't rename this one but actually a far better catch-all thread could be the "Bitmine.ch/Coincraft A1 Systems Hobbyist Thread". On the technical side they are the root of this mess regardless of the actual systems maker. Us AMT customers are far from alone in this. Man talk about a lot bigger audience.... |
|
|
|
<facepalm> The chips are soldered to the boards. If you have a hot air smt rework station and are good at it, yes they can be removed but not necessarily reused as the pads will need careful cleaning after. |
|
|
|
|
Show Posts
