Topic: New Official AMT Thread - page 91. (Read 149472 times)

Well not exactly. The cards will work even if one in the chain fails. My comment was more on the chips on the card itself. The backplane thing still sort of a daisy chain tho its not a dependent on the previous card in the chain being connected.

One clue that thanks to ISAWHIM made me consider its a software issue, is the fact that the cards actually COULD come back up after numerous reboots. Almost a chance event...except its not. It just means that there was just enough of a delay for the system to detect the card during the HW detection process. Also with the "dead" cards, it could just be a hardware problem. It could also be that the are taking longer than they should to come up. BUT might be able to revive them. So we got two fronts now to address these problems (your welcome AMT) Software and Hardware. More than likely the hardware side will yield better results, but the software at least can compensate for some of the other glitches as well.  

yes we are. Start at https://bitcointalksearch.org/topic/m.6754740
they are physically identical packages aside from the numbers on them. Someone screwed up badly there with what reel got loaded into the stuffing machine.

Those are the references I use They also have complete info 2-chip reference board including the gerber files for making a pcb and parts list for it. Whether or not the reference design works... who knows but it's all we have to go by.

You can't be serious here.

That actually would be great news if there is a way around it. I got my soldering iron unpacked in the event I need to use it

Hopefully even with that I can bring back the one card that is preventing an entire setup from coming up.

I will post results on the other cards tonight and see if a software method of bringing them up is possible. Either way we get a new firmware to work with.

The pidora based firmware I am using has the following:

screen
cgminer 4.3.3
And all the dev tools needed to compile packages from source.

SSH is also built in as well.

I have not added any web component at this point until I have it working (thats a later feature)

If it brings the cards back then project zombie will be a success!

Oh... the cards are daisy chained?  So if one card fails, the others don't work?

Anyway for chips,  read this:  http://bitmine.ch/wp-content/uploads/2013/11/CoinCraft-A1.pdf

and look at this wiring diagram:  https://github.com/bitmine-ch/bitmine

It looks to me like they are daisy chained.

BTW, anyone here done any PCB manufacturing and SMD reflow sodering?  You think we can have some of these dual boards created?

In the 2-chip reference design on github the chips themselves are daisy-chained for the SPI coms. Goes in one side and comes out the other. If you pull a chip I'd guess that you just need to place 2 jumpers on the boards pads to pass the data on to the next in line. Again, don't have any hardware here much less the new boards layout/schematic to verify that.

We will have to hear back on whatever isawhim's idea is with the not-responding board. It does seem that some of the local decoupling caps are missing/erroneously changed with resistors.

The traces we were talking about are 2 different things, one being the boards power-planes being shorted by a screw. In that case - or a chip shorting - then the single Vcore regulator they use for each board is going to try and feed all the power it can into the short until it is burned open. If the short is to sturdy to burn open it is going to kill the whole card and possibly also cause the PSU to shut down.

My ruminations are about the A1 chips internal power routing: There are 7 power pads feeding in Vdd: 4 very large ones, 2 smaller ones in 2 corners plus one more very small one. If for whatever reason all the big Vdd pads are not connected from bad soldering or not properly bypassed (missing caps) then it is very possible that connections inside the chip which are still getting power could overload and open up killing the chip. Methinks isawhim is going to try and get around that.

That is what I don't know. I know the cards are daisy chained on the backplane. But the chips on the card is what I am not sure of. Without any further info or documentation on the design we have no idea. I will need to take a closer look as I actually haven't. I probably will later after I get some time to sit down for a while and work on it.

Not even stressing it now. writing it was enough. If they want to communicate (or send the replacement hardware) they know how to. I am more than willing to bury the hatchet for my part if they were to just keep me in the loop on replacing the hardware with tested working units. That for me at this point would be enough to move on from this now. My frustration is mostly centered around the fact that I get a communication with options and I reply and then nothing. I imagine the lawsuit has them up their ears in crap, but they did state they were running business as usual. I know both legal parties are viewing all this anyway. My loyalty is to the truth of the matter.

I will say this....I still don't think they are/were a scam. A scam is really just a grab and run operation. They do have/had a product, regardless of quality and execution. Their handling of the business side of things as well as their over-promise under-deliver as well as lack of effective communication is what led to this whole mess for them. Bitmine is dealing with similar things as well (to a lesser extent). All said and done not a scam IMO....just poorly run is all. This would be a shitty scam if they didn't have an easy exit strategy to get people's money. And they don't. They actually are still around. Hopefully out of all this we get working miners with proper QA done on them or compensated for our losses.

Oh.... I had thought the chips were daisy chained.   You mean they are not?  You mean we can individually remove chips from the board and it will still work? 

Is that correct?   

Getting back to technical.....so is the issue with the cards blowing a hardware problem then? The traces shorting or some other component causing problems? I would presume a single chip blowing would still allow the rest of the card to function unless there was some kind of daisy chain that cuts the electrical flow.

Looks like a couple of my comments got deleted. I was pretty pissed that day over the lack of replies and had a moment. Oh well. I get inconsistent responses anyway after trying to work with them. As an FYI AMT put an offer out to get me two new miners two replace the two fried ones I have. I have not seen or heard anything since that offer. I accepted it even. I don't have any word on status of this. But I have given up trying to communicate as I, like everyone else is getting nowhere.

Since they are deleting days old posts they might want to be aware that the mirror hosts these posts as well and those cannot be removed. Plus I have a record of the removed posts in my PM box so I can always repost them. I likely wont though as its not worth it and I was pissed off when I wrote those (not that it makes it any less wrong what I wrote). Anyway in spite of that I am still working to solve what could be a software problem with these miners not just for me but for others. It might benefit AMT (if they are still in business after the lawsuit) as well the work we are all doing here, but ultimately we are doing it for the people who got screwed with faulty hardware. Those of us all working to get the miners working at least are working together to help each other out. We are not likely to see any compensation from AMT for that effort anyway.

I still maintain they were legit but screwed up royally, but just had alot of bad situations and fucked up royally on the communication. But I figure its just shitty planning that got them here and allot of false assumptions and not keeping on top of regulations and things like that. I don't totally think they were looking to screw people over but the shitty planning that led them to this mess just got worse and worse and was unmanageable. I said from the start if they were able to just properly communicate and set proper expectations none of this would even be an issue. Sure the lawsuit would have happened likely anyway as they were dragging things out. But the first problem was lying (or misleading us) about expectations and dates. The truth would have been just fine. Hey guys we have problems with IMET, we also have problems with bitmine we are sorting out....compensation coming....that would have helped to some extent. But if the goal was to not give compensation this is why they are where they are at now.

Anyway thought I would comment on my deleted posts.

We are all automatically included in it. The big question is whether we will come out of it or not "whole" If AMT doesn't have much we are stuck with the short end of it. It likely will be the case IMO no matter how its sliced. I guess it depends on what there is available for us. I didn't put in any claim. I think they got plenty from everything being said here. And at this point or even before it was irrelevant.

Man that 1st sentence was a long one...
Anywho, at least you got *something* even if it is a paperweight. I also ordered a 520, In my case I only went with that from AMT because the 520 was not advertised as pre-order. It was falsely advertised as 'delivery in 4 weeks or less' implying they were in production...  

I even asked for a DIY when they made the offer. Too bad they didn't follow through as I along with opieum2, ISAWHIM maybe a few others here are ideal folks to get the kits so we can figure out what is wrong with them and come up with fixes. Or - certify design & manufacturing issues caused by their contract manufacturer. Oh well, AMT's loss. And mine  

You're not the only one bro, they screwed many people like you and me..

Its beyond a joke pal i really wish i had never come across amt, worst decision of my life honestly im normally good at watching out for dodgy deals and scammers but this time i got caught out, 

Thanks for the link pal good look with your claim

Here you go: http://www.chimicles.com/lawsuit-filed-against-advanced-mining-technology-for-failure-to-deliver-bitcoin-miners

Also its likely they are watching this thread anyway. Seems to be alluded to in the complaint. I was supposed to get 2 miners to replace the broken ones I got....I have gotten no follow up email. So I presume nothing now until the lawsuit is done with. Also FYI google AMT lawsuit its the 3rd link.

So the issues with the board are indeed hardware then? Software won't fix it?

I will be finally plugging in the system and powering up tonight. I was stuck at a client site most of the day yesterday so I had no chance to work on it. The OS is pretty much ready (minus web interface I will bring the AMT interface over after this portion works)

Knock on wood and hope this works. I got all the software ready. Was not hard to set up the distro itself, most of the work was in making sure the right hardware support was compiled into cgminer. I am hoping there are not different driver registers that cause the cards not to read (which I think is why the bitmine firmware did not work for me). If so I will have to find the registers and add them in manually into the driver code and rebuild. This would be the only issue I anticipate right now.