New Official AMT Thread

[mellorbo]

I've got the two working boards up an running. I've been sick so after my PTSD class today I just came home and passed out. Didn't get a chance to call AMT.

I'll make that my job for Monday. Try to get in contact and get an RMA for the three non working boards.

I'll ask for that or a refund for 2/3 of the purchase price since I've only got 1/3 of the stated performance. I'd be happier to settle up with them the second way to be honest. That way I don't have to deal with thier bullshit anymore.

I'm just too worn to deal with it right now.

Curious,  I've seen two kinds of boards,  one with orange heat sinks and another with blue heat sinks.

Which ones are bad and which ones are good?

my boards work and have blue heatsinks, my friend tims has copper heatsinks/orange and his dont. i also noticed on my boards on bottom right (fan side) it says Rev 3.0A so not sure if that has anything to do with it either

[regtable69]

Alright. I got the miner mostly up and running. Anyone run into the 0ghs issue? All the modules are showing seems like everything is good. (I cleaned all the excess gunk on the one board but other than pretty decent buildwise). How did you guys address the 0Ghs issue if you had that happen? I am working on it now but if I got an answer it would go quicker lol. I imagine its software at this point as the hardware is being read correctly.

voltage possibly to one card lower than rest module one is the left most blade from the ribbon cable if using the amt interface. how many active cores has the 0g blade got?

[ISAWHIM]

Can anyone find out the layout of the chips with what is detected in software?  That way we can selectively remove the bad ones?

Got a heat-probe? (The bad chips will be cooler. However, it could be the sister-components causing failure or poor performance. That, or contact shorting from the heat-sink compound on poorly protected terminals and through-holes.)

Just for the record, testing is limited to a short burn-in period. Once cooled and settled-in, there is a chance that post-tested boards may later show signs of under-performance or apparent failure. That is not an excuse, just a real situation. Best to get an RMA, have a replacement sent, and the troublesome boards can be "reworked", or the post-test operation failures can be further diagnosed.

This is one reason why I do not recommend getting the "kit alternative", at the moment, unless you have the technical background to troubleshoot and test these things.

I am not going to suggest fixes beyond anyone's ability, which is why I was going to offer live support. To screen for those who, well, just don't have the understanding, ability or tools required for the more technical stuff. Those individuals would be directed to RMA the misbehaving boards, after a quick "test" that the non-technical individual could do, with guidance. (To ensure there is not just some other issue, like power or backplane issues.)

[FrictionlessCoin]

my boards work and have blue heatsinks, my friend tims has copper heatsinks/orange and his dont. i also noticed on my boards on bottom right (fan side) it says Rev 3.0A so not sure if that has anything to do with it either

Okay,  so far we have evidence that the boards with the blue heat sinks work.

Does anyone here have a board with orange heat sinks that work?

I checked the boards,  no difference between the orange and blue,  just the heatsinks are different.

Confirmed here, my boards are Rev 3.0A.

Well, still waiting for my backplane to be able to test out the boards.

[ISAWHIM]

In lieu of a non-existent AMT technical support staff, can we send our bad boards to you for inspection?

At the moment, no. However, soon, possibly. Let me get settled-in first... I just got out of the hospital a few days ago.
(Had an infection knock me off my feet. Another week of antibiotics and pain-killers, and I should be 100% back on track.)

[TooDumbForBitcoin]

Anybody know a good way to remove the coincraft A1 chips from the board?

Maybe this stuff is worth at least the A1 chips on the board?  

Maybe we can get these guys in Virginia:  http://ntekcomputers.com/products/1ths-asic-bitcoin-miner-1000gh  to re-build our systems?

Heat guns can be found as cheap as $20, with spreading and focusing tips.

Here are some SMT tools and more info, if you are looking to expand your collection.
http://store.curiousinventor.com/guides/Surface_Mount_Soldering/Tools/

If you don't have a good heat-gun, with a temp-control and focusing-tip... There is an alternative way to rework or remove the components. (This is NOT an AMT recommended solution for anything. Use this information at your own risk, and voiding of any form of warranty.)

You can use a hot-plate and focus the heat with an aluminum or copper block. Use an old pot-holder as insulation, to focus the heat where you want it. You can also use a top-side fan, to ensure the stray heat does not loosen other components.
IRON as a hotplate. http://runawaybrainz.blogspot.com/2011/11/surface-mount-devices.html
SKILLET as a hotplate. https://www.sparkfun.com/tutorials/59
BURNER as a hotplate. http://jimlaurwilliams.org/wordpress/?p=2101

1: Cut a hole in the pot-holder, to fit the block of copper. (Just a tiny bit shy of the actual size, so it holds firmly to the block, as you push it through.)
2: Place the copper block and heat-shielded pot-holder on the hot-plate, and get it "up to temperature". (You can use another pot-holder to contain the heat, like a cover, until the temp is high enough to melt the low-temp SMT solder. I use a junk PCB, which I can just lay on the block. When it turns shiny, it is just about ready.)
3: Brace the board with cardboard, so it will stay as level as possible against the block of copper.
4: Use a piece of electric tape, looped so the sticky-side is outward, hung from a pencil or paper-clip.
5: Tap the tape-wand down onto the chip, while it is heating, to try and "grab" the chip, to lift it off the melted solder. (Do not hold the tape down, just tap it down and pull it up quickly, or it will melt the tape. If the chip is loose-enough, it will stick and lift up off the board. Otherwise it will remain, when you pull up the sticky "grabber".
6: Use copper desoldering braid, to clean-up the excess solder from the chip, (Do that fast, don't linger with that focused heat on the chip.)
7: Clean the tape-gunk with a fresh piece of tape... Tap and pull, tap and pull.
8: Then clean any heat-sink compound off the chip with alcohol.

NOTE: The epoxied heat-sink on the top of the chip should/may loosen with the focused heat. If not, then it would be safe to remain on the chip.

NOTE: This board may not be safe to rework in a reflow-oven. There may be secondary components that my melt, which were added after the reflow process. (Such as the cable connections. Which would have to be removed the old way, with copper braid and a standard temp-controlled solder-gun. Prior to reflow-work in an oven. Never use an oven you cook food in, as the boards will release toxins into the air, which will later contaminate food.) You can quickly remove all parts at once, in a reflow oven, with an insulated fiberglass catch-tray, suspending the board upside-down over it. All the parts will drop-off, except for a few smaller capacitors and resistors. A quick tap should liberate them from the boards.
http://www.freetronics.com/pages/surface-mount-soldering-with-a-toaster-oven#.U1I35PldVK0

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)."

[NotFuzzyWarm]

<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....

[ISAWHIM]

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)."

The "Kit option", was not intended to be produced. It was for those who felt they could not wait any-longer, and felt they could live with the incomplete units. (With the missing components, stated.)

However, there seems to be a little more "hobby", than expected. The new revisions of the boards and backplanes and PSU's, (which was what they did not want to wait for), are being shipped now. There is still an issue with post-tested boards, which I am sure a better testing setup will resolve. Once they know the source of the failures. Josh has informed me of these potential issues, as well as the forums. These "kits", should not be a representation of the "final product". (For those who have gotten the fully assembled units, with shipping damage, that is another issue.)

The short-falling of the chips themselves, and the original erred designs, given by the MFG, is not making this easy for anyone at the moment.

[FrictionlessCoin]

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)."

The "Kit option", was not intended to be produced. It was for those who felt they could not wait any-longer, and felt they could live with the incomplete units. (With the missing components, stated.)

However, there seems to be a little more "hobby", than expected. The new revisions of the boards and backplanes and PSU's, (which was what they did not want to wait for), are being shipped now. There is still an issue with post-tested boards, which I am sure a better testing setup will resolve. Once they know the source of the failures. Josh has informed me of these potential issues, as well as the forums. These "kits", should not be a representation of the "final product". (For those who have gotten the fully assembled units, with shipping damage, that is another issue.)

The short-falling of the chips themselves, and the original erred designs, given by the MFG, is not making this easy for anyone at the moment.

So is AMT saying that once they eventually get everything fixed they would ship working boards and MPP upgrades?

[opieum2]

Can anyone find out the layout of the chips with what is detected in software?  That way we can selectively remove the bad ones?

Got a heat-probe? (The bad chips will be cooler. However, it could be the sister-components causing failure or poor performance. That, or contact shorting from the heat-sink compound on poorly protected terminals and through-holes.)

Just for the record, testing is limited to a short burn-in period. Once cooled and settled-in, there is a chance that post-tested boards may later show signs of under-performance or apparent failure. That is not an excuse, just a real situation. Best to get an RMA, have a replacement sent, and the troublesome boards can be "reworked", or the post-test operation failures can be further diagnosed.

This is one reason why I do not recommend getting the "kit alternative", at the moment, unless you have the technical background to troubleshoot and test these things.

I am not going to suggest fixes beyond anyone's ability, which is why I was going to offer live support. To screen for those who, well, just don't have the understanding, ability or tools required for the more technical stuff. Those individuals would be directed to RMA the misbehaving boards, after a quick "test" that the non-technical individual could do, with guidance. (To ensure there is not just some other issue, like power or backplane issues.)

My impression is that opieum2 got assambled miner from zipkin hands?
Am I wrong?
0 GH issues sounds like 100% CRAP
Do you still belive in Santa Claus
Will you be so kind to explain in behalf of Zipkin why folks need to buy 1500-1600 W PSU

Ask for your money dude

Truth:
https://bitcointalk.org/index.php?topic=569926.msg6297210#msg6297210
it is takin a minute to read but i warn you all it may take ages to understand who are you dealing with
Remeber: Poserkoto Zipkin has tremendous know how:
https://bitcointalk.org/index.php?topic=422243.msg6200301#msg6200301

So he is fixing georgio design and loosing ton of money every seccond bcause he is not even able to mine with the crap that he is producing.

But georgio bitmine boards are working Can you give me a reason why zipkin crap is not working then?
 
The only reason is that he is shiping crap intentionaly and minig with the rest

Hey losia if you are going to quote me quote me right. Also as an FYI I already solved the 0GHs issue. That was a detail I missed. Casey made mention of a power problem which I fixed...an issue which Josh also mentioned to me would likley happen. I recalled after what Casey mentioned. So for those who have a Lepa 1600W PSU make sure you use the instructions for Lepa in the bitmine thread. That fix applies here too. Meaning use the CPU connector to power the 5th card.

So both my miners are working so far. I am currently testing them out and getting my impressions together into a document and later on a post. I am getting 1.2 on non turbo at the moment on at least one machine I have tested, I have not gone turbo yet. Each module is running at between 225Ghs to 250Ghs so nominal as advertised. The other machine I will plug in once I have the second circuit fully freed up for this and review both at the same time. A few people have obviously run into issues.

Also I was very clear about what I got multiple times (hence quoting me right). I got a DIY kit 2 of them. Not a full miner. And they both work. For the most part minus the 0Ghs issue its been pretty smooth in my case. I have been taking my time making sure I cover every detail to prevent any issues as this was a major investment for me as well. The minor hiccup related to power but those fixes were already documented and I didn't bother to take a second look so thats on me. I have a electrical engineering and medical robotics training, besides an IT background so there are not too many issues with these machines that I cant fix short of bad boards or chips (which I do not seem to have from running the machines). I will test in turbo later on to see how well it works and monitor the systems carefully. I have my thermal gun (and heat probe) and multimeter to check on things. 

AMT uses the Yocto project linux distro. Basically if you never heard of it its a project that allows you to generate your own embedded linux. While the project itself is GPL AMT's stuff might not be so I will ask permission to see if I can repost an image in my post at some point. Mine has a single modification and that is a static IP address for bootup at 192.168.0.234. All I did was edit the /etc/network/interfaces file for static assignment (so be aware of that if you already have something on that IP address using the posted image). I will post more as I am still working on this. I may have the seperate post up today or more likley tommorrow as I have to get through this first.

[opieum2]

Its interesting to see how if this thread goes on track for a bit, the useless commentators show up again. They have less than nothing to contribute so I consider them trolls at this point. Loshia contribute something. You are an idiot who just takes perverse pleasure in seeing others misery. Otherwise you would be actually trying to help people out here instead of just going on and on about poserkos and all that nonsense and bad english you keep going on about. Contribute or GTFO troll. Goes for the rest of you who keep on talking about useless crap. I think the horse has been more than been beaten to death on the topic of AMT as a company. Those of us with issues with the hardware just want it working bottom line. If it can be done great, if AMT provides service great, if not EVERYONE here already knows their options and are adults. Having a group of trolls beating the drums of war just makes you look even more retarded. Seems like you have no life but to be on here and comment on this and take pleasure in the misery of others and lol indirecly saying I told you so. We fucking get it now go away and bother someone else.

Those of us with problems just want results and help to get our hardware working. ANd in this case there are a few people here willing to help. If you dont want to be apart of that and you have the ability to then you are useless if you just want to talk about poserko this and poserko that. We don't speak your language. But whatever it language its in you are clearly just talking stupid whether its bulgarian or english. Because its useless to everyone here at this point.

[opieum2]

Roger!
Dark side is getting stronger
anyway zefir mentioned a long time ago about proper way to use psu and spread the load
I am not bashing but for such a background it seem so stupid mistake
Happy mining dude...
So it turns out the amt happy customers are almost two
 

Yes he did. And like most humans I make mistakes. When you can claim you NEVER made a mistake you can talk shit. But until then stfu troll
Normally I do not expect a PSU to need a modification like that to have functioning hardware. And ASIC miners are new territory for everyone involved here. So yea mistakes are going to be made. Again you are no expert and likely have no usable credentials to make a comment like that so as I said before you are useless.

[opieum2]

my boards work and have blue heatsinks, my friend tims has copper heatsinks/orange and his dont. i also noticed on my boards on bottom right (fan side) it says Rev 3.0A so not sure if that has anything to do with it either

Okay,  so far we have evidence that the boards with the blue heat sinks work.

Does anyone here have a board with orange heat sinks that work?

I checked the boards,  no difference between the orange and blue,  just the heatsinks are different.

Confirmed here, my boards are Rev 3.0A.

Well, still waiting for my backplane to be able to test out the boards.

Mine work just fine. I have blue ones and black ones. If by orange I assume you mean the copper ones? If so then yes I have those. And they are working just fine.

[FrictionlessCoin]

my boards work and have blue heatsinks, my friend tims has copper heatsinks/orange and his dont. i also noticed on my boards on bottom right (fan side) it says Rev 3.0A so not sure if that has anything to do with it either

Okay,  so far we have evidence that the boards with the blue heat sinks work.

Does anyone here have a board with orange heat sinks that work?

I checked the boards,  no difference between the orange and blue,  just the heatsinks are different.

Confirmed here, my boards are Rev 3.0A.

Well, still waiting for my backplane to be able to test out the boards.

Mine work just fine. I have blue ones and black ones. If by orange I assume you mean the copper ones? If so then yes I have those. And they are working just fine.

So, as I understand it, these boards should work assuming that there is adequate power?

AMT sent me a 1500w SilverStone power supply,  do you think that is adequate to drive 5 boards?

They have yet to ship me a back plane, but they promised an extended one and an additional board.   Do you think 1500w is sufficient or do you think I need another one?

Can you tell me what kind of board is attached to the raspberry pi?

[ISAWHIM]

Are you speaking on behalf of AMT, whats the story?

No, I am not speaking on behalf of AMT. I am simply parroting the obvious and giving personal opinions. AMT speaks for AMT. I am, for what it matters, assisting AMT on the technical side, through my informal contract of services. My duties, as I have stated, are related to locating and isolating faults in post-tested boards, for reworking. (Loss prevention and recovery. To reduce the overhead of potential losses, from under-performing or faulting components on the cards.) Thus, my desire to have knowledge of the specific issues, which are reported here, are of strong interest to me. This gives me a specific guidance for things to look for, to resolve. (Just as Microsoft had issues with PSU's and heat-sinks, post production and post testing. The cause of 90% of the RROD situations.)

Hope that clears a few things up.

The guys with the REV 1 boards, essentially have a "reference design". That is what Bitmine provided for suggested builds. (They also had errors in the gerber-files and parts-lists, provided by Bitmine.)

The guys with the REV 3? boards, essentially have a "modified reference design". That is a board with some of the initial major short-fallings fixed. (Obviously, there are a few more minor issues with those boards, or the issues lie within the A1 chips themselves, which is still being determined, through the software side.)

[FrictionlessCoin]

Its interesting to see how if this thread goes on track for a bit, the useless commentators show up again. They have less than nothing to contribute so I consider them trolls at this point. Loshia contribute something. You are an idiot who just takes perverse pleasure in seeing others misery. Otherwise you would be actually trying to help people out here instead of just going on and on about poserkos and all that nonsense and bad english you keep going on about. Contribute or GTFO troll. Goes for the rest of you who keep on talking about useless crap. I think the horse has been more than been beaten to death on the topic of AMT as a company. Those of us with issues with the hardware just want it working bottom line. If it can be done great, if AMT provides service great, if not EVERYONE here already knows their options and are adults. Having a group of trolls beating the drums of war just makes you look even more retarded. Seems like you have no life but to be on here and comment on this and take pleasure in the misery of others and lol indirecly saying I told you so. We fucking get it now go away and bother someone else.

Those of us with problems just want results and help to get our hardware working. ANd in this case there are a few people here willing to help. If you dont want to be apart of that and you have the ability to then you are useless if you just want to talk about poserko this and poserko that. We don't speak your language. But whatever it language its in you are clearly just talking stupid whether its bulgarian or english. Because its useless to everyone here at this point.

+1

Loshia... you are no help to AMT customers who have been screwed.  

The options are clear and simple:

(1) Join the class action lawsuit.
(2) Try to get the AMT hardware working.

In both options, customers are trying to get back their investment.

I am of course very disappointed with AMT,   however since I did get hardware (albeit not working yet), I am going to research what I can to get it hashing.   If of course it never does, then option (1) would be the only option.

Well since opiem2 seems to have some success,  then there may possibly some light at the end of this tunnel.

[FrictionlessCoin]

Are you speaking on behalf of AMT, whats the story?

No, I am not speaking on behalf of AMT. I am simply parroting the obvious and giving personal opinions. AMT speaks for AMT. I am, for what it matters, assisting AMT on the technical side, through my informal contract of services. My duties, as I have stated, are related to locating and isolating faults in post-tested boards, for reworking. (Loss prevention and recovery. To reduce the overhead of potential losses, from under-performing or faulting components on the cards.) Thus, my desire to have knowledge of the specific issues, which are reported here, are of strong interest to me. This gives me a specific guidance for things to look for, to resolve. (Just as Microsoft had issues with PSU's and heat-sinks, post production and post testing. The cause of 90% of the RROD situations.)

Hope that clears a few things up.

Wait!  Are you saying that you will be providing technical support for AMT units?

Has AMT contracted you to provide technical support?  Are you also doing re-work on damaged boards?

Will you provide help on missing parts?  I got a missing back plane.

[ISAWHIM]

Wait!  Are you saying that you will be providing technical support for AMT units?

Has AMT contracted you to provide technical support?  Are you also doing re-work on damaged boards?

Will you provide help on missing parts?  I got a missing back plane.

My assistance to customers, will be technical support. (Though limited by several things.)

I am not personally reworking the boards myself, just identifying the point of failures, and marking the components to be reworked.

If I can, I will try to assist with locating or obtaining missing parts. (Once I have settled-in. I just got out of the hospital a few days ago, and have to finalize the "informal contract", with Josh.)

[FrictionlessCoin]

Wait!  Are you saying that you will be providing technical support for AMT units?

Has AMT contracted you to provide technical support?  Are you also doing re-work on damaged boards?

Will you provide help on missing parts?  I got a missing back plane.

My assistance to customers, will be technical support. (Though limited by several things.)

I am not personally reworking the boards myself, just identifying the point of failures, and marking the components to be reworked.

If I can, I will try to assist with locating or obtaining missing parts. (Once I have settled-in. I just got out of the hospital a few days ago, and have to finalize the "informal contract", with Josh.)

So are we supposed to send you the bad boards?

[freddyfarnsworth]

Here is the post info on P/S from zefir:
Old AMT thread is dead, and this is the info to make sure whatever you use complies with the miner Power needs.
Lepa, Enermax ect, if you use MultiRail P/S this is the info you need to understand.

If you have EU 240v power you get even better results.

Bitmine and AMT are darn near clones...

=================

Hello,
I have technical problem with 1 TH/s Desk.
I received 2 1 TH/s Desks yesterday.
I'm using 2 1600W Lepa power supply.
Miners firstly turn off after 5 minutes, I saw, that 1 original fan was set to air flow in, one to air flow out. A set both fans to IN. So now it working for approximately 30 minutes, then one temperature sensor reach 50 degrees - miner shut down. I have miners in room, where is 21 degrees. So I add 2 fans at front side, now it working for 1 hour, but its not a lot.
So, where is a problem? I must put it to fridge?

This is the same issue we observed with the Platimax 1500W: some PCIe connectors share the same rail, with that you are feeding two hashing units over the same rail and overloading it - the Platimax uses to shut down after a while. Since the Lepa's PCIe pinout looks similar to that one, here is the correct way to set up your CCD with the chosen PSU:

1) take a look at bottom of datasheet page 2 (http://www.lepatek.com/files/LE_ProductBasic_eng/PB_File/G1600-EN.pdf)
2) notice that the distribution of the 6 12V rails is as follows: 12V1:MB, 12V2:MB, 12V3:PCIe1+4, 12V4:PCIe2, 12V5:PCIe5+6, 12V6:PCIe3
3) ensure that you connect 4 hashing units to PCIe connectors 1, 2 , 3, and 6 (left 4 and 5 unconnected)
4) buy or build yourself an adapter from MB-20 (12V1) or MB-16 (12V2) connector to PCIe and feed the 5th hashing unit with that

Bottom line: ensure to feed only one hashing unit through one 12V rail. Even if the PSU uses a common rail, powering two hashing units through a single rail will either cause voltage ripples or over-current protection shutting down the PSU.


Good luck

Thanks a lot Zefir, you have right!

I did it exactly how you said, I connected 4 hashing units to separate PCI-E rails and then I modified CPU power supply connector to PCI-E connector, now working stable!  

Thank you very much again

can you make a small instructions for "us"... i think it whould be nice to know for many people...

Yes, of course.

Firstly, look at diagram of your PSU (mine Lepa 1600W)



and connect 4 cabels with PCI-E connectors to separate 12V rails (12V3,12V4,12V5,12V6)
!for each mining module, use only 1 connector on PCI-E cables!

Then cut off 8 pin CPU connector (12V2)



and solder other PCI-E connector (6 pin) to CPU wires

 

BUT beware, 12V wires is on opposite side, look at this diagram:
 


And finally, don't forget connect this connector:

or miner won't see mining modules.

For all of this, thanks to zefir.