New Official AMT Thread

[NotFuzzyWarm]

Here is the A2 board for Innosilicon that has the same electrical characteristics at the A1:
<picsnip>
I don't see any large capacitors near the A2 chips.

I just realized one difference though,  notice that it is clustered 4 at a time.   This may be hinting at an issue that the SPI daisy chain should be max 4 in lenght and not 8 as in the original specs.

Local decoupling caps is but one way for best-practice. Note at least one other major difference:

The bitmine/AMT boards use 12? Vcore supplies (I swear somewhere I have a pic showing 6) that from I can tell are all feeding one very large set of power planes. Unless the Vdd and return planes are incredibly thick there will be large local circulating currents. There will be interaction between the chips ergo the pretty much mandatory need for the larger 6uf local caps. to reduce that problem. The myrid of tiny local 100nf caps. surrounding the A1's take care of the dips & rebound spikes cause by the logic switching. ISa: are the power planes segregated at all?

Now from what little I've seen of it the TechnoBit version only addressed the thermal needs by making it easier to clomp a single large heatsink to the top. But they kept the same idea of the paralleled regulators feeding everything.

Sorry Isa but I gotta say bad idea. For one, if there is a major fault either the whole board goes down if like from a screw shorting things (a bit more care building and testing takes care of that barring something shaking loose in shipping). Or, rather more serious from liability standpoint, because by necessity that very large single source packs a lot of punch if a chip shorts, well we've seen the results of that. Burnt boards with hopes the PSU shuts down before fire.

The new InnoSilicon board addresses those points. They kept the slew of tiny low value caps around the chips and yes have the needed larger caps out at the edges of the sectors leaving access to the chips for the top heatsink.

However. Note 2 things:
1. Power wise the ASICs are actually in sets of 2. Each ASIC pair has its own set of regulator FET's and buck inductorss. Not 12 feeding 8 chips. If a chip goes kerfluey hopefully you just lose the one or at worst the set. Also as you mentioned they may have split up the SPI chain as well Good idea. Again at worst you only loose 2 chips.

2. Those power planes appear to be gold flashed/plated. Between splitting up the supplies and the gold the circulating currents I spoke of are taken care of. Good choice there and not really *that* expensive. Depending on how thick maybe 5-cents per board tops. Electrically silver would be best but with all that aluminum in close proximity can cause corrosion problems unless passivated (even more cost).

3. Goes fuzzy before zooming in much but looking at some of the holes for heatsink screws the board looks pretty thick. Evey with 4 layers too thick for being made of FR4. Tlam or such perhaps?

Given all that attention I'd like to see the contact side of their top sink... Bet it has copper riser plugs pressed into the aluminum.

[NotFuzzyWarm]

Polycarbonate (Lexan is a registered TM) great choice. Not only good temp before warping but also unlike acrylic it is hard to shatter.

On those mysterious 10k resistors - def a mistake on them. No way in hell are they right. Even from a size standpoint, those would be 2w resistors or more and @12v 10k would dissipate a max of 14mw. And no, no polarity. With resistors the band is just for vision systems on the pick-and-place to know how to verify the number faster. Assuming it uses one that is...

On the PSU switching and PSU in general, is the 12vdc only used to power the hashing cards? If so, ja using the power_OK line is a good idea. Better yet (and cheaper) - how about a lil 5vdc supply for the pi and whatever else needs it and a 1U server PSU for the cards? Leave the Pi on and use the PSU remote on/off to switch it.

For my Ant pharm I use ones from Amazon. $51 for 1,200w HP server psu (runs 2 OC'd s1's pulling 850w per-pair) and @ >95% efficiency they are better and smaller than any PC supply.

Sounds like with better attention to assembly and thermals the AMT/BMch rigs can do fine. The SPI chain thing scare me a bit though. From what I gathered per that SPI coms wiki it doesn't do ring config, just serial chain and at best a star & chain config if all 4 master lines are used. No matter what, all I/O goes through the 1st (master) chip..

[ISAWHIM]

Polycarbonate (Lexan is a registered TM) great choice. Not only good temp before warping but also unlike acrylic it is hard to shatter.

I would be using "Lexan", the brand-name... Unless I can find another source.

Managed to work-out a shroud that fits perfectly into a 12" x 24" pre-cut piece, perfect shape for the 5-card setup. Seems to fit nice and snug. Just have to remove the back-plane, slide it in, and insert the back-plane back into place.

Ordered more Nichrome wire and another LED-dimmer for a controller, to bend it better. Blew my last strand of wire a few weeks ago. Left it on a few hours too long.

Ordering some nice thick Lexan to build two separate custom cases, and some thinner sheets for shrouds. Just need to find some 20-pin ribbon-cables and get some "U" channel to support the cards. Getting rid of that whole metal frame. Was going to chop it into two halves, but it is setup so awkward. Now all I need is more individual cards!

Reduced the whole 5-card setup into a 16" * 7.5" * 14.5" (deep) setup... smaller than my home PC. xD

Going to make it smaller when I split it into 2 units of 3 and 3, and chop the large heat-sinks down by half, snugly packing them together. (The 24-pin ribbon-cables will give me the ability to move them closer together.) The excess length of the backplane just extends over the PSU, since 2 connections are not being used.) I got me a little GH-Cube. Though, adding the ribbon-cables will make it about 9.5" * 9.5" * 14.5" (deep). Ras-Pi also relocated to being above the boards, with the back-plane and ribbon-connectors.

[opieum2]

Quick update. I got the raspi hooked up with the pidora distro....backplane not detected. I need to figure that one out. Might just mean its a SPI driver issue I have to sort out. I will post more tomorrow if I solve the issue.

[NotFuzzyWarm]

Cool. Wish I had my miner...  'Course I'd like to fly and spit diamonds as well.  So far seems as likely

Doesn't the backplane have a coms chip on it for the GPIO/hash cards coms?

[mdiggy]

So........
It's been awhile since any sort of update from AMT.. Have emailed them plenty of times with no responses.... Has anyone got through to them? Still waiting on my machines........
Thanks!

[NotFuzzyWarm]

Not a peep from them even regarding the hosting they were hawking.

[opieum2]

Well they likley are barred from saying anything now. Considering their last statement that might be the case. I guess noone will know the compensation until this case is over.

In other news I figured out the driver issue. I am building out a driver for the at-spi interface which appears to not be set up (mistakenly assumed it was on pidora) I am hoping this works after a reboot.

Also I am going to setup PHPminer as the web interface for this firmware. Seems to work quite well for what we need.
https://phpminer.com/?page=screenshots

Once I get the driver issue out of the way and working I will update with PHPminer and put out an alpha 1 release. This will work for all coincraft A1 devices. Not just AMT's

[NotFuzzyWarm]

Looks pretty cool. Any chance it can do some basic 'ux stuff as well like the Ant's interface does? Mainly easy backup/update?

[ISAWHIM]

Also I am going to setup PHPminer as the web interface for this firmware. Seems to work quite well for what we need.
https://phpminer.com/?page=screenshots

Once I get the driver issue out of the way and working I will update with PHPminer and put out an alpha 1 release. This will work for all coincraft A1 devices. Not just AMT's

Nice... Control... Just what we need to tune things.

Trying to figure out why my cards do not like to run for more than 18 hours... I keep catching them slipping down from 590GHs (3 cards), down to 480-380GHs. (Heat is not the issue, it was cold as hell the last few nights.) Seems software related. Since rebooting fixes it instantly. (Which takes a few boots for the cards to stop from getting stuck on the "MONITOR" stage still.)

It is like it may have a memory-leak, or if it is hardware, the frequency-controller is drifting. (Thus, causing it to reach the point of generating more errors, operating at a higher frequency. I'll have to see if it is errors causing the issue next time. If it is drifting down, then it may be operating too slow to sustain the cores communications. Or, with two separate clocks, it may just be falling out of sync. But that would be a hardware error. Guess an oscilloscope would come in handy to check that. I don't have a digital counter that can count that fast, accurately.)

But it does take about 18 hours, which I can normally catch. That, or just reboot every 12 hours. Not that big of a deal with only one machine. (This is where the 2 days, running for 24 hours would detect.)

Selling my thermal imager... Before the BTC hike starts. (Looks like less than a week, and I don't think there will be a final dip before the hike.)

I'll use some money to buy the things I need to finish this lexan case, and try to get the other cards running. May end-up just selling this miner too. If I put enough glitter and glow on it, I should be able to get close to what I paid for it. (Minus the $60 in BTC I earned over the last few days. xD) In the hike, I should be earning closer to $300-$600 a day. For at-least 30-60 days, when it is all said and done. That will be nice, but I would have made more just buying BTC directly, again. Would be more if I could solo-mine alt-coins... But pools will have to do for now.

We'll see how things go, once I get some of those supplies in, after my thermal imager sells.

[opieum2]

Anyone notice the new changes to the AMT site? Must have happened pretty recent. Visual changes it appears at the moment. Nothing in content. But still no word from AMT. No email, phone or anything at this point. Wonder whats going on.....or if we are even getting our replacement hardware....I essentially paid 11k for 500Ghs at this point. Not a happy camper.

[opieum2]

Anyone who is a veteran use USAA to wire over their money? If so please contact me there might be some options there. Since Josh or anyone at AMT is no longer responding to anyone at all it seems, its time to get the banks involved as well and especially one that is for military and their families. Anyone getting any replies from AMT?

[NotFuzzyWarm]

Anyone notice the new changes to the AMT site? Must have happened pretty recent. Visual changes it appears at the moment. Nothing in content. But still no word from AMT. No email, phone or anything at this point. Wonder whats going on.....or if we are even getting our replacement hardware....I essentially paid 11k for 500Ghs at this point. Not a happy camper.

Compared to earlier Wayback snapshot they have removed pics of smaller miners from home. They and the 1.2TH are still there when ya click on 'see the specs' on the rotator for the 2.4TH.

Paid with check here, deposited at Eagle Bank. Nairy a peep from them since and nothing for some time now...

[NotFuzzyWarm]

Also, along earlier lines  re: caps... Seems the earlier Ant s2 have an issue with that very thing. Bitmine has the info on where to add more across the local chip caps... https://bitmaintech.com/files/download/Guideline%20of%20Soldering%20Capacitor%20to%20S2.pdf

[ISAWHIM]

Also, along earlier lines  re: caps... Seems the earlier Ant s2 have an issue with that very thing. Bitmine has the info on where to add more across the local chip caps... https://bitmaintech.com/files/download/Guideline%20of%20Soldering%20Capacitor%20to%20S2.pdf

Wish I knew where to do something like that on the A1 boards...

Today I had another card just bottom-out. When it was dropping to 480GHs, it was one of the cards dropping down to 50-20GHs. Sadly, that was a card that ran at 208GHs previously. Not even sure where to troubleshoot that card. Like the one that was shorted, ironically in the same location on the backplane, it has done exactly what the shorted one had done.

I got about 3 days out of it... running at 208GHs. Ran fine the first day. Then dropped-out to 20Ghs... Reset the miner, it ran about 18 hours, then dropped back down again... Then ran about 12 hours, and dropped back down again... Today, ran about 3 hours, dropped back down... Now it only runs about 5 mintues, before it drops back down.

Obviously, something is decaying.

Showed no hardware failures an no errors... it just didn't run any faster than 20GHs. The coils were hot, but not the chips, this time. (Voltage regulator issue, or a cap issue?)

Too tired to rip the whole unit apart again, to pull the card. Just unplugged the card power. Hope the other cards work, but that still leaves me with a total of only 1THs of cards. Running only 380GHs now, off two cards. Glad I waited to RMA.



Red is the unusual spike... It was hashing waaay higher than normal. (I assume the initial drift-up.)

Green is the resulting failure... I caught it here, and rebooted...

Blue is the end-result, failing after a few hours before I caught it... Resulting reboot did not "fix the issue" this time. I just left it unplugged. The final result is only 2 cards running for the remainder of the day.

[ISAWHIM]

Just for kicks and giggles.. I powered-up the faulty card... Seems to be running back at full speed again. (10:21 PM GMT -5)

Getting about 590GHs now... It got faster! xD...

Seeing how long before it faults again, dropping back down to 20Ghs... It had about 4 hours of cooling time. Might be that the coils are not mounted well enough, and could gain from having a hat-sink on top of them too.

I have a few CPU heat-sinks I can chop-up to fit those. Just need some thermal-transfer pads. Don't want to smear heat-sink compound across those suckers.

I did notice that power consumption is up to around 800w (796w) at the wall, while the machine runs about 590GHs. Seems to indicate the crystal failing, or the driver circuit relaying the clock-freq, failing or falling out of spec. (Higher freq translates to higher workloads, thus, it seems to be slipping towards turbo-mode, though it is still set on nominal settings. Not pure turbo, but faster than nominal. Or, if it is the voltage regulator failing, it is delivering more power, and thus, also slipping/drifting towards turbo-levels. That, or the higher voltage is the result of the clock-crystal slipping.)

[NotFuzzyWarm]

Don't suppose your meter has hi-lo hold function does it? If does you can wire leads across one of the local chip caps and record the core voltage and see if it changes...

[ISAWHIM]

Don't suppose your meter has hi-lo hold function does it? If does you can wire leads across one of the local chip caps and record the core voltage and see if it changes...

No, but I can get one... hard part is cooling while trying to probe, on a running card... Need a real big fan to run it without the shroud.

EDIT: (3:15 AM, oddball card is still running normal, so far.)

Going to make a per-card shroud and air-guide, to force air to the desired components. Seems to exit the fins and travel freely in the giant air-gap, which has less resistance, and not over the inductance coils enough. Again, this is where a tighter design and larger "wide" heat-sinks on the backs of the chips would really pay-off.

Got hold of some cheap Lexan. Came from picture frames someone was throwing-out. Free is always good. They are like Walmart frames, with cheap Lexan covers, instead of glass. Also got some insulation-foam-board, from construction they are doing at the place I work. That will make great space-filling material, to guide the air where it needs to flow.

Getting some click-pens, so I can pull the springs from them, to apply pressure to the copper heat-sinks. Just to test the idea. Just what I want... loose metal springs around all these electrical connections. xD (It was either that, or steel barrets, which use spring-steel, and are cheap at the dollar-store. )

[opieum2]

Just for kicks and giggles.. I powered-up the faulty card... Seems to be running back at full speed again. (10:21 PM GMT -5)

Getting about 590GHs now... It got faster! xD...

Seeing how long before it faults again, dropping back down to 20Ghs... It had about 4 hours of cooling time. Might be that the coils are not mounted well enough, and could gain from having a hat-sink on top of them too.

I have a few CPU heat-sinks I can chop-up to fit those. Just need some thermal-transfer pads. Don't want to smear heat-sink compound across those suckers.

I did notice that power consumption is up to around 800w (796w) at the wall, while the machine runs about 590GHs. Seems to indicate the crystal failing, or the driver circuit relaying the clock-freq, failing or falling out of spec. (Higher freq translates to higher workloads, thus, it seems to be slipping towards turbo-mode, though it is still set on nominal settings. Not pure turbo, but faster than nominal. Or, if it is the voltage regulator failing, it is delivering more power, and thus, also slipping/drifting towards turbo-levels. That, or the higher voltage is the result of the clock-crystal slipping.)

FYI if its a particular card, it's likely a precursor to failure. I had the same problem with 3 cards initially that worked before they died.

[NotFuzzyWarm]

Can always just solder a length of twisted-pair out to somewhere convenient outside the shroud from the chip cap.
A shot in the dark - is the flakely board one of the ones with the 10k resistors in place of caps? If so could be that the remaining caps are starting to fail from excessive ripple currents.

Buck inductor temps are something I've never been comfortable with for various reasons. They are actually rated to run pretty bloody hot with max ratings often 140C or more. At least they are one of the few things that heat does not really affect until insulation failure. If you can keep a finger on it without branding yerself then it's good. I'd worry more about the solder connections to them failing from temp cycling and the heat getting to nearby components.