I would be interested for sure but 6 weeks is too far out.   Sending PM

I know it doesn't make any sense. Maybe someone has an idea at what is really going on here but it sure isn't me.  This explains why some people find it hard to restart these prismas; they are in a cold room and have cooled off and need to warm up again. After rebooting and fighting it randomly starts to work, only because it has warmed up enough. Ignoring errors, power downs, non-hashing, until the heatsink is warm to the touch, only then will the power downs go away and so do the errors and boards start to hash.

I could throw open the vent to one of my mining rooms that has 2 prisma arrays and the cold air will put them offline within an hour. That room is too hot now for my cointerras(most heat sensitive miner I have) but the Antminers don't care either way.

So fighting with these machines for several weeks now I have found out a couple things:

1. These things are so temp sensitive that I don't see them running autonomously without being installed in a climate controlled environment. Not the heat - but the cold. If I fire them up from cold, or I turn on the vent to full, they will cool down to the point of not even hashing. They don't start to hash halfway stably until the room hits at least 70F+. I shut off the vent for too long and it overheats everything else. This makes me feel like I'm dancing on a wire rope.

2. IR Gun every hashing boards and look for boards that are running 2X+ times as hot as normal hashing boards. These need to be pulled as fire hazard, they may be affected by #3

3. The laminate on the bottom of the pcb/hashing board is VERY thin. One scratch here will reveal the copper substrate inside the pcb and will invariably lead to a short to the heatsink and fire is likely.

I run cointerras, S1, S2, S3, S4, BFL, BTCGarden, ZeusMiners in every flavor, Gridseed G-Blacks, SP20's... This was my very first RMA, flame out and hardware exploding miner, power supplies excluded.

That's a nice camera you got there    Thanks for sharing, it confirms the results I've been seeing. I am curious what they look like when they are being powered on from an off state, when the controller fires up, running, and when it fails to hash. Also if any change is seen from position in daisy chain, and record/log which hashing boards look proper to confirm new failure. Powerful being able to visualize all that information in real time. I suspect those hot spot boards were having problems when you captured those images.

I can confirm no response from the correct email address.

Last night a 3 prisma array wouldn't boot up with all hashing boards, so I took a physical examination.. One tested known good board has deteriorated to a 'hot' board where transistors see 80C+ and my IR gun measured over 95C on one BE200 chip, the same one that is the most burnt in my previous burnt prisma pic. I disconnected this newly bad board and resumed like normal, albeit another 300ghs short. I'm certain if these bad boards arn't serviced and removed they will lead to fire. I have found when this happens they do not recover from this state, I.E. once they run hot, they will always run hot (and not hash well because of it), and break/interfere with the daisy chain.

I have still yet to receive any response from AsicMiner in regards to my bad boards, out of 32 boards, only 18 are working, or 56% of the paid hashrate.   

I've noticed some of that too.. but the problem is, how do you run any amount of them in mass like that? Do they employ a team of peeps at the AMHASH1 data center rebooting arrays until they hash? What about a power blip that resets everything? What a headache 

That's actually how I have mine set up right now. On the other hand I have also noticed they don't like to run cold as well.

I bet if you looked under the PCB of that popped cap nearby you would find a nick in the PCB with a bit of copper inside that causes it to short to the heatsink and pop the closest cap it finds. That's what my burnt board looks like anyway.

Burnt prisma's should be posted over in this thread: https://bitcointalk.org/index.php?topic=791827.0

Sorry my updates are not more forthcoming I'm a little burnt (ha) from troubleshooting these:

So this is the back of the board that caught fire, I can see a spec of copper burned through the pcb in a tiny spot where on the opposite side is one of the burned out asics:



Here is the top side of the board, two views:





I had just fired it up from an off state, where it had previously ran for about a week now, fire started about 10 seconds in. Thank god I was there, I saw flame like a candle and was able to pull power immediately.

Just so you know it is very cold in my mining rooms right now, ~50-60F or even colder at nights. My cointerra is my most heat sensitive miner and it is running cool and stable.

Out of 8 prismas, or 32 boards:

4 have transistors that climb to 80C+ whereas all other boards exhibit 30-60C. They will sometimes still hash, but slowly, and the tube heatsink is really that much hotter because of it. Burning hot. And some of these only exhibit the problem on half of the hashing board.

8 have power down / hashing issues. Suspecting bad chips on these boards. They hash, albeit slowly and reset and generally I am blaming these boards for breaking the daisy chain and causing the others to have problems.

1 has burnt out on me now

I'm not trying to smear asicminer and have no agenda. I am disappointed with these problems. I had some plans that were in holding pattern waiting for the all clear and now they need some rewriting.

I would be happy if someone from asicminer would just offer me replacement of the bad boards. I've lost downtime but it will be more loss without any hashes. Overall this was more of a test run to see if it was worthy to order in bulk. The ones that do run, run well.

I am still having many random problems but especially: I had a board flame out on me last night, pics will be uploaded tonight. Be careful out there.

To Chris, try enabling the debug log and hit the play button on the dashboard in minera for the log, does that give you any more insight?

To others, I get the bet XX errors with the minera lots of times, seems random at best. Tried switching from rb to am cg branches, have yet to get 2 prismas connected to one pi properly.

Anxious to try your image CrazyGuy I'm hoping it will sort out these seemingly random power down issues on problem boards. I think one of my hashing boards would end up like the burnt one above with several transistors breaking 100C even without load. I think that's where it started with one of the transistors completely burnt off and damage downhill. The heat damage stops with the other side/half of the hashing board.

Has anyone tried to use these usb adapters to connect to PC and just use cgminer? Safe to assume to dip switch them 1-4 and 2x per raspi?

The hot board has been unplugged as soon as I discovered it with the exception of testing after inspecting/reseating (with no burn marks or damage), and the inital burn in before I knew I had problems. The other problem boards are not seeing this same high temp but I have not tested every chip to confirm there isn't a hot one somewhere.

Out of 12 hashing boards (3 prisma), 8 are working without issue, 3 work ~20-60% of the time and not for long and have resetting issues, 1 obviously has temp issues.

I am pointing the IR probe to the same transistor from tube to tube. This particular transistor is the worst case. I inspected/reseated the hot problem board already and I see nothing abnormal with it.

Another important point: The high temp will be reached even in the case of the prisma just being on, and not hooked up to BE controller and actually hashing.

I cannot confirm that all the problem hashing boards exhibit this kind of temp, some more research to be done today mixed in with some real work 

So this last batch of prisma's seem to have some problems. The first 4x I recieved from Crazy Guy do work, albeit with a dead hashing board that is totally my fault so I'll ignore that.

I dont think it is related to BE controller because I can swap two controllers and the 4x stack that works, works the same way no matter what controller I use, where the problem prisma array has problems no matter what.

It will reboot ever 10-120minutes. Power downs constantly on the same 4 out of the 12 hashing boards, after reseating the same boards, same thing. Switched order of the chain, and the problem boards move to where I put them.

Here are some odd pictures for you. I pointed my IR thermometer at the transistors at a problem board, and a working board.

Hoping the usb controller will help me isolate the problem more specifically, but my IR probe is pointing me to actual hardware issues.

Problem board (107.9C temp):


Working board: (50.5C temp)


I know hard to see it is reporting in Celsius - What is going on here?

Update:

I think I am making progress. I took the first prisma and took off a problem hash board and inspected: some spots of flux stuck in the holes where the BE200 chips are mounted to the PCB, cleaned those up best I could and screwed it back down.

Fired up the hash boards one by one and I currently have it without any of the PwrDn counts on the controller.

Prisma #1 is now hashing great 
On prisma #2 blade #4 it is getting PwrDn counts and not hashing, I will attempt the same procedure above here for it.
On prisma #3 blade #4 has not hashed yet and seems shorted somewhere I have not removed and inspected it yet.

I think I may have some problem hash boards with that last set 

Hashing fine for a few hours a little under but acceptable for sure then hashrate really dropped.

Isolated them one by one tried altering the way they are chained, two controllers, and two power sources.

One of the boards the row of large transistors gets so hot they will burn you just powered on without the controller, so I have it unplugged, but it will descend into ugliness anyhow like this picture:



gunna unplug it and fight it tomorrow 

Got my 3x more prisma in from CrazyGuy a few hours ago they'd be hashing @ 99% efficiency working to ROI Great little machines and CrazyGuy is easy, fast, honest, and reliable to work with. So refreshing to see a little integrity in this business.

Thanks Sidehack! A quick wiki search yielded this juicy picture:

(They look a lot more shiny and new than mine LOL)

Can I stray a bit off topic and ask the question: What's the back story of these surplus power supplies? They seem to have made a lot of them. Did they actually go in server computers? I was thinking maybe they used them for powering wireless telecommunications? That would explain the trainloads. Does anyone have any of these power supplies that came with a history?