BEWARE Bitmain S7/S9 failures and lack of warranty

[in2tactics]

I am really impressed that neither miner next to the burned up one were affected. In fact, I am impressed the whole house did not burn down.

Generally the materials used won't support combustion so unless something really catastrophic happens, the worst case should be as pictured. I don't remember what the certification series is called but I think boards are meant to self extinguish within 10 seconds or so.

True, I am aware of those facts and I agree with you. I think you are referring to the FR-4 materials grade designation for PCB construction. However, that does not preclude surrounding items, i.e. shelving or other objects depending on material composition, from catching fire.

[1714100078]

1714100078

[1714100078]

1714100078

[1714100078]

1714100078

[NotFuzzyWarm]

I am really impressed that neither miner next to the burned up one were affected. In fact, I am impressed the whole house did not burn down.

Generally the materials used won't support combustion so unless something really catastrophic happens, the worst case should be as pictured. I don't remember what the certification series is called but I think boards are meant to self extinguish within 10 seconds or so.

True, I am aware of those facts and I agree with you. I think you are referring to the FR-4 materials grade designation for PCB construction. However, that does not preclude surrounding items, i.e. shelving or other objects depending on material composition, from catching fire.

The self extinguishing ratings only apply  after the ignition source/condition is removed. In this case after the mains breaker feeding the PSU tripped shutting off power to the (massive) short. Odds are the PCIe power leads were the only thing setting a limit to the power being fed into the short circuit. This miner was tied to a 4kw psu no? If safety is a concern, um, putting a couple 1200-1300w loads on a single high energy source demands fusing EACH of loads to just over whatever current they take, for an s7 that equates to 110-115A to prevent exactly this sort of thing from happening.

Yes because of how they work fuses introduce a voltage drop, if that is too high then using DC circuit breakers is the only option.

Now as what started the meltdown by apparently shorting out the incoming 12V power planes...  Would certainly be nice to know. This also points to one of the few advantages to feeding each miner from a PSU more closely matched to its power needs. Certainly could still have a catastrophic failure similar to this but at least the maximum power that can feed the failure point is far less and more in line with reasonable worst case design considerations.

[philipma1957]

I am really impressed that neither miner next to the burned up one were affected. In fact, I am impressed the whole house did not burn down.

Generally the materials used won't support combustion so unless something really catastrophic happens, the worst case should be as pictured. I don't remember what the certification series is called but I think boards are meant to self extinguish within 10 seconds or so.

True, I am aware of those facts and I agree with you. I think you are referring to the FR-4 materials grade designation for PCB construction. However, that does not preclude surrounding items, i.e. shelving or other objects depending on material composition, from catching fire.

The self extinguishing ratings only apply  after the ignition source/condition is removed. In this case after the mains breaker feeding the PSU tripped shutting off power to the (massive) short. Odds are the PCIe power leads were the only thing setting a limit to the power being fed into the short circuit. This miner was tied to a 4kw psu no? If safety is a concern, um, putting a couple 1200-1300w loads on a single high energy source demands fusing EACH of loads to just over whatever current they take, for an s7 that equates to 110-115A to prevent exactly this sort of thing from happening.

Yes because of how they work fuses introduce a voltage drop, if that is too high then using DC circuit breakers is the only option.

Now as what started the meltdown by apparently shorting out the incoming 12V power planes...  Would certainly be nice to know. This also points to one of the few advantages to feeding each miner from a PSU more closely matched to its power needs. Certainly could still have a catastrophic failure similar to this but at least the maximum power that can feed the failure point is far less and more in line with reasonable worst case design considerations.

I have used finksy gear.
And I have used optimizer gear
And I have used sidehack gear
And I have used Pete m gear.

Along with evga 1600 t2 1600 p2 1600 g2 1300 g2

Seasonic 1200 plat

And more.

To me. The s-9 needs to be run under clocked with the 2k watt boards. Using the IBM dps-1600
The IBM dps-2000 or the IBM-2500 psus. At freq 500 to 600 depending on the batch number.

Bitmaintech has dropped from freq 650 to freq 550 batch 1 to batch 5.

My guess is they are having issues ruining the gear in summer heat.

Right now I am running two batch 1 s-9 at freq 600 in the array with finksy 4k board and two dps -2000
So far okay.
1 s9 batch two at freq 600 with evga 1600 p2
1 s9 batch two at freq 600 with dps-1600 optimizer 2k board
1 s7 batch one at freq 600 with evga 1600 t2

And
1 s-9 batch 4 at home freq 462 with optimizer 2 k board and a dps- 2500

[NotFuzzyWarm]

Hmm, a thought...
Problem: If using a (really) big psu or a bank of 1+n load share ones there is a need to limit power feeding each load.
Fuses not good choice due to the voltage drop they introduce.
DC breakers in the >100A range give very low vdrop but are very expensive.

So... how about incorporating electronic fusing into these high current breakout boards like the ones for 4kw IBM supplies. Say 2-3 banks of mosfets driven by adjustable current monitors (1 bank per-miner attached)? Just dial in the desired current trip point and problem solved. Near zero vdrop and response times from almost instantaneous to whatever is desired.

Oh Sidehack and friends, up for a new project?

[sidehack]

Sure, if there's dev money and a guaranteed market. I don't want to say it's not a good idea, but you're also doubling the cost of something that's already a race to the bottom.

[NotFuzzyWarm]

Sure, if there's dev money and a guaranteed market. I don't want to say it's not a good idea, but you're also doubling the cost of something that's already a race to the bottom.

No doubt about higher price vs an unprotected breakout board but -- when dealing with any supply capable of putting out 3-4kw or more I'd definitely pay the additional cost for peace of mind. To me it would be a nice selling point and considering there are pics a bit back in the thread to show the consequences of not having current limits methinks it should be an easy sell.

Don't suppose there's a way to monitor a few loads and externally trip a PSU's internal current limiter is there? IF there is a limit sense input that could shut down the PSU's internal DC on/off FET's that would certainly help with cost but then also becomes an all or none on situation.

[yslyung]

I am really impressed that neither miner next to the burned up one were affected. In fact, I am impressed the whole house did not burn down.

Generally the materials used won't support combustion so unless something really catastrophic happens, the worst case should be as pictured. I don't remember what the certification series is called but I think boards are meant to self extinguish within 10 seconds or so.

True, I am aware of those facts and I agree with you. I think you are referring to the FR-4 materials grade designation for PCB construction. However, that does not preclude surrounding items, i.e. shelving or other objects depending on material composition, from catching fire.

The self extinguishing ratings only apply  after the ignition source/condition is removed. In this case after the mains breaker feeding the PSU tripped shutting off power to the (massive) short. Odds are the PCIe power leads were the only thing setting a limit to the power being fed into the short circuit. This miner was tied to a 4kw psu no? If safety is a concern, um, putting a couple 1200-1300w loads on a single high energy source demands fusing EACH of loads to just over whatever current they take, for an s7 that equates to 110-115A to prevent exactly this sort of thing from happening.

Yes because of how they work fuses introduce a voltage drop, if that is too high then using DC circuit breakers is the only option.

Now as what started the meltdown by apparently shorting out the incoming 12V power planes...  Would certainly be nice to know. This also points to one of the few advantages to feeding each miner from a PSU more closely matched to its power needs. Certainly could still have a catastrophic failure similar to this but at least the maximum power that can feed the failure point is far less and more in line with reasonable worst case design considerations.

I have used finksy gear.
And I have used optimizer gear
And I have used sidehack gear
And I have used Pete m gear.

Along with evga 1600 t2 1600 p2 1600 g2 1300 g2

Seasonic 1200 plat

And more.

To me. The s-9 needs to be run under clocked with the 2k watt boards. Using the IBM dps-1600
The IBM dps-2000 or the IBM-2500 psus. At freq 500 to 600 depending on the batch number.

Bitmaintech has dropped from freq 650 to freq 550 batch 1 to batch 5.

My guess is they are having issues ruining the gear in summer heat.

Right now I am running two batch 1 s-9 at freq 600 in the array with finksy 4k board and two dps -2000
So far okay.
1 s9 batch two at freq 600 with evga 1600 p2
1 s9 batch two at freq 600 with dps-1600 optimizer 2k board
1 s7 batch one at freq 600 with evga 1600 t2

And
1 s-9 batch 4 at home freq 462 with optimizer 2 k board and a dps- 2500

S9 consumes approx. 1472W or let's say 1500W running at 650 mhz.

dps 2000bb output is more than sufficient to power it.

yes you mentioned to you only but i do not see any issues at all & there is no need to down clock them in fact those passed warranty, i oc them by a tad.

even with summer heat. over here OAT is around 37C indoor after some diy evap cooling it's around mid to high 20's C hot noon.

been running all sorts of miners since then for about 3 years.

so far so good, fingers crossed though.

i run them around 70 deg C board temp with dps 2000bb + jabber's board overvolt to 12.5V - 12.8v on load.

they run fine even on the s7's.

[sloopy]

Having PDUs with individual breakers gives me some of my peace of mind on the AC side.
I purchased the pdus I am running from Finksy and they are 12 AMP for each power supply.
12 AMPS is more than what is needed for each line going to the doublestack 4k b/o boards with 2000BB power supplies, but the 50 AMP breakers feeding each PDU are coming off their own subpanel which has another breaker in the main panel. Everything needed to be is at 80% or less. I run a mix of 2880 power supplies with the 4k doublestacks and "knock on wood" perform a good deal of preventative maintenance.

I will be the first to say there are many people on this forum who know a great deal more about electronics than I do, but I have also done a great deal of forensic "after the incident" troubleshooting of fires with electronics and air / water cooled components.

Finksy what were your temps averaging at the time it happened?
What firmware version was it running?
I believe you stated it was a batch 1, correct?
What was your fan setting set at? Was it checked with a value or unchecked with an empty field?
How long do you think it started before it was noticed?
Does it appear it started with one board and then moved to the others?

From what I see in the pictures it looks like the same or very similar to the issues we saw with S5s running away and burning up.
I hear you that the internet didn't drop, but I do not see a single 12 volt line causing this. In my opinion the guy saying one connection loose causing this is obviously not familiar with the scenario and is trolling you.

I have an S5+ running right now off a 4k doublestack which has been recently moved back to my main mining room. It was running in a room with very little ventilation. The temperatures in that room were easily 110 F ambient and I had it overclocked to 368 freq. Even right now it is at 362 freq, and I could go higher. I don't know how much higher but the temps have and are not limiting me. It had .0016 hardware errors at 368 over a week or so. I am confident using this power supply setup due to the variable input voltage has a great deal to do with being able to get another 1.2 - 1.4 TH out of that unit, and like I said, I am confident I could go higher. I'm sure the efficiency was probably closer to an S4 heh. When I run it at stock in my main mining room it never has a hardware error. After a month it has shown negative hardware errors and I usually give it a reboot when I see something odd on any miner.
I believe the S5+ miners were cherry picked hash boards. notlist3d had extreme success with his. Although I think Tupsu had problems with hashboards. I remember seeing very positive comments on the S5+ miners. There were not many sold, and it was a few days before the S7s went on sale and then we waited, for what? 6 weeks for batch 1 after the S5+?

I stripped my S5+ and most S7 miners. Yes I broke the warranty seal but the S5+ instructions clearly stated to confirm the heatsinks were not loose. I was unable to do so without removing each hash board. So I did.
I have not done this with my S9s, I have done the shake test but I do not feel near as confident as I did when I removed each board and inspected it.

I am considering placing my worst performing S7 boards in this S5+ chassis just to see how they perform. I don't know they will fit, but I think it would be good to see how they perform.
I have a couple of later batch S7s which run hotter than anything else, even the S9s.

Which my first S9 has / had one hash board which produces 95% of the hardware errors. That was temperature related. Once I cooled that hashboard down the errors dropped. It is back on the same power supply, same cables, running at stock 650 freq. It is strictly temperature related. I even spoke with Phil in the main BM S9 thread and I downclocked it with very little change. IT was my temps which looked fine. If they go over the 90ish range that board starts climbing much faster with errors.

My Batch 1 and Batch 7 S7s run perfect, but my batch 14, 15 units run hot as hell. I have swapped the physical locations, power supplies, cables, etc and nothing changes unless the temperature is lowered.

I pushed the S5+ so hard because I really didn't care anymore. I should have taken it offline by now anyway and I have fire protection where miners are. When I pushed the S5+ so hard it was in a room by itself on a metal rack sitting on concrete with a metal roof. If it went, it went.
(You might remember me doing something similar with 13 S3s on one 2880 power supply. Just to see kind of thing. No fire, but all of the S3s would not maintain their speeds.)

I keep the miners spaced out, but obviously the S5+ chassis are all connected. It is the coolest running setup I have. It always shows the lowest temperatures relative to other S5s, 7s, and now the 9s. Obviously there is more heat to dissipate with the 7s and 9s and the standard s5 "case" was a POS.
I think keeping the miners separated on metal racks allows for better heat dissipation. 
My main mining room has more CFM available than what is required. I keep one fan off most of the time and only use a single exhaust. There is another exhaust available at the top of the room.

Enough about my pressing of temperatures, but I do think your issue was temperature related in some way. Be it a runaway control board, etc., but I think you probably had one hash board which failed, pulled down the temperature limiting if it worked at all anyway, which I know mine has worked on a couple of S7s (I am using the latest firmware.) I know I said the S5+ handled the heat well at nice overclocks, but my S7s do not. They do not even run stock well at high temperatures. Ambients of 110 F require underclocks of my later batches. If I am unable to constantly monitor temperatures I downclock all S7s and S9s. 

Bitmain should be willing and able to give you a proper diagnosis. I have no doubt they could, but I also have no doubt they will not. The only real help I have ever received from Bitmain came via Yoshi and his connections which obviously carries some additional price. Not that he always charged me for assistance either, but I have always offered, and I have paid extra for miners. I plan to do so again with Yoshi and I will pay more than the BM website price I'm sure. It is a peace of mind not having to deal with inadequate support and even then there can be some issues I would never put my customer's through, or expect to deal with myself. Unfortunately I believe Colorado and California have their hands tied in many different ways, but Yoshi does what he can and has the ability to do more. I have previously said the support issues with Bitmain start at the top in China. It is a shame home miners are limited in our offerings. I firmly believe there would be a vast difference if there was a company in the West offering products. The competition would force it to be handled better.

I have heat issues with my Avalons as well. They actually did some good work with their last firmware release. I can put in the latest Pi image on a different SD card and the unit stays running, but using the one which came with the unit once the temps hit 93 it no longer mines or connects to the unit until it has been rebooted. This is also using the latest MM update.

I would be shocked at some of the trolling responses you received except I received many of the same type when dealing with Bitmain over an S4 purchase. I was not only screwed over by Bitmain, but many of the forum members who I thought would see and understand a situation where Bitmain had responsibility to do something but those forum members do not allow negative attention to be cast on Bitmain. It is only the few here who have always been honest and consistently present solid facts that provide a realistic opinions. Normally, it is the same people who are actually familiar with electronics / electricity, or at a minimum have years of experience running miners themselves and understand the behaviour through repetition. You do not have to understand how something works if you have seen the same behaviour hundreds of times. An open flame igniting an oxygen leak burning rubber lines will cause a fire every single time. If you see it enough you know what caused it and do not try to explain it away as being caused by a lack of coolant flow. That scenario looks completely different and burns in a completely different manner, but usually people need to see all of the parts to make a proper diagnosis. 

It sucks to realize you simply have to deal with being treated the way you have, pay more than others for your gear, or, not mine.
All I can say now is as a businessman, I know how many people have simply unplugged their miners and walked away from bitcoin mining or maybe bitcoin all-together. Most people will not go to the lengths some of us have to try and get assistance and many less will come to the forum and share the full details of their experience. This is especially true with some of your potentially best customers. If they cannot find a path of least resistance they give up and move on. Kind of like their fuse blew, or breaker tripped... there is no resetting it... they are done.

[DecentralizeEconomics]

I am really impressed that neither miner next to the burned up one were affected. In fact, I am impressed the whole house did not burn down.

Sheer luck.



Gotta love that Ching-Chong Chinese quality.  At what point does their cheapness and unwillingness to manufacture safe devices equate to highway robbery and criminal neglect?

[CjMapope]

I am really impressed that neither miner next to the burned up one were affected. In fact, I am impressed the whole house did not burn down.

Sheer luck.



Gotta love that Ching-Chong Chinese quality.  At what point does their cheapness and unwillingness to manufacture safe devices equate to highway robbery and criminal neglect?

i suppose at the point where their gov's will allow other countries to persecute Chinese criminals ;\

i am loving the cheaper hardware on the market right now with the 1/2, but threads like this stop me from building a farm
they make just what u guys show and say, cheap Chinese crap, but sadly it's easy to see why, noone especially the ones who make the hardware have an interest in it lasting.. ;(

[dogie]

Gotta love that Ching-Chong Chinese quality.

We've had burnt units from loads of manufacturers, and given that Bitmain sells miners in the 10,000 - 100,000, its surprising you don't see more issues appear.

[Finksy]

I am really impressed that neither miner next to the burned up one were affected. In fact, I am impressed the whole house did not burn down.

Sheer luck.

It was on a metal frame rack above concrete flooring, nothing combustible within ~2-3 feet.  The pictures in OP is misleading, as there is a metal fence and a wooden rack on the opposite side about 2.5 feet away.

Gotta love that Ching-Chong Chinese quality.  At what point does their cheapness and unwillingness to manufacture safe devices equate to highway robbery and criminal neglect?

While I agree that they are guilty of highway robbery in terms of price gouging due to lack of competition, I don't think they build miners of poor quality.  KNC Jupiters were infinitely more dangerous from the factory than any Bitmain product.

[Finksy]

Finksy what were your temps averaging at the time it happened? ambient temps down there at the time were in the 28-30*C range
What firmware version was it running? At the time I believe it was still the factory as-shipped firmware for 600M frequency, I believe Sept
I believe you stated it was a batch 1, correct? Correct, stock frequency 600M
What was your fan setting set at? Was it checked with a value or unchecked with an empty field? Manual, 100%
How long do you think it started before it was noticed? It had been running pretty much 24/7 except for very occasional power cycles for maintenance, moving things around, changing PSU's, etc.  I believe at the time of failure it had been 10-14 days continuously running
Does it appear it started with one board and then moved to the others?  I couldn't say man, all the failures seem to have started at the same place on the PCB's. Given the apparent shorting of the PCIe cables/jacks, I would say that the shorting happened at the same time on at least 2 hashboards, as opposed to a single board failing and causing fire damage to the other 3.  1 board still had cables/jacks intact.

From what I see in the pictures it looks like the same or very similar to the issues we saw with S5s running away and burning up. Agreed, with disruption of internet being the cause.
I hear you that the internet didn't drop, but I do not see a single 12 volt line causing this. In my opinion the guy saying one connection loose causing this is obviously not familiar with the scenario and is trolling you. Agreed, if I saw someone else with this failure I would assume that it was an internet failure, but that doesn't explain why it was isolated to one miner.  The Ethernet cable was still intact (apart from burn damage)

Enough about my pressing of temperatures, but I do think your issue was temperature related in some way. I also believe it was temperature related (internal temperature, not ambient), but can't understand why one miner would have runaway temp issues without other suffering of the same, and why the software didn't shut it down when temperatures started running off.  The front fan was still working after the failure, the rear fan was clearly destroyed by fire, but either way should not have been able to get to this point.

Be it a runaway control board, etc., but I think you probably had one hash board which failed, pulled down the temperature limiting if it worked at all anyway, which I know mine has worked on a couple of S7s (I am using the latest firmware.) I know I said the S5+ handled the heat well at nice overclocks, but my S7s do not. They do not even run stock well at high temperatures. Ambients of 110 F require underclocks of my later batches. If I am unable to constantly monitor temperatures I downclock all S7s and S9s.  

Bitmain should be willing and able to give you a proper diagnosis. I would have thought the same...  My only response was "hi. Sorry ,burnt board are not cover in warranty ."

Responses in bold


NotFuzzyWarm, you certainly make a good point about incorporating circuit interruption for the 12V load side of PSU. I just don't know if it would be economically feasible though.  Interesting how we don't see much of that even in standard ATX applications with dual-GPU's, even though it is possible for a similar failure in a high-end gaming build for example. 

[Finksy]

Just want to bump this thread for another user who's experiencing the same bullshit as I did:

Hi.

I bought 2 of Bitmain's Antminer S9's - the 2nd batch. 12.93 GH/s

I plugged both of them in at the same time  - in the same room.
They both started up and seemed to be running ok for  a few hours.
The temperatures were around 65 degrees C.
Then I noticed that on one of the S9's, one of the hashingboards was reading 0 GH/s
So i restarted it - and then the board was no longer listed on the Status Page.

This was day 1 - within a few hours of being plugged in.

Now bear in mind that I have also had 5, Antminer S7's also running in the same room + the other S9 - ALL WORKING PERFECTLY.
No heat problem, no electrical problem.

Ok, so I tell Bitmain what happened and they tell me to send back the hashing board - then I get this response:



EXCUSE ME?

So I just paid $1800 USD for a broken machine and you tell me that "because it's repairable" that is somehow my fault?!
Are you F**king kidding me?

Then they refer me to the warranty:

"3. In some cases, the warranty will be void, but Bitmain can still do repairs at your own expense if requested. The following events will void the warranty:

     a. Customer removes/replaces any components by himself without receiving permission from Bitmain first;

     b. Damage caused by poor power supply, lightning or voltage surges;

     c. Burnt parts on hash boards or chips;

     d. Miner/boards/components damage due to water immersion or corrosion due to wet environment."

Ok, let's see..

(a) does not apply - they told me to remove the hashing board and send it to them'
(b) does not apply - I have 5 other Antminers all running on the same power supply - same room - same temperatures. THEY ALL WORK.
(c) I didn't see any sign of burining on the board - if there is I assume it's because there is something wrong with it.
(d) does not apply

So basically they are trying to screw me.
I am probably going to have to sure them, luckily I have a very good lawyer in Hong Kong. So I will let you know what happens.

But anyway - I personally WILL NEVER BUY ANYTHING from these con artists ever again and I seriously WARN anyone against giving these people any of your money - as they are not willing to replace faulty goods.

My Antminer S9 was faulty within a few hours of starting up.
How any serious business can pretend that that doesn't fall under warranty is beyond my comprehension - only a total asshole thieving lying scumbag would refuse a repair in such a situation.

So BITMAIN you can go and F**K YOURSELF. GOODBYE.

[Eyedol-X]

No disrespect or dismissal to anyone's situation here but here goes:

My .02:

I believe Bitmain should honor their warranty, especially if they expect customers to remove hash boards -- effective disassembly of a miner, to send back to China for repairs. Bitmain should just honor the warranty unless they can prove without a doubt the customer abused the product or clearly violated some terms of their warranty agreement.

If Bitmain has no proof or cannot provide proof of this violation, Bitmain should by default replace it at their cost for the sake of simple customer service.

My .03:

I would be just as upset as everyone here if I experienced the same thing with my miners and I AM upset to hear about these events happening. This type of business practice wouldn't be acceptable in the United States. However since Bitmain is in China AND pretty much has a monopoly on mining Bitcoin and anything else SHA-256, we as "small-scale" miners pretty much have to roll with their punches if we want to play in their pool.

I've been fortunate thus far to have no issues with my miners but I tell myself each time I purchase one "there is no warranty when I buy this thing" because every piece of evidence I've seen from Bitmain regarding their warranty says that I would be better off using the paper for bird droppings.

My Conspiracy Theory:

I think since the S9 is such a new and "limited production" item that Bitmain doesn't actually keep a reserve of parts in stock for your every day customers, they reserve the parts for the large scale customers and when they are out of stock, they simply "void" the warranty for the little guys knowing that all they will do is come to a place like this and complain about it when that complaint will really have no effect on their business, their profits, or their day to day operations because they are in China and they could care less what the rest of the world thinks of them.... they got their money $$$$ and who out their is going to regulate or question their business practices? The Chinese Government?! ... I will stop right there with that punch line, LOL.

Evidence of this conspiracy theory can be observed by the fact you never see Bitmain doing jack shit about threads or posts like this anywhere, even inside of threads they started themselves.

In order to change the view of Bitmain its going to take a majority of people to exercise the power to NOT make a purchase until something changes.... or its going to take some serious competition from another company to force Bitmain to go out of its way to win our money over someone else. Until then, they could care less about the "small-scale" miners of the world.

What I'm doing in my "Small-Scale" Mine:

At this time, I have no plans to purchase anything more from Bitmain. The only way I will buy anything with a Bitmain brand is if I find a "can't pass this up" deal on something that is used or 2nd hand.

What I am doing right now is focusing on alt coins and then converting those profits back into BTC. Sure it's more work but the power vs profit is better so far for me, long term I'll be shutting down my remaining S7's entirely in favor of altcoin miners that I built with GPU's.

Why I'm sharing this opinion and perspective:

• I share this because I hope someone that is new out there to the world of mining reads this and makes some decisions a little better than I did when it comes to first time investments in mining equipment
• I hope that maybe someone at Bitmain reads this and realizes that the future for miners isn't with Bitcoin and the Tide is changing.
  

[DRKMSTR]

I know what happened.

PSU Failure, the other units didn't die because \#physics

Okay, so according to OP, he was running it all off of a DPS 4kw unit.

Red flag-ish here (you shouldn't run a singular rail across multiple systems), but let's keep looking.

"3 S9's"

Um....red flag, that takes up over 3800 Watts by Bitmain's numbers....so you're running the PSU at 95.6% and some change. HUGE RED FLAG

You should never run over 80% full load capacity on any power supply for Bitcoin mining (Constant load), power supplies aren't meant to run at >90% continuously.

So that sucker was going to fail....you know how these guys fail? Voltage goes crazy, either way up, or way down. Now typically the voltage drops and most times, this can be caught by a computer, which will shut down, but not a Bitcoin Miner, no sir....you know why?

Because #efficiency, that's what. People wanted more and more on smaller boards running at lower voltages at smaller manufacturing sizes.....so, we have buck converters.


You know what they do? They jump up the voltage. Up up up!

So your power supply begins to fail, the voltage starts to drop and the power doesn't change, remember, those thirsty boys need 3825 Watts!

But wait! The lower voltage means something has to increase, and that's the amperage.

Here it comes.......12V @ 1275 = 106.25 Amps (35.41 per buck converter).....now 11V @ 1275 = 115.9 Amps (38.63 per buck converter)

Poor mr buck converter's startin' to heat up, his coils are getting toasty, singing the enamel, but it's holding.

10 volts

Crap, Mr. Buck Converter's coils are getting red hot, the enamel is burning off....HOLY...............

Poor Mr Buck converter.

The wires all short, now that 4000 W power supply just became an ARC welder, He's blowing through the amperage required to weld 1/4 inch stainless steel!!!!!

KABLOOM! (ARC Flash) consumes everything in it's path.

But what about your other miners?

Why didn't they die?

The answer is they likely experienced some damage to their coils, but the miner that died saved them, electricity follows the path of least resistance, once that one circuit board started dying, it took it all.  (Replace "Missile" with "Voltage" and "Giant" with "S9"....tears man https://www.youtube.com/watch?v=SXy3f6f9DxI)

Sucks bro, but it happens when playing with fire. I used to run stuff near 100% until I learned about the 80% rule, burned up a PSU and nearly lost my motherboard. Sucks you have to learn this way though.

[sidehack]

Not sure what all those "#" are for, but it makes me read your reply in the voice of a teenager.

Buck converters do not jump up the voltage.

When you talk about the bucks frying, do you mean the bucks in the DPS2K or in the miner? Because I've load-tested a DPS2K at 196A total output for seven days in July heat with zero problems - not bad for a PSU rated at 165A. I've had them running 110% rated output for six months straight with no failures. They're friggin' tanks. If you mean the bucks in the miners, well, the output current is completely independent of the input voltage; the RMS inductor current can change a bit as you shift the input voltage closer to the output voltage but I think it actually decreases the peak inductor currents so the RMS actually gets more favorable and less heat is generated.

[Tupsu]

ˇHi,TupsuThe following is your order of scrap equipment:03320160813070114259mCdrl8rX06EC

RepairDetail ID 03520160813070114279oxTnD8GD06AC
    Name Hashboard S9(14T 650M)
Damage Causes    PIN is lost in circuits board

What the hell does it mean ?
Damage Causes    PIN is lost in circuits board

I have pictures taken before shipment.
Do not see anything that would be missing.


View original size - Click on the picture

View original size - Click on the picture

Link
http://nagi.ee/photos/photo_sizes.php?id=24747333&size=o
http://nagi.ee/photos/photo_sizes.php?id=24747331&size=o

[philipma1957]

ˇHi,TupsuThe following is your order of scrap equipment:03320160813070114259mCdrl8rX06EC

RepairDetail ID 03520160813070114279oxTnD8GD06AC
    Name Hashboard S9(14T 650M)
Damage Causes    PIN is lost in circuits board

What the hell does it mean ?
Damage Causes    PIN is lost in circuits board

I have pictures taken before shipment.
Do not see anything that would be missing.


View original size - Click on the picture

View original size - Click on the picture

Link
http://nagi.ee/photos/photo_sizes.php?id=24747333&size=o
http://nagi.ee/photos/photo_sizes.php?id=24747331&size=o

Your board looks clean as a whistle. Same as mine.
Mine has writing on one label as if the factory rejected the board.
Your label was not written on but your item truly looks good.

[sidehack]

Have they ever explained anywhere what that PIN is? I would guess it's a MacGuffin they made up to have a universal justification for denying warranties in otherwise passable cases.