ArtForz; You're absolutely correct.  So really, we're talking about a max of around 10A total on the 12V rail that the motherboard can provide..

ArtForz also shows an interesting part of hardware design - it's INCREDIBLY easy to miss a critical specification and design something that doesn't work because of it

Ask me how many times I've accidently laid out a part package for a PCB and not realized until after the boards were fabbed that the datasheet had in really tiny font a note that the package view was from the bottom instead of the top

Enigma

The plastic ones are alright, I just think the metal ones are overall better-made and usually have a faster return = more suction.  Especially for pins connected to thermal planes like these, once you start moving air, you start cooling down the joint.  The faster you can evacuate the solder, the less chance there is of it solidifying...

I hate de-soldering things, so I make sure I have the best tools I can possibly have to make it a little less painful...

With a good iron (Metcal is my personal favorite), a good solder sucker and a big solder tip, I kid you not - I could de-solder that connector in under 5 minutes... All 24 pins..  Less than 2 minutes later, the new one would be soldered in..  And, really, I'm probably being generous.. 5 minutes would probably cover the whole replacement...  Good tools make all the difference.  Use a good MX Series Metcal Iron for half an hour - you'll never go back to anything less..

Don't trust mouser's data (regarding 6A).  I just confirmed what I thought I remembered - Mini Fit Jr. pins are rated for 13A (MAX); per Molex.  That means about 75% (9.75A) is a nice safe number for 24/7 continuous operation.  Anything more, and you risk exactly what you have there...

Enigma.

In the very first picture - the ASUS board with the two burnt up pins.. I suspect it's not poor contact/arcing - it would be really unusual for two pins in one connector to be loose.  These are basically molex mini-fit jr. style connectors, and, if memory serves, they're rated to 13A per pin.  Those two pins are both 12V+, and in fact, they're the only 12V+ on the 24 pin ATX connector.  It's more likely that you're actually pulling more than 26A (312 Watts) and that they burnt up due to over current.  13A is MAX contact rating, so 10A (240 Watts total) is a safer number for continuous operation.

It will probably happen again, unless you power the cards with a separate power connector.

Enigma

P.S. To de-solder the female connector on the motherboard, you'll need a pretty good soldering iron.  Most of those particular pins are (most likely) connected to a power/ground plane on an internal layer, and that will sink a LOT of heat.  Use a big tip (biggest you have), put some no-clean solder on the tip of the iron to get it wet, and then set it on the pin and wait.. It will eventually go molten.  Use a solder-sucker to remove the molten solder.. Repeat for each pin.

Replacement connector here: http://www.mouser.com/search/ProductDetail.aspx?qs=g5f4E6w8l8gzEjxCgQ0R8g%3D%3D
Solder sucker here: http://www.mouser.com/ProductDetail/Apex-Tool-Group-Formerly-Cooper-Tools/7874B/?qs=sGAEpiMZZMt%252biMJH3c40u7BCFiF%252blJXomEfWxnrcMeE%3d

This is the last time I'm personally going to get involved in this discussion..  It's over in my mind and ztex stands as "NOT a scammer", to me anyway..


The specifications state that these boards may be operated from 4.5 to 16V.  WHAT MORE OF A WARNING LABEL DO YOU NEED?

Danger, if exposed to a bath tub, ztex boards may no longer function.

Caution, if exposed to a lightning strike, ztex boards may no longer function.

Warning, ztex boards were not designed for use at temperatures like those found at the center of the sun.  Operation inside of the sun is not guaranteed.

Notice, do not take ztex boards swimming with you.




You've really gotten to the point of hysteria with this whole topic.  NONE OF THE PRODUCTS YOU OWN are protected against over voltage - Literally NONE of them.  Hook your 30, 40, 50 thousand dollar vehicle up to 110V and watch what happens (it's pretty awesome, by the way).  Hook your cisco/linksys/d-link product up to reverse voltage or double voltage and see what happens.

ztex very clearly states that the boards operate from 4.5 to 16V.  I don't see how the specifications could possibly be any more clear.  It's like a vehicle tire - they have a speed rating - S, for instance means 112MPH.  There is no further warning.  The speed rating tells you that it's safe to operate the tire continuously at 112MPH.  At 113MPH, the tire may explode, but there isn't a bright orange warning label on the side of the tire..  It's been stated, and doesn't need to be stated again..

I'm glad ztex added a wiki entry about power supplies, but I will defend his old specifications - they were plenty clear.  He simply can't be responsible for you providing power far outside the specification - which from the burnt chips, you clearly did.

If you'd like to PM me, I'll even let you send me the power supply in question (3com) and I'll tell you beyond doubt whether it hurt the ztex board or something else did.  I have hundreds of thousands of dollars worth of test equipment at my disposal, and I will be able to see if that supply is dangerous to the ztex power supply or not.  Beyond that offer though, I'm ignoring this thread.

Enigma.

I really believe a customer should use a reasonably good power supply on boards that cost 500 bucks each..  I also believe a customer should be directed to a safe solution when they ask, not a cheap one.  I never try to break my customers' banks, but I also suggest they spend money when it makes sense to do so.

As DeathAndTaxes mentioned, the real cost of "Protection" isn't in the manufacture of the board (their bucks) but in Watts (customer's bucks).  "Protection" can be as simple as
a diode that prevents reverse bias (3 cents)..

or get more involved, and use
a zener clamp diode (19 cents) http://www.mouser.com/ProductDetail/Vishay-Semiconductors/BZG03C16-TR/?qs=sGAEpiMZZMstCHp3EWKGl1tP683WM1O433kxgBzanSM%3d
and optionally
a PTC (40 cents) http://www.mouser.com/ProductDetail/Littelfuse/2920L200DR/?qs=sGAEpiMZZMu7EFbsM1w0nTP0ee3bziE2pcZdmj40k18%3d

So leaving it off isn't to save money, it's to save power (increase efficiency).

Now, there is a way to add protection without really costing power, and that is to use just the zener accross the power input with a PTC (no reverse protection diode).  Interestingly, because of the way a zener works, it will still protect against reverse bias (kind of.. You can still get about -0.6V before the protection kicks in.  Some circuits can survive that, some can't.) but it won't be inline burning power and creating heat.  That may be a smart thing to add, if in fact the circuit can survive -0.6V (or be made to survive it relatively easily)...

Either way, MOST "boards" that you buy do not have protection.  Motherboards?  nope.  Video cards?  nope.  Home Routers and switches?  nope.  DVD Players and Xboxes?  nope.  I'm actually a bit hard pressed to think of something you might regularly buy that IS protected from over voltage and/or reverse bias... having a hard time..
The manufacturers of these products expect that their customers will use an appropriate power supply, and therefor do not need protection.  The caveat, of course, is that in most cases the manufacturer provides the power supply that the customer should use...

I won't agree or disagree with that statement - here's what I will say..
  If that power supply is putting out a peak of less than 18 volts (I'm not talking the average of 14.5 that you Hioki is displaying) than it DID NOT hurt the board.  Something else did.  You?  Maybe.  ZTEX? Maybe.  Shipping? Maybe.

There is no way for you to know what the peak voltage of that supply is without an oscilloscope.  Your Hioki is either an RMS or Averaging meter.  A noisy supply that peaks at 19.5 but troughs at 9.5 will be displayed as 14.5 on an averaging meter.  Similar results for an RMS meter..

Rundown:
     Is the ztex board poorly designed?  No, not in my opinion.  It has a well designed regulator with a very reasonable input range.
     What voltages can it tolerate? 0 - 18
     What voltages can it operate at?  4.5-16
     Is that 3Com power supply a piece of shit?  Probably
     Did the 3Com supply damage the ztex board?  Absolutely no way to tell without an oscilloscope
     Should ztex boards be protected?  Probably not.  It costs watts and money and you can never protect against everything anyway.
     Is that 3com switch protected?  God no.  Plug the wrong supply into that and see what majick smoke comes out.

     What can be done?  Use GOOD quality power supplies on your 500 dollar boards or at least use a cheap supply that is way under volted.  A 6V 'cheap' supply is not going to put out 18V under any normal circumstances.  If it fails, it might put out wall voltage (110/220) - but under normal circumstances, it's going to be 'safe' for the ztex board.

Enigma

I'm convinced after looking at the circuit that the power supply design is quite good, and that there is absolutely no protection.  Most devices do not have protection because in most cases, protection costs watts.  I'm also convinced that the input of the power supply will not tolerate more than 18V for brief periods or more than 16V for extended periods - hence the reason I suggest using a 9 or 6V supply IF you're going to use cheap crappy supplies.  If you're going to power it with a good, clean, accurate, switching regulator - feel free to use one at 15.99998V if you so desire - the board will survive and operate just fine.

Enigma

On the ZTEX Boards, If you're going to use a cheap supply, just use a 9v one..
Even at worst case, 9v is only going to be 12.73 volts peak - still leaving PLENTY of headroom..

If you want to be ultra-cautious, use a 6V supply..

I just checked over the schematics of the 1.15x product that your'e talking about...

From my perspective (several decades of analog and digital circuit design, EE), there are two possibilities if in fact the power supply is 'blown up'.

The supply is capable of handling input voltages of up to 16V:
     references: http://www.aosmd.com/pdfs/datasheet/AOZ1025DI.pdf  AOZ1025 Datasheet (1.2v Supply)
                     http://www.aosmd.com/pdfs/datasheet/AOZ1021AI.pdf  AOZ1021 Datasheet (3.3v Supply)
                     http://www.ztex.de/downloads/usb-fpga-1.15x.pdf  1.15x schematic, where a note is actually included on page 4 near the power connector (Input Voltage: 4.5..16V)

So, even if your supply is putting out 13.X volts, it's not really capable of hurting the power supplies.  It's possible that he means "blew up the capacitors", but I really doubt a guy as smart as ztex would use caps not rated to AT LEAST 16, and probably 25 Volts on the input.

So, as said, I suspect two possibilities here:
1) There was a real and significant problem with the board, and the output of the supply or supplies was essentially shorted.  Assuming ANY level of quality control, pretty unlikely.
2) You plugged in a power supply that was putting out > 16V (knowingly, unknowingly, admitting or not admitting).

If no shorts or problems can be found with the board, and he told you using a 12V supply was safe (he can't assume that you mean a 12V supply that isn't a 12V supply, by the way, and 13.5 isn't 12...), I wouldn't warranty the board either.  Sorry, but I wouldn't...  Now, I would offer a reasonable price for repair and retest, but that's me..

I strongly believe that the board was fed significantly more than 12, 13.5 and even 16volts..  I've worked a LOT with smps parts (although not the AOZ102X in particular), and they can usually stand quite a bit more than their rating for a short period of time anyway... In fact, the absolute maximum of the AOZ parts from their datasheets is 18V, meaning they can survive that for at least some amount of time...

Unless their is something else to this story that I don't know - I really don't blame ztex for not warrantying the board for you.  Should ztex provide (or at least sell) a compatible power supply - yeah, probably a smart move.. but He told you to use 5-12, you admit you used 13.5, and it won't actually fail until 18ish.. Something isn't adding up here...

Couple of questions:
  When will be the next batch of Icarus?  I actually have some use for these in a non-bitcoin-mining application (and can use that to justify the expense) - but I would then be able to use them for bitcoin mining whenever they're not in use (free mining hardware )

  For my application, I would need 5 boards.  Will the next batch be larger?

  Also, I see that you don't currently support chaining the boards for mining (I'm assuming that is what the P1 and P2 connectors were intended for?).  Is this a hardware 'issue' or a software one?  Will chaining be supported in the future so that I don't need 5 USB Ports to hook these up to a small PC?

Thanks!
Enigma

Not really sure.. It's not binary..

In Binary, the current difficulty is
100001010010010011011

Not hex either.. Thats
10A49B

1.5hgz - you're correct, EXTREMELY unlikely to get those speeds for SHA-256 style routing. 

Until it's released, the errata sheet is checked over (and there WILL be one), and pricing is confirmed - this chip is pretty much not worth talking about.

Just for reference, The Kintex-7 we're speaking of would have to manage 2025MH/s to beat the Spartan-6 LX150 in terms of MH/$

It's actually a little lower than that, since you only need one (bigger) power supply for the Kintex compared to 10 power supplies and support chips for the 10 spartans, but that's a rough estimate of where it would have to be to compete with the current Spartan-6 designs.  Can it get there... My gut feeling is no..

It will probably be able to beat the Spartan-6 in power consumption, but really, the spartan is already so low compared to GPU designs that it doesn't really matter.  Additionally, we're talking about such low power consumption that even a significant percentage of power saving would take a long time to pay off the more expensive chip.

Enigma

I haven't really looked all that much at the 7 series FPGA chips yet, so I can't speak very intelligently about them - but in general, I wouldn't expect the VHDL to be much different at all, and would assume a hashrate for that chip of somewhere in the neighborhood of 1000-1200 MH/s with good placement and resource usage.

These are ALL assumptions though.  Take those numbers with a very large grain of salt until someone takes the time to cram some cores in and synthesize a design on the new chip..

Edited to explain my math:
I'm guessing that 5-6 fully unrolled double SHA256 cores can be stuffed into the Kintex-7 with that number of LUTS (Guess Number 1)
I'm guessing that even in the new architecture, it's going to be damn difficult to place those cores with much better than 5ns delay = 200Mhz (Guess Number 2)
I'm guessing the new architectecture doesn't allow any magic that I don't yet know about/understand (Guess Number 3)
At one hash per clock, that's 200MH/s/core = 1000 to 1200 MH/s/chip

As you can see, lots of guesses.. but they're fairly educated..

The Kintex-7 Chips are fairly similar in price to the Virtex-6 Chips, but have about 50% more LUTs for that cost..

Digikey has a few of the chips in stock
http://search.digikey.com/us/en/cat/integrated-circuits-ics/embedded-fpgas-field-programmable-gate-array/2556262?k=XC7

Based upon that information, I would expect the KC705 to be similarly priced to Virtex-6:240 Evaluation boards

Enigma

Our company could do it for you - but we rarely are the lowest cost provider.  If you're planning on offering a warranty on what it is that you're building, make sure that the assembly company will be X-Ray inspecting the BGA.  Without X-Ray inspection, I wouldn't warranty anything with a Spartan 6 for longer than 12 hours.

Enigma

Could you please let me out of here?  I've been following the FPGA Threads in the mining section with great interest, and being an Electrical Engineer with a background in FPGA and RTOS development, I would very much like to add to those discussions.

I like the idea of the Walled Garden that "Newbies" are placed in to prevent SPAM, but honestly, I just don't have much to say in the "Newbies" section - Posting for the sake of posting seems.. pointless..

Thanks!
Ben

I AM NOT A TROLL OR A BOT
I AM NOT A TROLL OR A BOT
I AM NOT A TROLL OR A BOT



Seriously though, I actually wish some of the other forums I participate on would implement a rule like this 5 post/4 hour rule.  It gets pretty old having spam and trolls take over a forum..

Hello everyone,
I'm Ben.