Why would ASICs spike or not spike ETC?  The E3 has been announced and will be here soon. 

I would argue that ASICs would crash the price since the SUPPLY of ETC will massively increase once ASICs are online.  What is the counter-argument for this rationale?

UPDATE: I put all 3 miners on the 208v power and they are mining exactly as when they were on the 120v.  However, the power cords are cool to the touch instead of warm, so there are definitely fewer amps flowing through it.  I suspect that the power supplies might be more efficient using the higher voltage as well, but there's no way to confirm that. 

Yeah, I'm not sure.  I have special cables ordered so I can put them all on the 208v just to be safe. 

They have August, August, and June 2017 stickers on them if that means anything.  The power is very steady at 117.2v and the Kill-A-Watt I have says the PF is .99 and Hz are 59.99.  There is really no fluctuation in the readings.  Maybe it's because the power is so consistent? 

What I'm saying is that just because there is no Windows wallet, the price is supposed to drop astronomically?  Makes no sense to me.... There must be another reason for the decrease.

Not the best angle, but here is the sticker: https://imgur.com/dGCJrkV.jpg

I see, thank you.

I just wanted to ensure that an unbalanced phase load would not hurt downstream components like the transformers inside the AP3++ or even worse, the S9's.  I think I'll be fine running all 3 on the 208v bridged legs. 

The 110 socket is reporting 117v on my voltmeter.  I would like to power them on the 208v sockets, though, since I'm sure it's better for them.

The PDU supplies 110 to the outlets.  There are 21 outlets and 7 are on L1, 7 are on L2, and 7 are on L3.  Each line is a 110 hot just like a normal circuit.

The 6 big outlets with twist locks are (A) two outlests for L1-L2, (B) two outlets for L2-L3, and (C) two outlets for L3-L1.  These outlets bridge two 110's together.  Bridged, they provide 208v.  That is how the PDU can supply both.  On its spec sheet, it says "Nominal output: 120v, 208v."

Using the 20A tab in this spreadsheet, if miner A and B are operational and miner C stops for some reason, the line loads would be L1: 6.8amps, L2: 11.8amps, and L3: 6.8amps.   Would this phase imbalance hurt the miners?

Is there any danger in running the L1-L2, L2-L3, L3-L1 outlets?  It's 3-phase and I have heard horror stories about how unbalanced loads can destroy transformers and other components in power supplies and computer components.  I worry about having one miner stop for some reason and the load hurting the other two or some combination thereof.

What errors should I be receiving with the 120v input?  So far I've been mining for 5 days like this and the hashrate's been constant and I don't see any obvious errors under the Miner Status tab in each S9.  I've attached a screen shot of one of the S9's status, they all look pretty much the same with similar numbers and they all have "oooooooo"s.

https://imgur.com/BViz3pg.jpg

---

(Moderator's note: This post was edited by frodocooper to remove inline image tags.)

Thanks for the reply.  So far, there have been no errors on any boards and the hashrate has been pretty consistent at 13.6TH.

The AP3++ PSUs say on the sticker on the side that they run from 110v-240v.  Is it possible that there is a version of the AP3++ that works fine on 120v?

I have 3 Antminer S9's each with the Bitmain AP3++ PSU running on an APC 7863 PDU.  This PDU has 21 NEMA L5-20R 120v receptacles (normal household-looking American 120V outlets) and 6 NEMA L6-20 208v receptacles.  It has an L21-20p male plug connected to a 3-phase 20 amp circuit in a commercial building. 

Currently, each miner is plugged into a separate leg on the PDU using the standard L5 120V connectors.  Each L5 connector indicates its respective power leg to make it easy to balance the load.  One S9 is in an L1 labeled outlet, one in an L2, and the last one in an L3.  Each leg is reporting ~12amps on the PDU's LED display and the miners are getting about 13.6TH/s. 

My question is whether or not it would be of any benefit to use the beefier L6-20 connectors instead of the L5 connectors.  The L6 connectors are labeled "L1-L2," "L2-L3," and "L3-L1."  The output voltage of these connectors is 208v. 

Would I get any efficiency gains buy using three 208V outlets vs three 120v outlets?  I believe the amps/leg would stay the same.  Is it dangerous to use the 208V connections for 3 antminers in terms of power balance (ie. one miner shuts off and then L1-L2 and L2-L3 have miners, but L3-L1 doesn't... and vice versa)?  I am a total n00b as far as 3-phase power is concerned and this is my first foray into using it. 

I believe:

current setup: One S9 on each 120v leg (L1, L2, L3)
                     1400w / 120v = ~11.67amps per line

Contemplated setup: One S9 on each 208V line pair (L1-L2, L2-L3, L3-L1)
                               1400w / 208v = ~6.73amps per line pair
                               6.7 ÷ √(3)  = 3.89 amps at 3-phase
                               3.89 x 3 = 11.67amps per line

Questions:
Would there be any benefit to using the 208v plugs?  Is there any danger to using the 208v plugs in my setup?

Any help is appreciated, thanks!

The lack of a wallet seems like a ridiculous reason for such a price drop.  Something else must be at play...

Good points.  I am mining ETH myself so I hope you're right.  I'm just pointing out the fact that .1BTC or even .07BTC is historically high for ETH.  Be prepared for ETH to possibly revisit lows in the .025-.03BTC range until the market sorts itself out.   If BTC revists support levels of $6,000 and ETH is at .025BTC, well then ETH = $150.  I can't stay in business myself at that level, BUT it would be a huge buying opportunity.

Also, I mine ETH and sell it t current market value to pay my bills.  The market value is NOT set by the miners, but rather by supply and demand.  If miners/manufacturers in any market set the price, what would physical gold and silver actually be selling for today?  How about oil? 

This is true, but the ETH selloff has been due to the SEC in the USA issuing strong words about ICO regulation.  Since many ICOs are made possible by Ethereum, people are getting out.  ETH is just an alt, like any alt it is measured against BTC.  GENERALLY, When BTC goes down, alts dump... When BTC shoots up, alts dump (in BTC terms).  ETH is no exception.  So, couple the ICO regulation and BTC falling and it's a double-pronged attack against ETH.  

Look at the long term ETH charts.... It's still kind of highly historically.  $150 seems like a good spot to accumulate ETH again when it gets there.  Will ETH be up to $2k one day?  Maybe, but then BTC would likely be over $30k too.  

Indeed.  Well 2 RMA out of 96 flashed cards isn't terrible I guess, but I bet I'd be at 100% success if I did each one individually. 

They were bricked by mass flashing ROMs onto the cards.  I flashed about 10 cards with the same ROM and these few got bricked.  The ROM was one I edited myself from the saved stock BIOS of one of the cards --all cards are the same model. 

Yes, after trying everything, I think I will RMA the 2 cards.  Yes, I am running windows on the BIOS flashing machine, but the cards will go into a Xubuntu miner.   Bridging the pins has worked for a few cards, but usually the fans are spinning on a card that is able to be unbricked.  Hopefully the manufacturer will look past the fact that incorrect BIOS's  were flashed onto the cards!   

From now on, I will edit each BIOS individually to be safe. 

Got the one card unbricked!  The one with the fans that spin was able to be unbricked by bridging pins 1+"5"

Everyone seems to either be mislabeling what 1+8 means or some cards' pin #8 is different.   But I found a few posts and images showing that pin #8 is directly across the short side of the BIOS chip.  That is, the pins are counted counter-clockwise from pin #1 in numerical order.  Only 2 cards left to go!

I have 3 bricked cards...  All three of them are RX570's.  On two, the fans do not spin and they are totally unresponsive.  On the third, the fans spin, but that's all it does.

I have tried:

-Atiflash GUI in admin mode in Windows 10
-"atiflash -i" and "atiflash -p -f 0 bios.rom" in admin mode of CMD in windows.
-atiflash (same commands) on a DOS USB stick.
-bridging pins 1+8 on the BIOS chip on the back of the card and then turning on the computer and holding the bridged pins throughout running atiflash commands.

When I try "atiflash -i" in either DOS or CMD, it says "adapter not found."  The computer obviously can boot with the cards, but no program (atiflash, GPUz, windows device manager) can see the cards.  I have held pins one and eight together many times through several power on cycles. 

I am at my wits end with these cards.  I have unbricked a few cards previously using atiflash, but these three are a pain in the ass.  Does anyone have some suggestions or tricks I haven't found that could help?

Yes that is true, but then the wattage would be too high.  I have 6 miners with 6 cards each on a 16amp circuit.  With everything running, I'm using about 5,100 watts at 14.17 amps.  Thus I have less than just 2 amps left before the circuit trips so I do not want to push harder.

OKAY.  Tried a fresh reformat and it didn't work.  However, I then rolled back the kernel to 4.4 and uninstalled, then re-installed the AMD drivers.  This seems to have solved it.

Doing this, I was able to get the card running at 23.4MH/s which is acceptable.  I have 4 cards on risers now running all at 23-24MH/s.  Tomorrow, I'll try adding a 5th and then 6th card and see if I can get a reliable 6x570 setup on this new chipset.  Thanks for all the replies guys!

I have it on a powered riser now, still no luck .

Anyone else out there using this board?  I have 12 rigs of 6 cards each running flawlessly on Z170-AR boards and this fucking Z270 is causing me to pull my hair out!