Topic: We Ordered and received a new mining box. New photos as of Aug 26 (Read 491 times)

videos of box getting closer to it working.

https://www.youtube.com/watch?v=S_3Svub_TP0

https://www.youtube.com/watch?v=LnjSK0gldUk

okay the warehouse has a large garage door that opens to the outside.

So the humidity will match Clifton NJ humidity.

well under 45 percent most days.

But some summer days we go over 60 percent which means the water curtain will suffer a bit.
We rarely go over 70 percent humidity.

 which means the water curtain will work well over 300 close to 330 days a year.

This unit will move 40,000 cfm

and we do not have enough power to run over 75-80 s19s even though the unit claims to do 150 s19s or L7s

we have a 320 k-watt transformer to feed it which will allow us to do 250k-watts.

My gut tells me on a 95f day with high humidity we may need to down clock the gear just a little.  That means 5-15 days a year it could struggle

What is the usual ambient relative humidity there?

box is now at a height where the venting clears the concrete foundation.

In the current room we have
3x 220 amp panels giving a steady 225-235 volts.

they come from an 800 amp main panel doing 480 volts.  that feeds to a 160kva transformer. which feeds the  3x  220 amp panels at 225-235 volts.

we are looking to feed the box from the 800 amp 480 volt panel. and eventually shut the current room off.

I rather have a few smaller service panels than one huge panel, Our setup is 2 main 4-pole 1000A circuit breakers, each feeds 2*400 control panels (400A contactor attached to phase failure + voltage protection relays as well as SONOFF smart switch) and then each contactor feeds a 400A service panel which consists of 4*100A 3P MCBc (we run the netural directly from the busbar to the PDUs) each 100A supplies a 125A PDU so we have 4 PDUs for each 400A panel.

So despite having a total of 1600A on each phase, we don't have to deal with the complexity of heavy setup, just 4 small 400A service panels that we can control and shutdown individually, Keeping a perfect load balance between all of them isn't the easiest thing in the world, we were never able to keep them near even since some miners run at different clocks, some have dead hashboards, some would reboot some would not, but we were within 5-10% range most of the time.

yeah we are in talks with setting up a set of 4 Tesla superchargers.

This will take us over the 320kva transformer.

State of NJ will subsidy it.
Power company would like power sales for charging cars.

May or may not happen that is the if part.

But using what we have now to the max possible (the 320kva) is going to happen.

Which will be nice.

Will be interesting see how the power company handles adding a 2nd xmfr. I can't see them just strapping 2 together in parallel and running a single set of much heavier lines coming out of the pair. Parallel feeds to them from the utility HV lines sure, bonded together as a single output - no.

No matter what the incoming service panel would of course have to be upgraded and it would/will probably be cheaper and certainly easier to just have a 2nd service panel to distribute incoming power from the 2nd xmfr. Of course you will need to monitor the 2 service loads to they stay within their power limits. As I recall you have some decent 3-phase power monitors in place for that  

With the incoming lines from the transformers to service panels being independent of each other, balancing the load between them is not really an issue. Just 80% max out your existing transformer and add future loads to the new xmfr distribution panel or shift some loads around if desired.

Thanks for the explanations. After consideration I realised that MW/day was probably what was being discussed but thank you for confirming.

Mind blown! Thats one heck of a set-up, good luck with it.

I'll keep an eye on this thread to see the progress.

cheers.

G.

yeah mikeywith has it right.

we backfeed to the grid 280-60= 220 .

so the 280kwatt does around 1500k-watt  in the summer maybe 1200k-watt in the winter.

so 60 k-watts is used 24/7/365 and the grid is the battery which stores the energy.

we have around 700 x 400 watt in panels. giving the 280kwatt peak number.

the main transformer is 320kva

so in the day we can do around 250kwatt  from the transformer and 60kwatt from the solar.

At night all power runs from transfomer so the max night run is about 270kwatt as no other use the transformer..

this complex is two or more acres and a few buildings.

a second mining room is across the street and does 50kwatts.

we will need a second transformer in parallel  say 320kva + 320kva

for now we are doing 160-190 kwatt in the solar/transformer grid setup

and 50kwatts min in the building on the other side of the street.

we will keep the small room with gpus a few whatsminers and l7s

the box room will gain 60 to 90 kwatts of gear. all in the box.

this will max us and we will order a new transformer to run in parrallel.


best case is

250/260 in the box

50 in room on other side of the street

all no later than oct 2023..



if that all that happens in summer of 2024 we would get the second outdoor tranformer.

giving us 500 kwatts in the box  and maybe 50-60 kwatts across the street.

all that is a maybe 🤔 but

the october numbers should be good.

Ya i noticed that too, phill forgot to write the unit

instead of


No water as in liquid form would be coming out from the other end of the evap cooler, it will be cooler air with more humidity, as long as the miner's temp remains above dew point condensing won't happen, running evap coolers in humid places could be somewhat risky and would cause damage to the miners in the long run, but if the humidity levels are moderate or low, there should be no issue at all.

One thing you need to be careful about would be having a faulty internet connection while using custom firmware like Vnish, The devs of those custom firmware wanted to outsmart Bitmain by saving energy during internet outages, so they keep the fan spinning but the miner doesn't generate any heat (unlike the stock firmware), now if you got those evap coolers throwing in air with high humidity, and your miners' fans spinning, the surrounding of the miner fan will go below dew point and then kaboom.

Thanks for the info, nice deal. I can imagine industrial freezers chew up quite a bit of juice so exchanging it for a mining farm is a good one.

I'm a bit confused, you say your solar does 1.5MW for 20 S19 but grid supplied transformer 320KVA (~320KW PF1.0) is 90 S19's. Something's not quite right here.

Anyway, 500KW my head hurts just thinking about it, thats 1000 times more than my average home usage.

 

our solar does 1.5 megawatts a day. that is about 20 s19s nonstop.

our on site transformer is 320kva that is about 90 s19s nonstop.

the 320kva transformer is the master limit. So even though the box can do 150 s19s and even though the solar can do 20 s19s our practical limit is 90 s19s

we are in talks with the power company to add a second 320 kva transformer to run in parrallel.

We have a longterm prepaid contract for power.
The warehouse owner had purchased 5 years of power for the icecream business he was renting freezers to.
the icecream place went out of business and he was stuck with a prepaid power contract and was not using the power. we stepped in dec 2018 and had 4 s9s. we now have quite a bit more than that.

This may be a silly question, but how can you afford 500KW of power with todays energy rates do you have your own solar farm and power station as well?

Its mind boggling! 

Yeah I think it is 42 inches.

As for the moisture question if it is an issue we will put a layer of material like weed screen on the inside of the box along the water media. I also suspect at the rate we fill the box and when we get it running.
Say Oct 15 2023 we may not need the water curtain till may.

The fans are beasts 10000 cfm each so 40000 cfm will be far more than we use now. Say 22000 cfm

this should be fun project. we will grow a lot of hash from now till june.

this just shows my ignorance but with that kind of airflow from the water curtain so close (a few feet looks like?) to the shelves the asics will live  on that there is no water mist that could get sucked in. but that box looks to be pretty well thought out so obviously it must work somehow.

when you say you might raise the box to above the level of the concrete/cinder block wall are you talking a few feet (looks to be 1/2 way up that door so 3-4 feet)?

one of 8 pdu's



breaker box




4 shelves 8 plus in all


we may lift it higher than the concrete wall to avoid cutting it. (the wall)


the media for the water curtain is under the white intake slats. there is an aluminum screen to prevent bug damage to the media.

2 Side door shots and a view of the current asic room behind the cut-out people

---

---

---

water curtain one and two on the right also power cabinet on the right
shelving on the left the edge of the PDUs are visible



top shelf and first of 4 fans

second shelf same fan

yeah between the water filters and the aluminum screens the water media sponge should last a year or more.

this will be learning process for us.

hey your braiins+ is doing great on one test s19 pro

set to 2475 watts and it does about 95-98th runs cool.

I will be burning more cards and setting up a few more as I have 4 s19 pro 110th machines

I even saw some using fabrics, what they do is have a screen fabric in place a few days then switch with a second clean one they have ready while they clean the first. Its done weekly or something like that. There are also "proper" (large) filters they use for like industrial HVAC systems in data centers etc.