Topic: Running Miner off of Deep cycle batteries - expected drain? (Read 2922 times)

I'm unable to sell back also. As long as your load exceeds your maximum production potential, at all times, you can use grid tie and never back feed. That is how I can run solar gear and grid tied inverters with no contracts, permits, inspections, and barely any install costs; by doing it myself super cheap. Everything has to be synced so that power is coming from same phase in breaker box as solar inverters are feeding into, now power company never sees the solar production, just the varying constant load from the miners.

This is good info thanks. The only issue is selling power back is not an option at all. So I'm left with battery and simply having downtime for powerless hours.

The battery only serves to store what the miner doesn't use during that day off the solar. It could run maybe a few hours after light and then I'll need to switch over to wall power. A arduino unit controlling a physical relay switch could handle a task like this.

Yes I think that is a good plan. Also think an Arduino is a good idea, when I was thinking about a similar setup I had in mind monitoring the voltage at various points down the chain with the Arduino and then with some changeover relays and a simple program to get the best from the setup.

Rich

I'm running 10kW of mining gear and ~5kW of grid tied solar. Solar supplies about 10-15% of total daily consumption, in Florida. Realistically, I would need more than a 50-60kW solar array, at minimum, to cover 240kWh daily use. So you either need a shit ton of expensive battery or are going to be grid tied, back feeding to power company most of your generated power, only to buy it back later at a higher price. So you need at least ~6X in solar what your intended load will be if located in a 4 sun/hr geo zone.

Yes this is what i'm thinking, except with the addition of a direct loop from the panels to the buck converter to bypass charging in the day cycle. If the batteries are low for some reason (bad math on my part or a few shitty days in a row) I would want the Panel to prioritize powering the miner over charging dead batteries.


So the circuit would resemble this:



Solar Panels >>>> MPPT Charge Controller >>>> Battery Bank >>>> Buck Converter >>>> Miner
                   | |                                                                    |   |         
                  >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

This gives the buck converter the ability to run off battery or pure solar. This might require a second buck converter in parallel with the existing. I could also run a Arduino post second buck converter and write a simple python program to cut power to the miner in the event its running low on power from both sources, this prevents unexpected power loss and unintended low voltages going to the miner.

I would keep the overall setup as simple as possible using some off the shelf modules from Ali Express.

Solar Panels >>>> MPPT Charge Controller >>>> Battery Bank >>>> Buck Converter >>>> Miner


Rich

You;ll need to rent/hire a forklift to move them 

Indeed. Forklifts have massive counter-weights on them so they don't tip over when they lift a heavy load. Electric fork-lifts tend to use these heavy batteries to their advantage.

Ja they are massive power -- which in turn means massive weight. As is several hundred to over 1k Lbs each so plan on needing an engine hoist or other heavy lift device to handle them...

A few people mentioned this to me in the past too! Forklift batteries are usually a pain for any company to have to deal with ;P   I could def take them off their hands.

If you can make a purpose built building for battery storage, look into some Good used/Refurbished forklift batteries, Extremely durable if properly maintained. They are huge, heavy, clumsy to move, but they should last forever in an application like yours with a good charge controller and proper battery maintenance. Frequent battery replacement will negate any cost savings made by solar.

Go big!

I have one of questionable strength/life left in it that i acquired for nothing. Can't find any marking or stickers on it but it is a 36v battery with a few dead cells, you can bypass/hack off terminals to the dead cells and make it a 24v battery. That's how i have mine set up, it's @ my friends hunting cabin in Mexico. I have a cheap $40 30amp charge controller and 5x 100 watt rated panels(supposedly) People are rarely there and its allways charged.

If you aquire a 36v battery and many cells are dead you can make it a 12v battery, . Whatever suits you needs. Right now i'm looking @ one that looks kinda funky that survived a warehouse fire intact'ish.  Look around, these are out there.

Good luck, sounds like a fun project, free power makes a few miners mining solo more interesting.

Sounds like you have some really interesting stuff going on. I'm very excited to see your final results. Saving on power could be a life saver for a lot of people.

Interesting news boys. Ran a mock test on my solar panel i made.

These cells are claimed to run around 0.6v per cell at roughly 6 amps totaling around 3.6 watts each.

I made a 4x9 array totaling 36 cells all wired in series to match a 18v minimum inverter.

Some measurements i took today...

@9AM with zero cloud cover i recorded : 22v @ 6.5 amps!!!! That's 143 watts total and 3.97~ watts per cell!

@11:30AM with zero cloud cover i recorded : 22.2v @ 7.77 amps!!! That's 172.5 watts total and 4.79 watts per cell!!!

That's insanely good!


I should mention the other amazing side of these cells is; they are rough 165 micron thick or .006" thick, Semi-flexible to around 30 degrees bend, are relatively space saving in area compared to conventional pv cells, and weigh virtually nothing.


This panel is only 45" x 20"

Once i finish my solar tracker i should be able to get some real-time data that maximizes its output.

the nvidia will work (in hand is the key)  they will give decent watt to mh ratio

yeah the amd are still a little better.  but in hand is the key since in hand = 0$ cost and profit  as soon as you start to mine.

the rush is on but we all know 2 to 4 months and this will level out. 

Yeah i covered that here;


https://bitcointalksearch.org/topic/m.14461340

 
More or less i would run something like this

http://www.ebay.com/itm/LTC3780-DC-5V-32V-to-1V-30V-10A-Automatic-Step-Up-Down-Regulator-Charging-Module-/322049388287?_trksid=p2141725.m3641.l6368

but a bit more powerful obviously. Dirt cheap and efficient.


I want to be clear here too, i don't want to just run on low voltages, i want to undervolt/clock the unit itself to increase efficiency and decrease power usage. If the S7 has dc/dc converters, it sounds like that would bypass any undervolt i would do? Or does it merely run DC/DC converters to run the control board, while allowing the chips to operate at lower voltages.



philipma1957


Do altcoins still rock mainly ATI/AMD cards? I don't have any of those but i have a few top of the line model nVidia's. GTX 980, 780, a couple more.

this is why I mentioned the  gpu rig for eth coin.

right now I spend 311 watts to earn  about .44 eth coins a day that is about  0.01 btc

so 311 watts =     0.0101 btc    eth rig
590 watts     =     0.0034 btc    stock  s-5
1350 watts   =     0.0141 btc     stock s-7

so my point is  the eth rig  loses   power   solar panel to inverter say 97% eff back to atx 90%
that is 87.3 %

same battery cost


but 311 watts for 0.0101 btc
vs 1350 watts for 0.0141 btc

or 1244 watts for 0.0404 btc  eth rig rates
vs 1350 watts for 0.0141 btc  s-7 rig rates.

just saying a mix of gear here may be better then all s-7

This is not true for Lead Acid batteries under normal conditions. Their charge efficiency is about 90-95%. It decreases under special conditions like temperature or charging current (very low charging currents have bad efficiency, but it is not our case).

S5 and 165 chips batches of S7 are not happy of voltage variance caused by dis/charging. Fully charged battery at cycle charge is 14.6V, but fully discharged battery has 10.5V. S5 likes 12V at full speed, if you put 14.6V you may overheat it, side effect is much worser efficiency, growing to S3 level. Opposite, if you keep S5 running at low voltage, it will stop hashing or you need to slow it down. My S5 can work from 9V, but at slowest seed of 100MHz.

High efficiency DC/DC converter will stabilize your voltage, but it is more conversion loss and more expenses.

135 chip version of S7 has DC/DC convertor, so it can be used at voltages 11-16V. I did undervolted version of S7 which can be used from 10V.

 In my case, it's that the hundreds of millions Alliant is pouring into it's major wind farm projects have been a driving force behind the almost 50% rate increases (through new added "fees") they've pushed through here in the last 7 years or so. They USED to be a hair below US average power cost, now they're one of the highest-rate (all factors considered) utilities that is NOT on one of the coasts.


 I'm also less than a mile from the Clipper Windpower factory - which company is an interesting story in how to NOT take federal research funds and use them. VERY poorly done design, so even they were the first to market with a 2.5 MW windmill design, the thing broke so much that warrenttee repair costs has almost killed the company outright.

This can be true but that is typical with quick charges on duty cycle batteries. A marine battery is not a deep cycle battery. Those are somewhat hybrids. On a pure deep cycle where you only want to discharge maybe 30% of the battery max, you get a much slower charge rate and the efficiency is much higher.

Yeah its horrible. WE energies is among some of the worst. I read they recently shut down a Michigan plant they had, and then due to loss of profits from running that plant, decided to push that cost onto Wisconsin residents (where i live) to coop profit loss due to the Michigan site being shut down. That could be the reason for the extra hatred towards green energy programs. This is a company that is down right bitter about the fact that people creating clean energy directly affect their revenue stream. They aren't losing money by comparison.... they are making less profit. It costs them pennies on the dollar to move green (as customers are absorbing this cost all on their own) and they still won't do it.

There are two costs on my bill from them currently.

1) Service charge of $9~ a month. This covers the cost of the connection, monitoring, up-keeping and maintaining the lines, and repair.

2) The cost of the kWh you consume. For me... 0.135$ per kWh.



Now for people with grid tie solar....

WE said people who run their own solar don't contribute to the standard maintenance cost of people who pay for actual power usage. Even though they themselves state all those costs are wrapped up in the initial service charge standard customers pay.

So in addition to the standard $9 i would pay closer to $18. That's without using or sending any power at all. Now, they also put a tariff on power generation capacity. For every 1kWh capacity you generate they charge an additional $3.80 per KWh capacity from what i understand. They more or less make it completely senseless to go solar with them and i have the most efficient solar cells on the market.


This article covers the topic quite well


http://www.dailykos.com/story/2014/9/17/1328688/-Wisconsin-Electric-s-War-on-Solar-A-Case-Study


What's also funny about the claim that they lose large amounts of revenue is this:

https://www.we-energies.com/residential/acctoptions/eft.htm

Their own site... where if you want to be a customer that uses specifically or exclusively green energy that they themselves create... you must pay a higher rate

"Pay additional charge per kilowatt-hour. Because renewable energy costs more to produce than energy generated from traditional sources, the price is slightly higher."

Now that is somewhat as expected in my book (green energy costs more), but then don't make the claim that you lose money on green energy...

Don't forget to consider the charge efficiency of your batteries, which can be as low as 2/3 - which is to say, 900Ah in to get 600Ah out.

That is sad.  As solar works well with the gird tie in. NJ and MA  are good but many states are bad.

I'll certainly look into this. I wont rule anything out. The only downside to running an inverter is the efficiency loss. Same thing back into a psu.

I could lose 5% stepping up to ac through the inverter, and another 5% to drop it back down to DC through the power supply. The charge controller itself is probably a good 10% alone. it adds up quick. Any regulators etc all seed power on a low scale. Running right out the gate with 12v of the panel, battery and a regulator, I'm looking at minimal loss through the regulator. Still probably a good 5%. I also need the 3x factor in terms of solar power. If I'm using a 500 watt draw 24/7 assuming I get full power 8 hours of the day I'm going to want a 1500 watt solar array minimum. This will not only power the miner during the day, but provide an additiona 1000watt charging capacity to recover the 16 hour draw.  



The best idea I have at this point is to use the power company as my battery. If I simply grid tie these units with an inverter and make up the power during the day, the solar should still run my meter backwards for what I consumed at night no?

I wanted to stay off grid with my mining but perhaps the battery's just are not pheasible right now. I know my state put the kabosh on solar efforts. The power company basically said fuck off if you made your own power. Pretty shitty world we live in.

 Car battery AH are figured at 20 hours also, unless that got changed in the last 5 years or so.
 Reserve minutes on a car battery (and most car-sized deep-cycle batteries) is figured at a 25 amp draw.

 Deep cycle batteries are expected to drain down close to zero and designed for that, normal car "starting" batteries are not.

 You need more than just a buss bar to run batteries in parallel - they very enough in manufacturing that you can AND WILL drain one down while another is still nearly full charged. Current equalization is THE WAY - one way or another - for paralleling batteries.

 It would be more viable to run one hashboard off one battery, the other + controller off a different battery, which would keep the draw closer to 25 amps that most car and deep-cycle batteries are designed to easily handle. Then use a proper diode isolation circuit to keep one battery from feeding off the other while feeding them all from your solar setup, something similar to what some high-end "boom car" competitors or some RVs use to have 2 batteries, one to start the vehicle (and possibly run a FEW low-power accessories) the other running the (high-power) accessories.

 Low end "cheap" invertors generally put out square waves, NOT suitable for running most ATX power supplies on.
 There are exceptions, many of them designed specifically for use with higher-power Solar and Wind power installations.
 *ALL* "Grid-tie" inverters are designed to put out sine waves (or very close), those would be fine to power ATX power supply(ies) with.

I am a bit lost as to the why of this.  but hear me out.

you want

solar >>> battery >> Regulator >> s-5


why not Solar >> battery >> invertor >> atx psu >> gpu rig mining eth coins

yeah I know we are talking alt coins sin sin.

but if you are going large scale right now at this moment  the eth coin idea pulls 195 watts vs   590 watts

 easier to do  .

I suggest your first tests are with an eth coin rig like I just said.  read my thread in alt coins on an eth rig


https://bitcointalksearch.org/topic/my-1-gpu-now-11gpus-5rigs-for-ethereum-45-days-of-mining-photos-thoughts-1429151

my idea has more conversion and loses in the conversion  but is ⅓ the power draw and actually earns more then an s-5 does in a day.


if you are going big scale the ability to do either or could help you earn profit over your startup cost .

Found it!~


https://bitcointalksearch.org/topic/m.12451252



RichBC


"Best efficiency with reliable starting and operation was at 150MHz,  9.4V supply (Core Voltage 0.63V) consuming 166W at the Wall giving 495GH with an efficiency of 0.335J/GH."

I read something like that somewhere and it was said to be around 350 watts down around there i thought. Ill see if i can find it.

Yeah, a good efficient buck/boost would solve a lot of problems.

Wasn't one of the later revisions of the S5 capable of running down around 10V?

And you cannot drop below around 11.7 or the miner will be very unhappy... More than anything the time it takes to drop to that is what matters and that small of a voltage swing puts you back to normal (but big) batteries.

Deep cycle batteries are designed to routinely withstand going from full charge 12+v nominal to below 8v or less @ sustained high load which would rapidly kill a normal battery. Remember, they are for trolling motors, electric pallet jacks, and RV power etc. Now if they feed an inverter or a switching buck/boost regulator to hold @ 12V out, whole different story.

edit: NM. Just read reply ^^

I do plan on running a solar charge setup inline with a few batteries. I was thinking running three to four 600 amp deep cycles on a solid buss bar in parallel. That should be enough capacity to maintain a small draw. Sidehack, I intended to use a step up/down voltage controller to regulate the output to the miner. Honestly i wanted to underclock/volt the unit to get larger efficiency so i could run two or three units. I'm shooting for a specific power maintenance level around 1kW. I could even fit a undervolted s7 if that's been done. I made my own panel out of some 21% efficiency cells. I plan to make a solar sun tracker for this panel and about 5 more.

I do realize however this is a very large expense for running two s5's     But this is more or less a proof of concept for a larger plan.

That 600 amps is over 20 hours, not 24 hours. AH rating for deep cycle batteries = how many amps over 20 hours, to drain the battery to zero (which is bad for the battery) and car batteries are figured on 10 hours.

So, to get the battery to last for years (and not months) you only want to pull around 50%, so figure 300 amps.

Plus, as sidehack was talking about, the fully charged battery is going to be around 12.7V.

What is your goal of the batteries?  Are you going to be running solar setup on them?

I'm missing what main reason of needing the batteries is.   But few cent electricity beat's almost anything is why you will see "naysayers shunning the use of batteries".  Just normally a good initial investment keeps most from battery solutions most common being solar.

Well I had my S1s and have my S3 units on 450 watt UPS units that are supposed to hold for about 5-10 minutes under full load.  Batteries are brand new and the S3 will drain in in about 15 seconds when the power goes.  Now my S5 units are on 3kva UPS units with servers and other stuff and they don't drain those nearly as fast.  So I guess it all depends on the batters and equipment but I would not expect anywhere near normal operating times out of the batteries.

Capacity is also a function of draw. A 600Ah battery drawing 600A probably will not last for one hour. It really depends on what the "expected load" of the battery is, what the draw was at capacity measurement.

You also need to consider regulation. Is the nominal voltage of this battery 12V? Because if it's more like 14V you're going to sorely overvolt your S5 and likely burn it up.

Hey guys, I'm wondering if anyone knows how a 12v Deep Cycle battery would do running a miner like an S5? Assume a 590 Watt draw just as a baseline. Is it as simple as pulling watt hours from the battery? IE an S5 running @ 12v using 590 watts is pulling roughly 50 AMP. Is it as simple to say that if the battery is 600 AMP it would provide 600/50= 12 hours? Deep cycle's don't really like to drop below 50% for long life. So 6 hours? I'm more concerned with my concept being incorrect than actual data numbers here.

I've seen plenty of naysayers shunning the use of batteries as they typically have voltage drop as load is applied and battery resistance rises. That is how automotive batteries work, however deep cycle are specifically designed to maintain voltage across a usage curve with a steep drop-off and resistance increase generally happening around the 80% drain mark.

Happy hunting