02 October 2024 – by ertech
Author’s note: Thank you to Arman the Parman for reviewing and improving this paper. Any remaining shortcomings are mine.
Technical level: Beginner to intermediate.
Disclaimer: Nothing in this paper should be construed as financial or technical advice, you must do your own research. The veracity of instructions is subject to changing conditions and limitations of the author.
==========
==========
1. INTRODUCTION
This is a record of setting up a home bitcoin mining operation without an extra electrical service, extensive garage space or exterior shed. The experiment is this: a. can a non-technical person set up a modest home mining operation to stack sats (acquire bitcoin incrementally) with a relatively low expenditure; and b. can it be justified?
Mining is the computing operation, done as a competition the world over, that brings new issuance of bitcoin into existence on a steady schedule. The winning miner receives not only the network fees associated with the transactions within the block, but also the block reward (newly created bitcoin), which the Bitcoin protocol currently sets at 3.125 bitcoin. For a detailed review of how mining works, see this explanation by Arman The Parman.
Mining with a CPU or GPU was done in the early years, but is now superseded by specialised computers called ASICs (application-specific integrated circuits). Because the odds of “finding the block” are small, miners typically pool their efforts and then share the rewards, according to the mining pool’s policy (there are many types).
Choices for mining for the individual now include renting a remote hosting facility and maintenance services, enabling a more powerful operation than can usually be achieved at home. The risk is that in trusting such a service, you may lose control over your equipment. Since both remote options and elaborate shed facilities are of a different scale and character than our proposed home operation, they need separate analysis and so are out of scope for this paper.
2. RATIONALE
Discussion of home mining often jumps to consideration of its short-term profitability in terms of fiat (government-issued) money, accounting for equipment and power costs. From that perspective, home mining, in the sense that I am using the term, will likely have to be run on a break-even basis or at a loss. Profitability is also now questioned due to the low block reward (currently 3.125 bitcoin).
There is little doubt that small-scale home mining, even if profitable, will not substitute for establishing a substantial position relatively quickly via bulk or DCA (dollar cost averaging) purchases. Therefore, forecasted profits should always be compared with the opportunity cost of early lump sum purchases.
Is there a financial rationale for the home miner to begin operations? The first positive consideration is that, despite the lower block reward, a general rise in transaction demand is expected to extend into the future, resulting in ever-higher mining fees. The Bitcoin network’s aggregate computational power, or hashrate, while volatile, is also expected to increase indefinitely. A rather complex market dynamic plays out, roughly as follows:
Price increases in Bitcoin raise the profitability of mining for all miners of the world. As new miners are attracted into the market, mining becomes more competitive. Less efficient miners are gradually forced to unplug their rigs, leaving mining to those with more efficient equipment and lower power costs. The Bitcoin protocol’s difficulty adjustment, (whereby the computational problem to win the block is made harder in up markets and easier in the downturns) helps to regulate these market actions, so that the pace of issuance of new bitcoin is held quite steady. Moreover, it virtually guarantees that mining will always continue.
In this constant flux of Bitcoin price, transaction demand, level of difficulty, and costs, miners must take care to withstand the volatile hash price (the rate at which mining rigs earn revenue). In the world of industrial-scale mining, cash earnings must keep up with costs, which usually must be paid in fiat.
Perhaps the home miner is less affected by this brutal market discipline. The home miner’s operational cost (the electricity bill) is marginal in the context of a household budget. Home miners are generally not dependent upon a cash flow generated by bitcoin earnings to cover costs, as long as they remain small. The need for the subsidy is mitigated by two factors: 1) the dual purpose of home rigs as miners and space heaters, reducing the electricity bill; and 2) the possible effective discount in bitcoin purchase price that mining would obtain. These, then, are the cautions as well as a few positive indications forming the rationale for home mining. Let us see how the experiment was executed and its results.
3. HARDWARE; SOFTWARE; SERVICES
Devices/OS:
Notebook (desktop) computer running Linux (or other OS)
Android phone (or iPhone);
Canaan Avalon Nano 3 – marketed as a space heater with mining function. I am using quantity = 3.
Software:
Parmanode system on a notebook (or desktop) computer including:
Bitcoin Core (hosting own node)
LND (hosting own Lightning Node as Layer 2 solution on top of Bitcoin core)
Zeus Wallet integration – a “hot” (software) wallet as opposed to a cold wallet device
Alternative to Zeus (to be explained): Wallet of Satoshi software wallet (app on phone)
Canaan Avalon Remote app on phone – used to configure and monitor the Nano 3 mining device(s)
Braiins mining pool monitoring app on phone (see next)
Services:
Braiins mining pool account online: set up on desktop; link device (phone)
4. CONFIGURATION and TRADE-OFFS
The choice of hw/sw and services (as outlined above) and their configuration require a series of trade-offs.
MINING POLICY and PAYMENT
Let`s assume that we want regular, predictable payments. In that case, it would be preferable to receive fairly large bitcoin amounts from the mining pool, because (with Braiins mining pool, for example) there is no fee charged for payment above 10k sats. This would also permit stacking UTXOs (analogous to paper bills in a wallet) of a practical size.
However, it would take many weeks of mining to build up sufficiently large payout amounts, and this increases custodial risk. By contrast, acquiring in one’s hardware wallet a stack of small-value UTXOs via frequent payments from the pool does reduce custodial risk, but presents the problem of high fees to eventually consolidate or spend all of these tiny values (referred to as “dust”).
The question becomes: can we receive frequent, sizable payments cheaply, and maintain self-custody?
LIGHTNING NETWORK
My solution involves the Lightning network, a fast and low-fee service built on Layer 2, i.e., “on top of” the Bitcoin blockchain. (its transactions are eventually resolved, behind the scenes, on the Bitcoin base layer.) A Lightning wallet allows you to accumulate a balance, just as in a conventional bank account, rather than a list of individual entries (UTXOs) as is done in Bitcoin itself. Eventually, a large balance in a Lightning wallet can be sent to one’s hardware wallet as a single bitcoin UTXO for self-custodial cold storage.
MINING POOL
I have not conducted a review of mining pools, but for now, I accept the default association of the Nano 3 with the Braiins mining pool. Its policy (Full Pay Per Share – FPPS) gives steady payments, commensurate with the hash (computational) power contributed, even if the block for a given period is not found by the pool. Additional reasons for choosing Braiins are as follows:
(a) since it is the Avalon Nano 3 default, we know this small device will actually be recognized by the pool;
(b) it offers payouts on the Lightning Network, suitable for small amounts;
(c) the payment threshold amount (and thus frequency) is flexible, set by the user;
(d) there is no fee to withdraw funds via Lightning.
The only cost is the pool fee of 2.5% share of one’s mining proceeds.
ZEUS WALLET
The (seeming) ideal setup would be to receive frequent Lightning payments from Braiins to the Zeus Wallet. It is a software (“hot”) wallet. All software wallets are considered less secure than hardware wallets because they touch the internet and so could be hacked. Braiins requires a static Lightning address to which it sends your funds, and sure enough, Zeus gives you such an address.
I would like to thank Parman for taking fast action to set up the Zeus integration in Parmanode when we first discussed this back in August 2024. This enables taking self-custody on Lightning and reduces custodial risk. But I must apologize: I’m not actually using it at this point. My reasons are as follows:
Zeus does not permit the automatic receipt of Lightning payments to a static address. Rather, as I discovered: “Users need to log in to their ZEUS wallets within 24 hours to claim the payments”. The device (cell phone) must be on. The payment threshold from Braiins I am using is 5000 sats, which occurs roughly once a week. This makes the use of Zeus impractical unless one were willing to set a notification, try to hit the right day, and re-create the invoice for every payment. Finally, Zeus Wallet charges a flat fee of 250 sats (5% of the proposed 5000 sat payout) to receive over the Lightning network.
WALLET OF SATOSHI
My current solution is to use Wallet of Satoshi (available in my country, although not in the US). It is a custodial wallet. Yes, I am accepting custodial risk here in exchange for convenience: there is no need to manage payment channels (to get inbound liquidity) as with most Lightning wallets. Further, there are no fees for receiving Lightning payments. Sending bitcoin to cold storage will cost 1% plus a mining fee.
I realize I could just as easily send bitcoin payments straight from Braiins to a hardware wallet cold storage. That seems the most rational, as it can be done without fees. I might switch to that. But as it is, I am at least getting the experience of using a Lightning wallet, which I may use for other payments.
I confess that I went to my current setup when I felt a bit overwhelmed. It certainly can be improved, so I invite Parman’s comments here.
My goal was to automate the receipt of payments from the mining pool, as cheaply as possible. Whatever your solution may be, pay attention to two key points if you use the Lightning network: a) ensure that you get a static receive address (it has the same format as an email address) rather than having to manually create an invoice for each payment; b) ensure that you understand how inbound liquidity is achieved.
CANAAN AVALON NANO 3
Computers dedicated to professional mining, such as the Antminer S19 or S21, are not suitable for home use. By contrast, the Avalon Nano 3 gives a solution for the home miner in respect of five key factors shown in the following table:
| Bitmain S19 Pro | Avalon Nano 3 | |
| 1. Cost | USD 1300 | USD 129 |
| 2. Power consumption | 3250W | 140 W rated; 125 W output |
| 3. Voltage required | 220V – requires special installation for home use in US (or similar) jurisdiction | 120V |
| 4. Hash rate | 110 Th/s | 4 Th/s |
| 5. Heat produced | major problem to dissipate | modest; not out of place in a cold office or bedroom |
| 6. Noise level | 75dB = intolerable for home use unless isolated in a shed, basement or garage | unobtrusive (but see comments on my atypical setup) |
S19 data from ASIC Miner Value site
The Avalon Nano 3 is marketed as a space heater with a mining function – fair enough. One reviewer characterized it as “the Honda Civic of miners”, as it is clearly suited to home use, and is a better performer in terms of the hashing power/purchase price ratio than other 110V devices I have seen so far. The longevity of the Avalon Nano 3, if it is a typical miner, should be 3-4 years with no maintenance. The cooling devices (rack; heat sinks; fans) added to the set-up should extend its life.
The Nano 3 throws off enough heat to replace (in a small room) or supplement (in a large room) a baseboard heater. The Nano 3 is rated at 140 W, drawing only about 1.16 amps. Note that the PSU (power supply) heats up considerably and this must somehow be managed. Do not use anything other than the Canaan-specified PSU for this device, as it will likely overheat and fail.
MY SETUP
Since the Nano 3 generates only 4 Th/s in comparison to a pro machine I decided to order 3 of them for my setup. This is a subjective decision; it seems to optimise the trade-off of rewards vs power costs. I am able to use the heat function in colder months by doing my install in the garage, below the living room, so that the baseboard thermostat upstairs will trip on less frequently. At the same time, the noise the rig produces is isolated.
I also decided to supplement the devices with the following, incurring, of course, more initial cost:
- a power bar with surge protection
- an aluminium server rack of my own design/build
- cooling fans installed at the output side of the devices, with covers removed, to draw more airflow
- heat sinks installed on each of the PSUs, supplemented with scrap pieces of aluminium flat bar
These expenses are justified, because the supplementary fans should extend the life of the devices, which now can run 24/7 unsupervised and so avoid heating-cooling cycling. The PSUs are physically isolated at the top of the rack and considerably reduced in temp (I did not measure it). The Nano 3 plastic enclosures themselves are positively cool to the touch. Finally, as noted, each device is drawing only 1.16 amps and so the whole setup runs on one 15 amp circuit.
Canaan Avalon Nano 3 Mining Rig
(quantity = 3 units) custom setup.
PROCEDURE
1. Set up your preferred Lightning wallet (if you are using one; presumably with private email).
2. Establish an account at Braiins pool, https://braiins.com/pool or the desired mining pool; designate the receive address for payments.
3. Set up Nano 3. Instructions are included. No ethernet connection – you must connect to a wifi router.
4. Download the “Avalon Remote” app and let the app discover the device(s). Go to “Mining management”. Ignore the initial warning. It takes several iterations to get the settings right. Use the default “pool address”. The password is not needed. It seems to designate the name(s) of the device(s) – called “workers” – itself, and overwrites the names you may try to invent. Further, it takes time for the hash rate for each machine to ramp up. Persist until your full hash rate is showing (see next two points).
5. Under the app’s “Heater control” I have set the devices to run on “high” in order to get the maximum hash rate – some have commented that this draws too much power or produces too much heat. But don’t forget I introduced cooling measures. You can see the power output (about 125 W) on the small screen on top of the device. Set the “Light control” as desired.
6. Access your Braiins account on your desktop. Then go to your phone and download the (Braiins) mining pool app; pair your mobile phone with your account (User icon top right> Devices> Add new device).
As the Avalon Nano 3 miners are discovered by the app, they show up on your account. Note that only one miner will appear to be connected. But by looking at the hash rate (= approx 12 Th/s) you will see that the mining pool is picking up the total hash from the rig. Power interruptions are not a problem; the Nano 3 devices quickly reboot, find the wifi and the pool, and resume working all on their own.
See screenshots:
Braiins mining pool account desktop UI – Dashboard showing the current hash rate.
Braiins mining pool app UI – Rewards showing several weeks’ earnings and estimated daily reward.
5. RESULTS
CAPITAL COSTS/PARTS LIST in USD (approximate: you must verify, check shipping, etc.)
| ITEM | QTY | PRICE | REMARKS |
| Canaan Avalon Nano 3 devices (incl PSU; power cord) | 3 | $687 | NB: this is for quantity 3, not a single device |
| Fans – Iceberg Thermal | 3 | $20 | |
| Wiring harness – PC Fan adaptor | 1 | $15 | |
| Heat Sink – Aluminium | 3 | $33 | size 100 x 69 x 36mm – fits the PSU |
Server rack:
To make a design similar to mine, buy materials from a local industrial metals supplier (not hardware store): 1” aluminium flat bar, 1” aluminium angle, SHCS fasteners. $120.
======
$ 875 TOTAL CAPITAL OUTLAY
======
DESIGN NOTES
See photo p.5. Rack is 24″ high x 12″ wide x 12″ deep. Four shelves @ 6″ height spacing. Use double screw joints for stability (or use diagonal cross-braces). Aluminium flat bar shelves act as heat sinks. The top shelf for PSUs is offset towards the back. The miners/fans are towards the front. This makes wiring easier. You will also need tie wraps and duct/thermal tape. The extra set of fans is not strictly necessary.
OPERATING COSTS and PROFITABILITY
The above spreadsheet shows the miner, as space heater, offsetting costs to a modest degree ($11.56/mo). The home miner must be willing to absorb a $10.71/mo cost, which results in a discount in the effective price of Bitcoin. With that benefit, the operation overcomes the extra electrical cost – but these calculations ignore the capital cost outlay.
SIGNIFICANCE OF BITCOIN ACCUMULATION OVER TIME
Assume 330 days (approx 90%) uptime @ 850 sats/day.
Current bitcoin price is USD 65K
Projection of bitcoin total and fiat price. Costs, and changes in bitcoin and hash price ignored:
** This is Michael Saylor’s prediction for the fiat price of Bitcoin by the year 2030.
6. CONCLUSIONS
The rationale was that anticipated conditions in the Bitcoin market might favour a modest home mining operation, although it would likely be a supplement to, but not a replacement for, bulk purchases.
This expectation was fulfilled, in a limited sense. The home mining rig does overcome, at least marginally, the extra operating (electrical) costs when using the miner as a space heater. In that case, bitcoin is mined at a discount compared to the current spot price. However, the whole endeavour can only be financially justified if the capital (rig) costs are ignored.
To emulate this home mining solution, one must be willing to regard as sunk costs the initial capital outlay ($875) and focus on ideological reasons, or other ancillary benefits, such as stacking in a non-KYC fashion.
There are increasing concerns about the concentration of Bitcoin mining capacity. Until individual Bitcoiners can find efficiencies and realize value in home mining, it will be difficult to achieve more decentralization, consistent with the Bitcoin ethos.
(See the addendum – comparison chart of home mining devices, next page.)
ADDENDUM – COMPARISON CHART
This chart shows Bitcoin miners in the Terahash range that run on 110-120V (suitable for home use in US). Users elsewhere with 220V service (or those willing to build it) have more market choice, but must still manage heat and noise.
These are the devices that I found time and again online, but it is not an exhaustive list:
| NAME | COST USD | HASHRATE | POWER CONSUMPTION | HASHRATE per USD | REMARKS |
| Canaan Avalon Nano 3 | 229 (incl PSU) | 4 Th/s | 140W | .017 (best in its price range) | quiet operation;modest heat |
| Braiins Mini | 299 (incl PSU) | 1 Th/s | 40W | .003 | quiet operation |
| FutureBit Apollo | 799(PSU integrated) | 10 Th/s | 400W | .012 | quiet operation; pairs w/ node |
| BitAxe | 179 (incl PSU) | 1.3 Th/s | 20W | .007 | tiny device uses S19 chip; found the block Jul24 |
| Urlacher S19 Pro 115T | 1,749 (incl PSU) | 56 Th/s | 1200W | .032 (best overall) | modified S19; space heater; see video demo- quiet operation |
Please do your own research. I tried to find the best source/price but can’t guarantee it. Scan the market also for any new developments in home mining. (Much of the online discussion has to do with shipcoins.)