Off-Grid HandbookPower and water for cabins and homesteads
About

What We Don't Recommend, and Why

The short answer

Every product below was researched for a page on this site and then not linked, or linked and then withdrawn. Most affiliate sites never publish this list because there is no money in it. It exists here because a recommendation only means something if refusing to make one is also on the table, and because the reasons are useful in themselves.

Most buying guides are lists of things to buy. This is the other list.

Everything here was researched for a page on this site, and then not linked. In several cases the product was most of the way to being recommended before a published number ruled it out. One of them was recommended, wrongly, and then corrected.

The reasoning is the useful part. A verdict tells you about one product; the reason tells you what to check on the next one.

MRBF terminal fuses, on a 48 volt bank

Was going to be: the recommendation for the “fuse as close to the battery terminal as practical” rule, which is a real rule and MRBF fuses are the obvious way to satisfy it.

Why not: interrupt capacity is published per voltage, and it collapses.

10,000A at 14V DC
 5,000A at 32V DC
 2,000A at 58V DC

Class T, for comparison, is 20,000A at 125V DC. At the voltage that matters for a 48 volt system the MRBF has one tenth the interrupt capacity, and a 48 volt lithium bank can deliver far more than 2,000A into a dead short.

Separately, the fuse and its block are both rated 58V DC maximum, and a 16-cell LiFePO4 bank charges to 58.4V. The part is over its voltage rating at full charge before the interrupt question is reached at all.

Where it is fine: 12 volts, where 10,000A is real protection and terminal mounting is a genuine advantage.

The full argument is fuse interrupt rating, which exists because of this rejection.

Marine battery switches, on a 48 volt bank

Was going to be: the main battery disconnect. Search “battery disconnect” and these are what you find.

Why not: none of them is rated for 58.4V. Maximum voltage, from the manufacturer:

Switch Maximum voltage Continuous
m-Series Single Circuit and Selector 48V DC 300A
m-Series Dual Circuit and Dual Circuit Plus 32V DC 300A per circuit
e-Series 9003e 48V DC 350A
HD-Series 3000 32V DC 500A

Note the inversion. The Heavy Duty switch, the biggest and most expensive in the range, has the lowest voltage rating of the four. Reasoning “buy the beefiest one” picks the worst part on the list.

The 48V entries are the more dangerous ones, because the number matches what you call the bank while meaning the opposite thing. That argument generalizes well past disconnects: why a 48V part is not for a 48V bank.

Unbranded high-amp fuse holders

Was going to be: nothing. These surface high in search on price.

Why not: a $6.59 “250A” ANL holder and a $12.99 “300A 58V” MRBF, both from unidentified sellers, publish no interrupt rating anywhere. Not a low one. None at all.

Linking a part with no published interrupt rating on a site whose wiring pages argue that interrupt rating is the thing people get wrong would contradict the page it sat on. Not a close call.

The general rule: where a specification decides the purchase and the seller does not publish it, the absence is the answer.

Battle Born batteries

Why not: in June 2026 Dragonfly Energy sued a well-known reviewer over statements made in monetized content, citing his affiliate income from them.

Whatever the merits of that case, the only thing this site has is that nobody is paid to reach its numbers. A brand that has demonstrated willingness to sue a publisher over review content is a brand this site will not take money from, because the arrangement itself would compromise the position.

This is enforced rather than remembered: a test in the codebase asserts that Battle Born is not a linkable merchant, so the decision cannot be undone by someone forgetting it.

Not a rejection. Victron makes excellent equipment and this site has no criticism of it.

Why there are no links: Victron runs no affiliate program at all, by policy. They sell through distributors only. So the site cannot earn from Victron even where Victron is the right answer, and a page that recommends one is a page that earns nothing.

It is listed here because the absence of Victron links across a site about off-grid power would otherwise look like a judgment. It is not. It is a business model.

The one I got wrong

On 2026-08-31 the page on LiFePO4 versus AGM linked a battery marketed as “Low-Temp” under a paragraph about charging below freezing, and described it as having internal heating that removes the below-freezing charging limit.

That was wrong, and wrong in the worst available direction. The manufacturer’s own page states that the BMS cuts off charging below 32°F and resumes at 41°F. The word “heat” does not appear on that product page at all. It is protection against cold charging, not a way to do it.

A reader in an unheated cabin following that link would have bought a battery that refuses to charge in exactly the conditions the paragraph was about, and the claim sat directly above a paid link.

What changed: the link now points to the same manufacturer’s genuine self-heating model, which publishes a charge range down to -4°F. The page also carries a warning that “low temperature” in this market usually names a cutoff, not a heater.

The general lesson, which is why this is published rather than quietly fixed: a product name is marketing copy, and on cold-weather battery products it frequently names the opposite of what a reader wants. Read the published charge temperature range, which is a number, rather than the badge, which is a word.

What this list is for

Two things.

It is the test of everything else on the site. A page that only ever recommends is an advertisement with paragraphs. These pages are supposed to reach a conclusion honestly, and sometimes the honest conclusion is that the obvious product does not work.

The reasons outlive the products. Every entry above comes down to the same handful of habits: read the number and not the name, check the specification that decides the purchase instead of the one in the headline, and treat a missing specification as information. Those apply to products this site has never seen.

Where something here changes, because a manufacturer revises a rating or a situation changes, this page changes with it and says when.

Specifications last verified . Every specification quoted here was read at the manufacturer, from the datasheet, manual or specification tab, and never from a retail listing, which routinely omits the figure that decides the choice. That keeps each comparison small: I would rather name three products I can check than thirty I cannot. No price is shown beside anything I recommend, and prices are not tracked, so check the current price at the retailer. Nothing here was supplied free, no manufacturer has any say in what appears, and where I own a product its card says how many and exactly which variant. A guide describes components, not your system: whether one fits depends on your loads, your worst month of sun and how long you must run without any, which is what the calculators work out.

Common questions

Why publish a list of products you do not recommend?
Because a recommendation is only worth reading if not recommending something was a real possibility. Every product here was researched for a page, and several were most of the way to being linked before a published specification ruled them out. The reasoning is more useful than the verdict.
Are these products bad?
Mostly not. Almost everything here is a well-made product being used outside what its manufacturer rates it for, usually because a name or a nominal voltage made it look like a match. The marine battery switches and MRBF fuses are excellent at 12 volts. The problem is the crossing over.
Do you get paid to leave products off this list?
No. Nothing on this site was supplied free, no manufacturer has ever had input into what appears, and no arrangement exists with anyone to include or exclude a product. Where a page links a product, the commission is disclosed on that page.
Have you ever published a recommendation that was wrong?
Yes, once, and it is described below. A battery was linked under a paragraph about charging below freezing with a description of what it does that was the opposite of what its manufacturer publishes. It was corrected the same day it was found, and it is documented here rather than quietly edited out.