Years ago with my own product I decided to seal the battery inside to keep the cost down, and that same decision today would shut it out of one of the world's biggest markets.
Every battery in this countdown works perfectly on your bench, and every single one of them has ruined somebody's product.
One of them will get your shipment refused by your freight company before it ever leaves the country.
Another one puts your product under a federal safety law most people building products have never heard of.
And unlike most design mistakes, a bad battery choice doesn't just mean a product that won't work.
It can physically hurt your customer, and your company is the one liable when it does.
So in this email I'm counting down seven batteries you should never use in a product you plan to sell, and I'll tell you exactly what to use instead for each one.
Battery #7 - 9V Alkaline
The 9V alkaline is the battery almost everybody reaches for on their first prototype, and it's the first one I'd get rid of.
It looks efficient, but a 9V alkaline only holds around 500 to 600 mAh, and once you account for the voltage difference you're paying three to four times more per unit of energy than with a couple of AA cells.
The voltage is a problem too, because your electronics almost certainly run on 3.3V or 5V.
If you drop 9V down to 3.3V with a simple linear regulator, you throw away about two thirds of your energy as heat.
A switching regulator fixes the efficiency, but now you've added cost, parts, and noise to solve a problem you created by picking the wrong battery.
The snap connector is a mechanical weak point, since those thin wires flex every time your customer changes the battery, and eventually one breaks off inside the enclosure.
If you need a replaceable battery, two AA cells in series give you around 3V, which many microcontrollers and sensors will run on directly with no regulator at all.
And when you do need a regulator, dropping 3V to 1.8V with a cheap linear part, or boosting it up to 3.3V, is a much smaller job than stepping down from 9V.
If you'd rather go rechargeable, a single-cell lithium polymer battery with a charger IC is a better fit for almost any modern product.
Battery #6 - Alkalines in High-Drain Products
Alkaline AA and AAA cells belong in wall clocks and TV remotes, not in any product that pulls real current.
The AA format itself is fine, it's the alkaline chemistry that gives out when your product leans on it.
The problem is internal resistance, which eats into the voltage the second your product starts drawing power.
A fresh alkaline AA sits at 1.5V doing nothing, then sags down toward 1.1V under a heavy load, and that internal resistance keeps climbing as the cell drains.
So your product browns out and resets while there's still plenty of capacity sitting inside the battery, and your customer is convinced they got a bad pack of cells.
Cold weather makes it worse, since alkaline performance drops off badly near freezing, which is why I steer anyone building outdoor products away from alkalines.
Then there's leakage, and that's the one that fills your inbox with warranty claims.
Alkaline cells leak potassium hydroxide when they sit in a device too long, and that stuff corrodes your contacts and destroys the product from the inside.
Your customer doesn't blame the battery brand for that, they blame you.
For high-drain products, low self-discharge NiMH cells are a solid swap, just remember they're 1.2V per cell instead of 1.5V, so design your voltage range around that.
And for the toughest conditions, lithium iron disulfide cells like the L91 class handle heavy current and extreme cold far better than any alkaline will.
Battery #5 - Coin Cells for Anything Beyond Tiny Loads
Coin cells look like free energy when you're laying out a small product, because a CR2032 is tiny, cheap, gives you 3V, and the datasheet promises around 225 mAh.
That number only holds up if you sip current out of it very slowly, a few hundred uA or less.
A coin cell has internal resistance in the tens of ohms when it's fresh, and that climbs into the hundreds of ohms as the cell drains.
So the moment your Bluetooth radio fires off a transmit burst and pulls 10 or 15 mA, the voltage collapses and your microcontroller resets.
You end up with maybe a third of the rated capacity, and the product dies early in the field for reasons that never showed up on your bench.
Adding a bulk capacitor across the cell helps ride out those current spikes, but that's a patch on a battery that was never sized for the job.
There's a legal side to coin cells now too.
Reese's Law passed in 2022 here in the US, and it means any consumer product using a button or coin cell has to keep that battery in a compartment children can't get into, plus carry specific warning labels on the product, packaging, and manual.
That's an enclosure design requirement and a certification item, not just a battery choice.
If your product needs more than a trickle of current, step up to a AA-sized lithium primary cell or a small rechargeable.
Battery #4 - Sealed-In Batteries
Using sealed-in batteries is one of the big mistakes I made with my own product.
And today, that same mistake would lock the product out of the EU, because of a new battery regulation I'll get to in a minute.
My product was a small LED light running on coin cells, and I wanted those cells to be user replaceable.
But every version of that idea made the product worse.
A battery door meant a separate piece of plastic, and that meant another expensive injection mold.
It also meant a bigger enclosure, more parts, and a contact design that had to survive people jamming cells in backwards, on a product that sold for a few dollars.
Replacement coin cells cost almost as much as a whole new unit, so sealing the battery in looked cheaper for me and sensible for my customer.
But most customers still wanted the ability to change the battery, and the complaints kept coming.
It got to the point where I recorded a video showing them how to fully disassemble the unit to replace the battery, even though it was never intended to be user serviceable.
Article 11 of the EU Batteries Regulation says portable batteries in a product have to be readily removable and replaceable by the end user, and it applies from February 18th, 2027.
The replacement has to work with commercially available tools, so adhesives that need heat or solvents are out, and you have to stock replacement batteries for five years after the last unit of that model ships.
This hits rechargeable products hardest, because a portable battery under that rule is anything sealed weighing 5 kg or less, which covers a glued-in lithium polymer pouch cell the same as a coin cell.
There are exemptions, and the Commission keeps adding to that list, with wearables like smartwatches and fitness trackers, plus electric toys, in the most recent round.
Article 11 might sound like it collides with the coin cell rule, since one law wants the battery locked away from children and the other wants your customer changing it.
A battery door held shut with a screw satisfies both, because a screwdriver counts as a commercially available tool and reusable fasteners are allowed.
The danger is that this isn't a component you can swap later, it's an enclosure decision, a tooling decision, and a contact and layout decision, all made at the beginning and brutally expensive to reverse once you've cut steel.
So decide this at the concept stage, before you have a single mechanical drawing.
And if you think you fall under one of those new exemptions, confirm that with a compliance expert before you commit to tooling, not after.
Battery #3 - Unprotected 18650
The 18650 is an excellent lithium ion cell, and that's exactly why so many people drop a bare one straight into their product.
An unprotected cell has no circuitry sitting between the chemistry and your customer, so nothing stops an over-discharge, an overcharge, or a dead short across the terminals.
Take a lithium ion cell below about 2.5V and you cause permanent internal damage that can create a short weeks later, long after the unit shipped.
A shorted 18650 can dump 10 A or more into whatever happens to be touching it.
That's how you end up with a product that catches fire in somebody's living room.
Your customers will find every one of those failure modes, since they'll install cells backwards, charge them with whatever adapter is in the drawer, and toss a spare cell loose in a bag with their keys.
There's a counterfeit problem on top of that, because the online marketplaces are full of rewrapped 18650s advertising 9,900 mAh when the best real cells top out around 3,500 mAh.
Buy through authorized distribution, and use either protected cells or a proper pack with a battery management system that handles charging, discharging, and cell balancing.
Build it on cells from the top-tier makers like Samsung, LG, Molicel, or EVE, and get the real datasheet before you design anything around them.
Battery #2 - Bare LiPo Pouch Cells
Bare lithium polymer pouch cells are the most dangerous habit that carries over from the hobby bench into a product you sell.
These are the flat silver packs with two wires hanging off them, the kind sold for RC planes and drones, and they ship with no protection circuit at all.
That's on purpose, since RC users want every last amp out of the cell and accept that risk themselves.
Your customer never agreed to any of that, and your name is the one on the box.
The pouch is soft aluminum laminate, so a screw boss pressing into it, a pinched corner during assembly, or a drop onto a hard floor can puncture the cell.
A punctured or overcharged lithium polymer cell doesn't just stop working, it swells, vents, and can go into thermal runaway.
What you want instead is a pouch cell with an integrated protection board already attached, from a supplier who'll hand you a datasheet and test reports without you chasing them.
On a protected cell, that board usually sits at the top of the pouch, under the semi-transparent yellow tape where the wires come out.
Pair that with a proper charge IC that does constant current and constant voltage charging with a hard 4.2V limit, and add temperature monitoring so charging stops when the cell gets too hot or too cold.
Then design the mechanical side around it, with a little gap for swelling as the cell ages, and nothing sharp or rigid pressing against the pouch.
Battery #1 - Uncertified Lithium Packs
Uncertified lithium packs take the number one spot because this is the one that stops your entire product from shipping.
These are the cheap packs sold online with no datasheet, no test reports, and a seller who can't tell you which cells are inside.
UN38.3 is the transport safety standard for lithium batteries, and it's a set of eight tests covering altitude, thermal cycling, vibration, shock, external short circuit, crush, overcharge, and forced discharge.
Without that test summary in your hands, your freight forwarder won't accept the shipment, and you can't legally move those batteries by air.
It blocks you at certification too, since IEC 62133 is the battery safety standard your product certification leans on, and you can't get there with a cell you can't trace back to an actual manufacturer.
The liability side is even worse, because if one of your products burns down a customer's house and you can't produce a single document about the battery inside it, you're the one holding all of it.
Buy your packs from established manufacturers who hand you the full documentation package, and ask for the UN38.3 test summary, the cell datasheet, and the IEC 62133 report before you send anyone money.
If a supplier gets vague the moment you ask for those documents, that's a big red flag.
Talk soon,
John
P.S. If you need help navigating battery selection and compliance for your product, then you can get help from me and other experts inside the Hardware Academy.