SEC1210 Raspberry Smartcard Hat Assembly & Design Discussion

This is a longer, more technical piece than I usually do, but I wanted to document how the Smartcard Hat for Raspberry Pi comes together from start to finish. It’s aimed at DIY hardware builders, privacy-focused Bitcoin users, and anyone who wants to understand or reproduce the design so they can connect a Satochip/SeedKeeper JavaCard to SeedSigner.

Watch the video.

The hat is small, but it has enough surface-mount parts that hand assembly takes some care. The good news is that once you understand the layout, the protection circuitry, and where shorts usually hide, it becomes much less scary than it looks.

Build your own hardware wallet

Kits, smartcard hats, cases and parts for SeedSigner, ShieldSigner, Specter DIY and Satochip builds, tested with the firmware shown in these videos and shipped from Canada.

Why this Smartcard Hat exists

The goal is simple: give SeedSigner a way to talk to an SEC1210 smartcard controller so you can use a SeedKeeper or compatible JavaCard without needing a separate USB reader.

For basic operation, the hat only needs a few GPIO pins on the Raspberry Pi:

  • 5V
  • Ground
  • UART TX
  • UART RX
  • Extra grounds for mechanical strength

That last point matters more than people expect. If you solder too few header pins, the stack can feel loose and become annoying to remove later. I usually solder enough pins to make the connection solid without overdoing it.

The board also includes USB-C power input so the Pi and smartcard controller can be powered through the hat. That makes compact SeedSigner builds easier, especially when you want everything stacked tightly together.

Starting with an older board: what changed?

I started by disassembling an older revision of the board to salvage parts and show how headers come off cleanly.

Older Raspberry Smartcard Hat PCB showing smartcard slot and USB-C connector
Older board close-up

The main reason I upgraded from that style is protection. The older boards used a resettable polyfuse, which gives you some current limiting, but not the full voltage/current protection I wanted for USB-C input.

USB-C power can be more forgiving on paper than it should be. Depending on the supply or cable condition, you may see voltages higher than expected. A Raspberry Pi Zero is generally rated around 5V/5.25V territory, and I don’t want to tell people to build a custom hat and then find out the hard way whether their power source is safe.

So the newer board uses an eFuse. That gives you:

  • Current limiting
  • Overvoltage cutoff
  • Better protection for the Pi and other stacked hardware
  • A visual indicator through the status LED

If your power supply is out of spec, I would rather the hat simply stop working than risk damaging the Raspberry Pi or other parts in the stack.

Tools that make hand assembly much easier

You can absolutely build this without fancy equipment, but two tools made a huge difference for me:

  1. A solder sucker for removing headers
  2. A cheap USB-C powered reflow hot plate for surface-mount components

The solder sucker is great when you need to remove pin headers from an existing board. Instead of fighting with wick and heat, you can heat the pins, suck out the molten solder, and lift the header off cleanly. One thing to remember: point it away from your work area before releasing the button, because it blows excess solder back out.

For the surface-mount parts, reflow soldering is the easiest path I’ve found for hand-built boards. You apply solder paste, place the components approximately where they belong, then heat the board so the solder melts and pulls the parts into alignment. It’s much more forgiving than trying to solder every tiny part individually with a pencil tip.

I used a USB-C hot plate set around 215°C. That is more than enough for typical PCB assembly in this context. You don’t need a complex temperature profile for a small DIY board, but you do want to avoid leaving it on the plate too long and overheating everything.

Files, BOM, and how I keep assembly sane

All the design files are available in my GitHub repository under the SmartcardHat electronics folder. This includes production files such as Gerbers, part placements, build materials, KiCad project files, schematic PDFs, and an interactive BOM file.

GitHub repository showing SmartcardHat files
SmartcardHat GitHub repository

The interactive BOM is the most useful file for assembly. It shows the bill of materials on one side and highlights parts on the PCB layout when you click them. That makes it much easier to work through components one by one without constantly flipping between a schematic, spreadsheet, and physical board.

If you are getting the boards fabricated, I recommend using 1mm thickness if you want to squeeze the hat into compact SeedSigner builds with display hats or other stacked hardware. Thicker boards can still work, especially if you are just stacking directly onto headers, but thin PCBs give you more clearance when everything is packed tightly together.

The layout is also designed for low-cost fabrication and easier hand assembly. Tracks are deliberately generous, pads are large enough to accept common component sizes, and the design favors cheap, available parts over maximum density.

Solder paste: stencil first, touch-up by hand

For this build, I applied solder paste using a 3D-printed stencil made from KiCad data. It’s not as perfect as an expensive production stencil, but for a couple of boards it works well enough.

Solder paste being applied to Smartcard Hat pads
Applying solder paste with stencil

The process is:

  1. Clip the stencil over the board.
  2. Spread low-temperature solder paste across the pads.
  3. Remove the stencil.
  4. Inspect for missed pads or connector feet.
  5. Touch up small areas by hand with a syringe or tip.

I used low-temperature solder paste because it makes prototyping easier when you need to add, remove, or fix parts later. For production boards I send out, I would be more careful about the exact materials and process, but for DIY assembly this approach is practical.

You want enough paste on the pads, especially around the SEC1210 controller and USB-C connector feet. A little extra solder can actually help here because reflow will pull things into place and you can clean up excess later with wick or an iron.

Placing the SEC1210 smartcard controller

The most important placement step is the SEC1210 chip. If this is oriented wrong, the board will not work correctly.

Interactive BOM next to physical board showing pin-one orientation
Checking SEC1210 orientation

The chip has a pin-one marker, usually a small dot or circle in one corner. On this design, I kept the orientation consistent across revisions: the notch/dot side needs to match the PCB marking. If you are reusing a salvaged chip and flux is covering the markings, clean it with isopropyl alcohol so you can see the text and pin-one indicator clearly.

A useful trick is to look at the underside of the package as well. On some small packages, the underside notch or shape can be more reliable than trying to spot a tiny dot under flux. If in doubt, use magnification: a phone camera with zoom, a magnifying glass, or a microscope all work.

The SEC1210 comes in variants such as the single-channel URT and dual-channel UR2. For this application, they can be interchangeable because the driver supports the CCID interface. The important thing is that you get the orientation right before reflow.

Placing passive components: approximate is okay

Once the main controller is down, the rest of the board is mostly passives: resistors, capacitors, TVS diodes, LEDs, and eFuse-related parts.

The interactive BOM makes this much easier. You click a component, see its value and location, place it roughly on the paste, then tick it off as you go.

A few practical notes from the build:

  • Capacitors are usually not polarized, so orientation matters less unless the part is specifically marked.
  • Resistors can technically work upside down, but I keep them facing up because troubleshooting becomes much easier later.
  • The eFuse package needs to be placed carefully enough that you can still see the top and bottom pads after soldering. If too much solder pulls it out of alignment, you may have a bad connection or short.
  • USB-C plug orientation matters. I tend to slide the connector slightly toward the edge during reflow because if it sits too far back, it can cause shorts under the package.

The 5.1k resistors on the CC lines are important for USB-C power negotiation. Without them, newer USB-C or PD supplies may not detect the device and simply will not provide power. Older cheap supplies might work anyway, but if your board does not power up consistently, check those first.

eFuse voltage and current limits

The eFuse is one of the main design reasons to use this hat instead of just wiring USB-C straight into something fragile.

Schematic showing eFuse protection section
eFuse schematic section

The board is configured so that if input voltage goes above roughly 5.2V, the eFuse cuts off output. It also has current limiting behavior to protect against shorts and excessive draw. The exact threshold depends on resistor values in the divider network, but the design goal is simple: if the power supply is too high or something is drawing too much current, the hat stops rather than frying the Pi.

The status LED sits on the eFuse output side. That gives you a visual clue that the eFuse has not tripped and that 5V is actually reaching the rest of the board. If the LED is off but input power looks fine, that is an important diagnostic hint.

For current limiting, I wanted enough headroom for a Raspberry Pi Zero with SeedSigner running, while still protecting against unexpected shorts. The traces on the newer boards are thick enough to handle reasonable USB-C current without getting uncomfortably warm under normal use.

Reflow soldering the board

Once all components except the card reader header are placed, it’s time to reflow.

Board mounted and ready for reflow
Reflow-ready Smartcard Hat

I set the hot plate to around 215°C and used a timer for about five minutes. During heating:

  • Watch parts as they begin to move.
  • Use tweezers or a non-conductive tool to nudge tombstoned components back down.
  • Press gently on areas with excess solder paste to help squeeze out blobs.
  • Use solder wick on large solder bridges if needed.
  • Check the USB-C connector and slide it slightly toward the edge if it looks too far back.

Reflow is forgiving, but it is not magic. If a part is completely in the wrong place, you still need to move it before or during heating. The surface tension helps pull parts into alignment once they are roughly close.

After reflow, I visually inspect for:

  • Solder bridges between pins
  • Missing solder on important pads
  • eFuse top and bottom pads visible
  • USB-C connector seated properly
  • No obvious blobs connecting neighboring traces

Finding and fixing a short under the USB-C connector

The most common problems I have seen with this board are shorts under QFN packages or inside/under the USB-C connector. Fortunately, those are usually fixable without scrapping the board.

In this build, the first continuity check showed 5V and ground shorted together. That means nothing will work until it is cleared. The likely culprits were:

  • Solder bridge under the eFuse
  • Solder bridge around the SEC1210 controller
  • Short inside or under the USB-C connector

I applied flux around the QFN packages and ran a soldering iron over the edges. Flux helps melt existing solder and allows it to flow back onto the correct pads rather than staying as bridges between pins.

Multimeter checking for short on Smartcard Hat
Checking continuity after reflow

The USB-C connector was still suspicious, so I reheated the board and slid the connector slightly toward the edge. Once it was seated correctly, the short cleared and the meter showed no continuity between 5V and ground.

This is why I recommend checking shorts before powering anything up. A dead short across USB-C input can be annoying to diagnose later, especially if you have already stacked a Raspberry Pi on top of the board.

Power-up test: current draw and eFuse behavior

After clearing the short, I powered the board through a current-limited supply. With nothing else connected, the smartcard controller alone draws only a few milliamps, with the LED adding another small amount.

Powered Smartcard Hat showing status LED and meter readings
Power-up test

The green LED came on, which told me two things:

  • The eFuse was not tripped
  • 5V output was present on the Pi side of the board

I then raised the input voltage above the overvoltage threshold and watched the LED turn off. When I brought the voltage back down, it came back on. That confirmed the protection circuitry was doing what it should.

This is a useful test to run on every hand-built board before connecting it to anything valuable. If the current draw is much higher than expected, stop and look for shorts or misoriented parts.

Soldering the card reader header

The final hardware step is adding the smartcard slot or SIM-sized reader header. I usually add this after the main board has been tested because a full-size card slot can block access to chips and pads while you are debugging.

Soldered contacts on smartcard reader PCB
Card reader header soldering

The pads for the card reader are intentionally large, which makes hand soldering much easier. The main things to check are:

  • No bridges between adjacent contacts
  • Card presence detection pins soldered properly
  • Connector aligned with the PCB cutout
  • Reader seated fully before inserting a card

Full-size card slots tend to be more reliable for regular use than very small SIM-sized readers, especially if you are cutting cards down or handling them frequently. The smaller form factor is useful for compact builds, but it can be fiddlier.

End-to-end SeedSigner test with SeedKeeper

Once the reader header was soldered and inspected, I connected the hat to a Raspberry Pi Zero running SeedSigner and booted it up.

The first check was in the settings menu: list card readers. The SEC1210 controller appeared, which confirmed that the Pi could see the smartcard interface over UART/CCID.

Then I ran the generic smartcard test. This is important because it verifies communication with a physical smartcard regardless of what applet is running on it. If this passes, the hardware path from SeedSigner to SEC1210 to card reader is working.

Finally, I inserted a SeedKeeper card, entered the PIN, and loaded a seed from the card into SeedSigner. That completed the full end-to-end test:

  • Pi boots
  • Smartcard controller detected
  • Generic smartcard communication works
  • SeedKeeper applet responds
  • Seed loads successfully

If you are building one of these hats, that final sequence is what I would consider a successful build.

If you don’t want to hand assemble it

You can get the board fabricated and assembled elsewhere, or order prebuilt versions from my site if you just want a working Smartcard Hat for Raspberry Pi with USB-C input power: Smartcard Hat for Raspberry Pi with USB-C Input Power.

If you are in a country where labor costs make hand assembly expensive, fabrication is usually the better route. If you want maximum privacy, custom modifications, or just enjoy building hardware yourself, doing it by hand with reflow soldering is very doable once you have the files and a few basic tools.

Final thoughts

This hat is not complicated electrically: mostly UART, 5V power, protection components, and a smartcard controller. The challenge is in the physical assembly and testing. Get the SEC1210 orientation right, check for shorts before power-up, verify eFuse behavior, and then test end-to-end with SeedSigner and SeedKeeper.

If you have electronics experience and see something that could be improved, I’d rather hear about it than keep a suboptimal design around. The point of publishing this is to make the reference design useful for other people, not just for my own builds.

Build your own hardware wallet

Kits, smartcard hats, cases and parts for SeedSigner, ShieldSigner, Specter DIY and Satochip builds, tested with the firmware shown in these videos and shipped from Canada.

Found this useful?

These guides are free and ad-free. You can support them by sharing them, subscribing on YouTube, or sending a tip.

Stuck on this?

Book a one-hour session and I’ll walk you through it, or ask about a trusted recovery.

Recovery guides

More guides