Interface PCB design.

[Spud2233]

I have been playing around with an interface circuit design to combine the MOSFET level shifter and an NPN transistor network #7 onto a single PCB.

The above circuit is a work in process, It's working for me, but circuit design is not my specialty so I'm open to any critiques or recommendations.

The current layout is designed to plug straight into raspberry Pi GPIO header but I will also look at designing some other layouts, particularly one that can connect directly onto the battery.

[Tool-Scientist]

Good stuff. One addition could be an optocoupler above/below R3 that is driven by Serial 3.3V. That way if you connect an unpowered adapter to the battery, it won't pull J2 high and increase the charge count. This depends on the serial adapter being able to supply enough current to turn the opto on. There may be a better way to do this, but optocouplers are usually the easiest high-side switch.

It might also be worth considering an opto-isolator level shifter, so that we're not linking the 21V, 300-amp-capable battery to the computer.

For the problem of bringing J2 high when the script isn't running, maybe we could use the default state of DTR to hold either J2 or Serial-TX low until the script is running.

Here's a falstad example. I've set the opamp to output 2V on J1 to show the working of the Spud-isolator™. Sorry it's so messy, I switched J1 and J2 to match the battery layout (I wish Milwaukee had numbered from ground up like everyone else) and falstad isn't very good at moving things around. All the internal stuff (inside dotted boxes) is entirely fictional.

Adapter unpowered has J2 at 0V, Adapter powered has J2=3.3V (not enough to increase charge count), then turning on DTR allows TX to put 12V signals on J2.

Falstad example

[Spud2233]

I picked up a few optocouplers from jaycar today, so i will start prototyping a fully isolated design this week.
I am also looking at including an optional 3.3V power supply to power a microcontroller i.e. ESP32 or raspberry pi zero / pico from the M18 battery.
I hoping to build a unit that will provide basic info like serial number, cell voltage, cell balance on an LCD to allow quick checking of batteries without needing a PC.

[zRCrackiiN]

Hey all, so I've been kinda watching from a distance, I had a crack at doing a pcb for my batteries.

Unfortunately I didn't read about the issues with cp2102's 😂

Back to the drawing board for me i guess!

[zRCrackiiN]

so when i run m.idle() J2 = 0.109v and J1 = 0.975v

m.high() J2 = 12.6v and J1 = 2.258v

[Spud2233]

I've been playing with the below design for an isolated interface, hopefully I will be able to build a prototype this weekend.

• Fuses on battery terminals
• Optoisolators for all data lines
• Isolated 9-27V to 3.3V DC-DC converter with output control via opto to power down the converter when the board is powered via external source. This can be disabled by leaving the CTRL pin floating by removing the 68ohm resistor.
• DTR signal to hold J2 low until called in script

Falstad M18 isolated interface Demo

[Jason-Root]

Please keep us updated this is exactly what I am interested in a stand alone unit to check batteries in field. Custom pcb usually have a minimum quantity you can add me to the short list of buyers when your ready

[tanker665-pixel]

Count me in too

[kotor2knight]

i would throw in for a PCB as well, this whole project is great!!

[Spud2233]

After working on this on and off for the past few weeks I have the first version of the isolated interface prototyped and tested on the breadboard, I have 5 PCB's on order, once I have confirmed the PCB works I will upload the BOM and gerber files.

@Tool-Scientist - I was able to prevent incrementing the dumb charge count when the script is not running, but was unable to prevent it incrementing with the USB interface unpowered, if this version works the next version will include an isolated 3.3v regulator powered from the battery that should eliminate the issue.

[Tool-Scientist]

Was TX unable to drive opto directly? I solved the "unpowered adapter" problem by having TX and DTR in series instead of parallel. That way it needs both of them on to bring J2 high

[Spud2233]

Tx could drive the opto fine, but it inverts the signal, It's fine for the genuine FT232's as you can invert the TX pin using FT_PROG 3.12.70.678, but it wouldn't work with any other serial adaptors.

When plugging in the FT232 adaptor I have i found the DTR pin bounces a few times before settling at high, a few of these bounces exceed the threshold that triggers the dumb charge counter.

[DarthTaters]

That's fantastic. Nice work, I'll be excited to see how this comes out.

[Tool-Scientist]

Tx could drive the opto fine, but it inverts the signal

This is what I designed a few weeks ago. Untested, though.
TS Opto

When plugging in the FT232 adaptor I have i found the DTR pin bounces a few times before settling at high, a few of these bounces exceed the threshold that triggers the dumb charge counter.

Maybe capacitor?

[Spud2233]

My Interface PCB is complete and tested.
PCB Schematic, BOM, Gerber & Case STL files are on my GitHub, and will be added to this project in the next few days.
https://github.com/Spud2233/m18-protocol/tree/4503db3b5ae36d714da77aad857d4dc701c842c0/Interface%20PCB%20Isolated%20SMD

Schematic_Isolated-M18-interface_2025-11-03.pdf

Battery + & - are protected by 1.5A fuses as well as a 50mA PTC fuse
J1 & J2 are protected by a 50mA PTC fuses
Battery side is electrically isolated from the PC side by 4N33 Opto-isolators (5000V isolation voltage)

To prevent incrementing the Dumb charge count the serial interface must be powered

[mnh-jansson]

My board also works, but i will make some revisions before releasing it.

Nice job! @Spud2233