Simple extra battery for the expansion bay

It’s more the current levels, than Wh capacity. Lower voltage and higher Wattage (Power) gives higher Amp (Current). Which means you need to size up wiring, heat dissipation, etc.

60W@3.7v nominal (single cell LiPo) would be about 16.2a of current. 4 cell series is 14.8v, 60W gives about 4a of current. A LOOOOOT easier to handle.

Something like Protection Circuit Module (PCB) for 14.8V Li-Ion Battery Pack (10A limit) with Fuel Gauge Socket would be a fine BMS for this, able to handle up to 10a in/out continuous. I’d limit to 6 or 7 max just be safe of course.

And for max capacity, I’d use pouch cells, since they’re square-ish, so less void space.

If you’re physically removing it as well to “hot swap”, you need to also have some kind of protection container around it, so that cuts slightly into the volume as well. Plus you actually need an hot swap insert plug connector to hook up to the interposer to hot swap the battery in and out.

If you’re not hot swapping the battery in, that does simplify it. And yes, if the total size of the expansion card is enlarged a bit you can get a bit more space just fine, and not worry about protective case and insertable connector.

No, as @Blacksmith12 said, the challenge will be the interposer & logic circuit to handle the power out, and power in (charging from the laptop when plugged into the wall). In looking around for some other use case, I found BQ25713 data sheet, product information and support | which actually looks perfect for this, as it supports talking about the USB-PD stuff, talks I2C for a controlling MCU (probably some small Atmega or M0 or something, could be running Arduino perhaps?). Still not simple or easy.


That for sure, I mean there isn’t a big difference in capacity. This was in response to

Ah. It’s more that as you pull more current, actual usable capacity goes down after a certain point. Thus with a higher voltage (more cells), you can have lower discharge current, and get more usable Wh of power. See the Discharge Rate Characteristics charge in this battery cell specs sheet At 0.68a, capacity is ~3500+mAh, whereas at 10a it seems to be more like ~3300mAh, give or take. So by having more cells, even though each cell has lower capacity, you end up with more usable capacity. Also a lower discharge rate lowers the stresses and typically means longer life span.

This is a lot of super helpful info. You obviously know more than I do about stuff like this. Thanks for the help. I am currently just trying to get physical dimensions right now. The logic will come later; I plan to have an external contractor create the PCB and logic for the Laptop side of things and, if I need to, the battery side. Obviously, I would like to design it myself or get some sort of collaboration, but we will cross that road when we get there. In response to @Term_Grecos, that’s a great idea. I intend this expansion to be as sleek as possible since I’m figuring people who need the extra battery will be traveling to where slimmer is preferred, but I’m sure a little bump at the back can be justified for some extra capacity.

LOL, that’s oddly endearing. At least if you don’t have to put up with a friend stalking your forum use and trolling you regularly :wink:

Actually, I’m thinking that single chip might be all you need, almost. Doesn’t seem that it does balancing or per-cell low voltage/over charge detection. However, through a few short searches, found that BQ77915 data sheet, product information and support | can do the balancing and protection. Or perhaps BQ40Z50-R2 data sheet, product information and support |, which has SMBus, so can do battery gauge/etc.

I’ve actually been looking at the BQ25713 myself to do a custom board to combine a lot of things for a LED light up cane, and having to have the charging be external is a bit annoying, so I can’t just have a USB-C input and charge off of any USB-C source, even if slower with one lower than 50W source capacity (running 4s2p, figuring at 3a ~50W is the fastest I can go). That’ll be for another day though, to design the custom PCB and get it made. I’m NOT an electrical engineer, and I’m going to have to find some that’ll be willing to review the design to make sure I’m not totally screwing things up. At that amount of power the pack can contain, it’s potentially quite dangerous.

What’s also nice, is it’s right at about 99.2Wh nominal capacity, which fits under the 100Wh battery limit the FAA has, so I don’t need airline approval. =D

I do really want to design that charger though sigh


So found the Expansion Bay Specs thread, had a question on it, around this Expansion Bay Developer Program - #9 by Aaron_Baff

Ideally you would want a hot swappable system that gets screwed in as normal and then you can yank the batteries out and shove in a new one. But given how much battery capacity you get with the regular battery, anyone who is going to complain about size is going to have to accept compromises. Small battery packs are going to degrade far more than larger packs, especially if there is no way to set charging and discharging capacity. On that note please do not touch pouch cells if you are going that route. They are compact, yes, but they always fail badly and damage components when they do.

You will get far more people wanting a large battery pack over a small one.

Rather than that, have it stick out at the bottom like this:

That way it can have more volume while still being transportable.
Many reviewers complained about the FL16 being less mobile with the GPU module attached.


I thought about potentially having a bottom tray underneath the laptop, but I don’t know if intake fans will allow it. Also, that can make it less transportable, which would be why a hot swappable expansion bay would be best.

I believe the fans mostly come in from the back and vent out the sides (or reverse?), rather than anything underneath. I could be wrong about that. Obviously that’d add to the thickness, but could add a LOT of battery. However, something to worry about is total Wh, since there’s an FAA limit of <100Wh before needing airline approval. Probably would be fine if it’s just a large battery pack on the bottom of the laptop, they probably wouldn’t even notice it even if it’s >100Wh.

The fans pull air in through the bottom and top (the opening above the keyboard) and exhaust out the sides (and with a GPU module the back, but since this battery would be instead of a GPU module that’s not relevant).

Just leave the intake area out. You’d have to do that either way, even if not expanding to the bottom.

Just have the expansion bay act as a connector for the hot-swappable batteries.
That way the expansion bay itself doesn’t have to be hot swappable; only the battery packs have to be. And as they are hot-swappable, just take them out before transportation, if necessary, and reattach them again upon use.

Working on version 2? 1.5?, I haven’t really shown the version 1 publicly. Anyway, I have been taking a bunch of the considerations you all have made and improving the original design. The calculated capacity for this one is 17.76 Wh. I’ll post a larger update soon.


Battery Pack Version 2


  • 17.76Wh battery
  • 66.5mm x 132mm x 9.7mm
  • LED battery indicator at the back
  • Housing and case are two separate pieces for ease of assembly and repair
  • less complex
  • thumb grip for ease of pulling in and out of slot


  • The back of the battery is not flush with the module
  • capacity is still limited and not within the legal flying limit**
  • shell is thin to accommodate the most capacity
  • The rear stock cover needs to be modified along with other parts of the original frame to fit correctly (will most likely need to be replaced with a 3D printed chassis)
  • tolerance issues with the frame rails
  • The battery pack can only fit one way due to the Z asymmetry of the rails
  • The cage is flimsy and warps out of the shell (I am guessing that the added PCB and cells will be enough to stiffen the frame)

**The FAA is unclear if two separate batteries carried by one device are considered one size.

Further development:

  • I need the physical expansion bay module on hand to test Z clearance with the laptop. I also need to check if replacing the stock frame with a modified one is possible.
  • Currently, the dimensions and circuits for the battery pack and the expansion module have yet to be specified. The most that has been discussed has been from @Aaron_Baff.
  • The battery cells used to get the Wh estimate are from here (Model No. LP703051). I’m considering having a custom pack made to increase capacity further.
  • I’ve been looking at the USB PD standard to see if the theoretical 14.8v the pack can supply will be adequate for the 45w spec.
  • Looking at routes for development and ways of distributing products of this type. Thus far, the plan is to keep everything open source (it makes no sense to do anything else). 3D printing the cage, along with a list of parts for assembly and instruction, seems the way I’m going to go with it. Along with having assembly kits and/or complete units available. Currently, the battery shell and cage are hard to print; I’m going to work on optimizing the design for printing once a final design has been reached.

I would love to hear feedback (criticisms, improvements, thoughts, ideas, etc.)

Some more photos of version 2
note: the USB C plug is not designed yet but does fit within the current dimensions


For the FAA limit…just pull it out and carry it separately. Fixed. They’re fine with multiple <100Wh batteries that are carried/stored reasonably.


This is awesome!
I probably won’t be using this as I’ll be running the dGPU, but I love that you’re putting in the work for us :​)

Regarding distribution: Wouldn’t it possible to use the framework marketplace for that, if the result is good?

I also plan to use the dGPU, so I won’t be using my own product. But I plan to create a hideaway second monitor that clips onto the screen. I could technically make a similar design for a clip-on battery module that mounts on the display frame, mainly for people who use the dGPU on the go and need the extra capacity. But that’s if it seems reasonable to develop.

I do plan to do this, but it’s too far down the road to promise that’s the case.


Hello Blacksmith,

it looks like you can get this working elactrically. If you can get it to work with 4 cells can you please test it with 4 18650 or 21700 cells externaly wired.
I dont know if it is possible with 5 cells, because that would be better for capacity and it needs lower current for the same power output.

For capacity you can see what is possible in my first post here:

1 Like

The PCB/BMS is not nearly that large, generally. And I think it’s also been mentioned, but putting the 18650s there will interfere with the cooling substantially, which is a problem.

The non-GPU expansion bay fans only exhausts out the sides, so 18650 along the back won’t interfere. Nrp even mentioned that location for cells in an early video about the FW16.

1 Like