5V * 0.0011A = 0.0055W, So it would discharge a 61Wh battery in just about 11k hours (ignoring conversion losses and stuff of course) XD.
Not 100% sure you divided the right things from each other there.
I divided the rated mAh value from the battery (which is btw. 85 Wh) by 1 mA. With 1.1 mA I get pretty much exactly 5000 hours.
Edit: Is the error here the conversion from battery voltage to the chip’s supply voltage?
Edit 2: Also, why did you use 5 V instead of 3.3 V for your calculation?
That only works if you are consuming the power at battery voltage directly. For stuff like this working with watts and watthours is much easier even if the conversions are never 100% efficient.
Yes, the pico would not be too happy with 18v and neither would most other usb devices XD.
There are all kinds of dc to dc conversions going on in a laptop, 5v rails for the usb, very high current ~1.3V for the cpu, 1.2V for the ram, 3.3v for internal devices and a bunch of other stuff.
Your picture looked like you were using an ldo for the 3.3v rail, on an ldo the input current is the same as the output current (it just burns off the voltage diff as heat) and the laptop is supplying 5v to the usb port.
1 Like
That would be about right. Remember the oscillator is still going so there will be a certain percentage of the gates inside the chip still switching. If this is using a crystal it is certainly in the ball park I would expect.
This actually works remarkably well! Had to size down the voltage regulator, though, compared to the suggestion from the framework design files.
This is a 4-layer version, with ground pours and signal lines on the top and bottom and a ground and a 3.3 V plane internally.
The next step is now finding the right connector and switch for the final v1 design.
1 Like
Yeah an ldo should do for most cases instead of the fancy switching regulator the pico uses.
4 layers probably makes the layout less of a pain. I made a very small rp2040 board once but I used 0805 components and soic8 flash so I could hand solder it and the routing got extremely cramped. I should probably figure out how to do layout on more than 2 layers at some point, especially since it got quite cheap now.
Edit: you may want to add at least some test-points for reset and bootsel for at least the one connected to usb, probably also for the other one and it’s usb lines.
That was already in the design, I just replaced it with a smaller footprint package.
Sure does, I really wouldn’t have wanted to route all the power on the front or bottom layer, too. Might not even have worked out in this case.
I’d say it’s pretty much the same as doing two layers. Doesn’t make much of a difference to me, at least.
Good point ^^.
My projects so far have all been hand soldered, too. Went down to 0604 components, that’s where the limit is for me with hand soldering. (I guess if I really had too, 0402 would also still be possible, but really painful ^^.)
This is the first project that requires me to go even smaller.
It certainly wasn’t easy but the board works. I did only need a few pins though so I was able to ditch a few components. the board was only about 1.5 rp2040s wide and about 3 or 4 long XD.
That’s great! I certainly have missed some very important connections before or had to make manual adjustments afterwards 
.
A sharp knife and good soldering skills are certainly very helpful in that case.
From my experience this really is something that needs time to learn and will get better and easier with each project. This is probably somewhere between the 10th and 15th board I designed.
If you want a really deep dive into PCB design, I can also highly recommend this video.
Four layers is the absolute minimum I would use for these sort of chips. Use of ground and power planes is essential to minimise power problems.
Kicad can do multiple layers quite easily, far more layers than you could dream about. It is quite easy to configure for this.
I have done multiple boards with the rp2040 on only 2 layers and all of them worked fine so far.
I don’t doubt it but I do have to do it XD.
Hi! I’m sorry if I’m not following the thread of conversation correctly, I’m new to this format.
I wanted to know if you’d made any progress with this.
I also wanted to point you to this device, the T-PicoC3, which has an RP2040 and an ESP32-C3 on the same board. They communicate via UART. It might be useful for you.
T-Pico C3 schematic
type or paste code here
Hi,
Thanks for asking!
I was basically ready to order the boards, but then realised that because of the size constraints there’s some extra cost (e.g. for smaller vias) I had missed and especially getting the parts placed is quite expensive. My previous projects I always soldered by hand, but that’s not possible with this.
As I’m currently just a student, I haven’t yet been able to afford that.
It’s actually not that expensive for a single board, but you have to order multiple and the parts placement has quite a high set-up fee regardless of how many boards you order. I believe it was about 200 € for 5 Boards overall.
If there’s a couple people interested in sharing the cost (and risk of it not working) and maybe get a working board out of it, I’d be happy to take care of all the work and share the design files with you so you can check them yourself. It’s just the financial side holding me back right now.
I will definitely post an update here if I find another solution.
That’s interesting, thanks for sharing!
If you’d be interested in selling them on Tindie and having someone look at the boards to double check they’re good, I’d be willing to fund the up-front cost.
1 Like
That would be amazing!
I had actually already thought about selling the additional boards to reduce the cost for me, so I don’t mind that at all.
What do you mean with having someone look at the boards? The design files or the actual physical boards?
I already have the design files on my GitHub. But I’ll clean the repo up a bit, it’s currently still a bit of a mess .
I’d also be happy to send someone a board to check it or something like that.
Edit: I just saw that Tindie only allows to be payed via PayPal as a seller and I don’t use PayPal anymore. Are there any alternatives to Tindie?
I’m sure there are. I’ll look around. Yeah, I just meant look at the files. I only mentioned that because you said ‘and risk of it not working’ haha.
1 Like
There’s lectronz.
Alright, thanks for the clarification.
I mentioned it because there’s always a certain risk with a first prototype that something might not work as intended. Often that still can be manually patched, but I’ve had issues before where that wasn’t possible either. So I just wanted to make that clear, this is a relatively complex project after all.
Thanks, I’ll check it out!
Edit: I just cleaned up the repo. I left the mechanical part from the reference design in there for now but in the electrical part there’s now only the relevant project left: Project on GitHub.
1 Like
So is there any update on this?
From what I read this project is ‘in hiatus’ due to sizing issues, is this correct?
No, size wise this project is quite doable. Currently it’s an issue of financing the order of a few first boards, as discussed recently.
@Earnest_Williams_Geo did you have a chance to check out the design files? Or did you want someone else to do that?
Btw. I’ve set up a store on the site you suggested: Lectronz.