A while back I bought enough parts to build two of these, but as life does it intervened on my project. I want to move forward now but I do need two mounting boxes. Do you have these available? I don’t have a 3D printer.
Thanks.
Which box? The project is designed to replace the faceplate, and the switch already installed in the wall. He does offer the stl file for a desktop box. I use both.
Before I bought a 3D printer, I sent my stl files to Bellair3D (bellair3d.com). He does high-quality prints at a very reasonable price. Highly recommended. Send the stl files to him and he will reply with a quote.
Thanks for the reply. I’m looking for the cover plate and the box that the pcboard gets mounted into.
This will allow me to install all my parts and mount it in the wall.
Thanks.
All of the stl files are here.
I can’t speak for Luma if he still sells the faceplates, but the original kit I ordered from him included the faceplate, and I can’t match the quality.
I have four of these, three in the wall and one on a desktop. They are remarkable devices.
I’ve been able to successfully print new faceplates, as well as modify one for a custom need I had.
He has recommended settings, which I used, however I used ABS myself for extra strength but I think he suggested PETG.
The key for me was using a glass buildplate (for that shiny, extra smooth surface) and many bottom layers (I don’t recall, maybe 6?).
I know this is an old thread but do you have the stl file for this plate? I like your mod.
Not sure if you were asking me but the faceplates are on his github page.
The “custom need I had” was that I needed to make it oversize to cover some wall damage. I can upload that if you want but mine is a 4 gang decora version.
Thanks for getting back to me. Sorry for the delay. I did finally find the single gang stl files and started to print them out. Unfortunately, I am having a couple of issues with my 3d printer so this little project is on hold til I sort the issues out. Thanks again.
I’m still moving forward on my HASPone projects. In viewing Luma’s stl files he has one labelled Assembly. Do you know where I can find any discussion of this?
Thanks for the help.
That’s a jig to help with soldering - if you print that, you can place all the components on the board, place this jig over the assembled PCB, then it’ll hold everything in place when you flip it over to solder the joints on the bottom of the board without the parts all falling out. Coming from someone who has soldered up several hundred of these things, it’s a life saver 
Thanks Luma. I will definitely have this printed. Also, I want to thank you for this awesome project. I’m backlogged with projects right now but I can hardly wait to get these running and integrated into Home Assistant.
1 Like
Luma:
I got my insert nuts from Amazon, only to discover that they sent me the injection molding type. OK amazon a return is coming your way. In the meantime another question has come up.
Have you ever tried using the ESP32 C6 version. I have a dev kit from ESPRESSIF that has one in it. They claim one can run it on WIFI 6. I have never been a fan of 2.4 G as it has all kinds of issues with interference. I’m a ham operator so I have had experience with this. The 5G band does not have this issue and the neighbors 2.4 G wifi running at full blast won’t interfere. Plus I am a big fan of Zigbee and have a ton of stuff running on it, plus a Zigbee Hub in my barn. So, a couple of most intriguing alternatives are presenting them selves. I would be interested in your comments. Thanks again.
I call BS.
There is potential interference, but rare. The real issue is RF desense. WiFi operates at higher power levels than Zigbee, so in a saturated 2.4GHz area, a Zigbee packet has to retry more often than not to be successfully received. Multiple Zigbee routers increases the likelihood of packets being received.
As I said, actual interference is rare because data packets in both protocols are small and there is dead time between packets as short as 10 µs or stretch into hundreds of microseconds. This interframe spacing is intentional as to let other clients have an opportunity to send a packet.
I have more than 100 WiFi clients and a couple of dozen Zigbee devices. All live together happily.
Oh, to your original question, Luma shares the source code, so help yourself to port it to the ESP32. It should be pretty straightforward. Share it when you get it working.
Well, consider yourself a lucky man then.
Happy Friday.
I can confirm Marshall does EXCELLENT work. My faceplates and boxes were well made and fit perfectly. I vote YES.
You’ll run into two challenges. The first (mentioned above) is that the code would need some work to port it over as ESP8266 and ESP32 have some differences in how things like the WiFi stack are handled. I haven’t tried this but it’s simple enough that I imagine modern code gen tools might be able to do most of that work.
The second challenge would be re-designing the PCB to fit a new board footprint, then getting a production run of the resulting boards. At one point there was an ESP32 D1 mini-ish dev board but it added a second row of pins which made it unusable in the very tight space we have inside this device. Since then, smaller ESP32 boards have been made available so it seems do-able. I likely won’t be doing that simply because the ESP8266 meets all the needs here but, being open source, the option is there for you. I don’t know of any AI tooling that can handle KiCAD quite yet so the PCB redesign is likely to take some EDA skills. It’s also a simple project and might be a good place to learn those skills!
I used to do a lot of pc board design work in a past life so that doesn’t put me off. It’s the code that will be a challenge. I’m an engineer not a developer (you know engineers program with soldering irons). However, if I can convince my developer friend then we might just take a run at it.
Now, this is all I need, another project on my desk, hee, hee, hee but I think in the long run it would be a great achievment. Thanks for the reply.
However, I want to get your version running on my first HASPone before I start anything else. Cheers.
Thank you for that clarification. I buy my Wemos D1 Mini modules five at a time for exactly that reason. They are cheap and have all the resources I need for my projects.
I purchased about a dozen HASPone bare PCBs, Wemos D1 Minis, and (the biggie) Nextion displays quite some time ago with the intention of modding them to use PoE (i.e. Cat5e wired) infrastructure instead of WiFi. I’m finally getting back to that part of my renovation project and wanted to pick your brains on how to accomplish this.
Let me say first-off that my decision to not use wireless is set in stone. We can discuss why in a separate thread. Here, I’d like to focus on technical approaches to deploying HASPone with minimal firmware changes in a PoE environment. From 20,000 feet this seemed plausible since I’m “just” replacing the Phy layer. Of course it’s never that simple.
For the basic PoE interface I’m using the PoE Featherwing + RP2040 shim developed by Silicognition. I use this for my regular light switches, which will talk Webhooks to HA automations over Cat5 instead of switching 120V.
I’ve had several approaches rattling around my head but nothing has really “clicked” yet. They all boil down to where to cut the HASPone stack and stick in the wired ethernet part:
- Lowest level - ESP8266 controls the W5500 on the PoE Featherwing
- Mid level - Handle all MQTT request/responses in the RP2040 PoE Featherwing. Replace those layers of the ESP8266 code with a shim to the RP2040.
- High level - Port the HASPone FW to the RP2040 (replacing the L1 code) and have it talk to the Nextion.
- Use OpenHASP instead, requiring it be ported to the RP2040 as well as abandoning those Nextion displays I bought. An RP2040 port however looks like it would be 100% on me to do.
- Hope that Silicognition successfully crowdfunds the ESP32 Single Pair Ethernet with PoE Patrick is considering, and use that as an OpenHASP platform. No PoE Featherwing + RP2040 needed, allowing their use elsewhere. Since the OpenHASP folks are all-in on the ESP32 that looks easiest from a SW angle. Should work on my Cat5E wiring. I’d still have a dozen Nextion displays left behind that I’m unlikely to use.
So if you were in my boat, what would you do?
Use WiFi.
What you are describing is yet another dashboard. You could build on the excellent work by @luma but in the end it would not be HASPOne.
Limiting a dashboard to Ethernet is a market of one.