Hydro Core
HydroCore is a compact interface board designed to connect a Pixhawk flight controller and Ethernet-based systems directly to a mobile phone through USB. It integrates Ethernet-to-USB, UART-to-USB, and a USB hub in a single self-powered PCB, allowing mobile phones to handle Mavlink telemetry, control, and video tasks. HydroCore also includes an upstream charging system so the phone can stay connected and powered during long underwater drone operations.
Created by
Shades
Tier 3
8 views
1 follower
CAN ⚡🚀
approved Hydro Core ago
Tier approved: 3
Grant approved: $55.00
Nice project
Shades
submitted Hydro Core for ship review ago
Shades
added to the journal ago
Updated BOM
I have 2013 Excel software. That's Why It made everything unicoded. But luckily online cs editor exists and I fixed my BOM file properly. Hope this is the end!

Shaurya Bisht
requested changes for Hydro Core ago
Can you please fix your BOM.csv, its improperly formatted, I think I don't see commas when I look at it raw. https://github.com/samdoes-stuff/Hydro-Core/blob/main/BOM.csv
Shades
submitted Hydro Core for ship review ago
NotARoomba 🚀
requested changes for Hydro Core ago
Looks good, but please include a BOM.csv in your Github repo!
Shades
submitted Hydro Core for ship review ago
Shades
added to the journal ago
Removed the Via In Pad
Removed the via in pad and updated the repo with new gerber and pcb file. And still the manufacture price is same. JLCPCB Charges the standard 6 layer price.

NotARoomba 🚀
requested changes for Hydro Core ago
You still are using via in pad (literal via under a copper pad), looking at your design it is being used for some capacitors and inductors and is not nessecary. This process costs more to manufacture so please remove any vias under passive components (resistors, capacitors, etc) the via for the microcontroller is fine.
Shades
submitted Hydro Core for ship review ago
Shades
added to the journal ago
Revision
Differential pair Revision



100R Differential Pair Width - 0.127mm Spacing - 0.253 mm Tolerance - 0.254 mm
90R Differential Pair Width - 0.147mm Spacing - 0.253 mm Tolerance - 0.254 mm
Pick & Place file updated , Gerber Updated, PCB file updated
technical_.
requested changes for Hydro Core ago
Please fix your board. There are many DRC errors, as well as conceptual errors...
Make sure to length match ETH traces, and do not use via in pad.
Shades
submitted Hydro Core for ship review ago
NotARoomba 🚀
requested changes for Hydro Core ago
Are your ethernet/usb lines properly impedance matched? That part seems a bit sketchy to me but if you're sure about it then resubmit.
Shades
submitted Hydro Core for ship review ago
NotARoomba 🚀
requested changes for Hydro Core ago
Please upload screenshots of your entire LCSC cart, and we cannot fund tools so please remove the soldering station from your BOM. Also please add in a BOM.csv in the root of your repository with what you are going to buy.
Shades
submitted Hydro Core for ship review ago
Shades
added to the journal ago
Driver Files Upload
Driver Files Uploaded in Repository for Ethernet & Uart Controller. For Ethernet it is realtek driver & for Uart controller it is silicon lab driver. You can install the driver after connecting to host device.

Shades
added to the journal ago
Whole Project Complete
Really Sorry! I can't keep up with this journal. My main journal has been posted in gitub.
Initial Concept Research (1.5 hours)
I began exploring the initial idea by studying various YouTubers and technical websites to understand different design approaches. I noticed that using separate controllers required excessive wiring and physical space. My first plan was to combine open-source Ethernet-to-USB and UART-to-USB schematics and merge them using a USB hub IC.
Ethernet to USB 2.0
Schematic
UART to USB-C Repo
Studying Open-Source Schematics (2 hours)
I followed open-source hardware designs, but discovered that many components were unavailable, back-stocked, or incompatible. For the Ethernet port, the magnetic-shielded connector required four differential TX/RX pairs, and I found mismatches between open-source examples and my needs. Still, the reference schematics helped a lot.

First Schematic Attempt & Failure Realization (1.5 hours)
After around five hours of work integrating open-source parts into my schematic, I realized the design would fail. The referenced example used USB-C-to-UART only, creating back-feed issues. The PHY controller I picked was too powerful and poorly matched because everything had to operate over USB 2.0 to match my mobile phone’s OTG limitations.

Learning Datasheets for the First Time (2 hours)
I began reading datasheets seriously for the first time in my life. I learned that datasheets include example circuits that were more helpful than any online answer or AI suggestion. ChatGPT was not reliable for exact electrical values. Going through LCSC, I finally discovered the correct controller IC that matched my USB 2.0 requirements.



Day 2
Restarting the Entire Schematic (5.5 hour)
I restarted my schematic from scratch. The earlier open-source designs still helped because some connections matched. Missing passive component values were found from Google searches, not ChatGPT. The UART-to-USB section was easy because the chip had a simple example diagram. The Ethernet controller was extremely difficult. It had so many complex connections that I nearly had to pull an all-nighter to finish it. I had to stay focused because if I left it halfway, I would forget the connection logic. I had to select a hub IC capable of upstreaming data reliably. FE1.1 was an option but had too many ports. I ultimately chose SL1.1 because its specs matched my project’s data flow requirements. I designed the AC input section that steps down to 5V. The example power circuits in the datasheet were extremely useful. Component sourcing was difficult because LCSC filtering takes practice. I added the upstream charging IC, letting the mobile phone charge while connected. LED indicators were integrated to show phone charging/battery status. I also built the 3V and 1V converters.

Beginning PCB Layout (4.5)
After completing the schematic, I started PCB layout work. This was the most time-consuming stage. I tested several board outlines and dimensions to fit all components properly. Autoroute failed every time. Proper component placement became critical. I had to switch to full manual routing. I learned about impedance, trace gaps, crosstalk, thermal vias, and multilayer routing through Altium Academy and PCB guru videos. I switched between 4-layer and 6-layer designs until I understood how to use bias and planes properly. After several routing revisions and placement adjustments, I finished the board. I ran DRC checks, ensured power traces were thicker, fixed clearances, and finally completed the PCB.

Initial Layout


Designing Back & Forth

Final Layout

Completed One

Final PCB - Pretty Clean; Proud of Myself



Shades
started Hydro Core ago
11/23/2025 4 PM - Whole Project Complete
Really Sorry! I can't keep up with this journal. My main journal has been posted in gitub.
Initial Concept Research (1.5 hours)
I began exploring the initial idea by studying various YouTubers and technical websites to understand different design approaches. I noticed that using separate controllers required excessive wiring and physical space. My first plan was to combine open-source Ethernet-to-USB and UART-to-USB schematics and merge them using a USB hub IC.
Ethernet to USB 2.0
Schematic
UART to USB-C Repo
Studying Open-Source Schematics (2 hours)
I followed open-source hardware designs, but discovered that many components were unavailable, back-stocked, or incompatible. For the Ethernet port, the magnetic-shielded connector required four differential TX/RX pairs, and I found mismatches between open-source examples and my needs. Still, the reference schematics helped a lot.

First Schematic Attempt & Failure Realization (1.5 hours)
After around five hours of work integrating open-source parts into my schematic, I realized the design would fail. The referenced example used USB-C-to-UART only, creating back-feed issues. The PHY controller I picked was too powerful and poorly matched because everything had to operate over USB 2.0 to match my mobile phone’s OTG limitations.

Learning Datasheets for the First Time (2 hours)
I began reading datasheets seriously for the first time in my life. I learned that datasheets include example circuits that were more helpful than any online answer or AI suggestion. ChatGPT was not reliable for exact electrical values. Going through LCSC, I finally discovered the correct controller IC that matched my USB 2.0 requirements.



Day 2
Restarting the Entire Schematic (5.5 hour)
I restarted my schematic from scratch. The earlier open-source designs still helped because some connections matched. Missing passive component values were found from Google searches, not ChatGPT. The UART-to-USB section was easy because the chip had a simple example diagram. The Ethernet controller was extremely difficult. It had so many complex connections that I nearly had to pull an all-nighter to finish it. I had to stay focused because if I left it halfway, I would forget the connection logic. I had to select a hub IC capable of upstreaming data reliably. FE1.1 was an option but had too many ports. I ultimately chose SL1.1 because its specs matched my project’s data flow requirements. I designed the AC input section that steps down to 5V. The example power circuits in the datasheet were extremely useful. Component sourcing was difficult because LCSC filtering takes practice. I added the upstream charging IC, letting the mobile phone charge while connected. LED indicators were integrated to show phone charging/battery status. I also built the 3V and 1V converters.

Beginning PCB Layout (4.5)
After completing the schematic, I started PCB layout work. This was the most time-consuming stage. I tested several board outlines and dimensions to fit all components properly. Autoroute failed every time. Proper component placement became critical. I had to switch to full manual routing. I learned about impedance, trace gaps, crosstalk, thermal vias, and multilayer routing through Altium Academy and PCB guru videos. I switched between 4-layer and 6-layer designs until I understood how to use bias and planes properly. After several routing revisions and placement adjustments, I finished the board. I ran DRC checks, ensured power traces were thicker, fixed clearances, and finally completed the PCB.

Initial Layout


Designing Back & Forth

Final Layout

Completed One

Final PCB - Pretty Clean; Proud of Myself



11/23/2025 5 PM - Driver Files Upload
Driver Files Uploaded in Repository for Ethernet & Uart Controller. For Ethernet it is realtek driver & for Uart controller it is silicon lab driver. You can install the driver after connecting to host device.

11/26/2025 - Revision
Differential pair Revision



100R Differential Pair Width - 0.127mm Spacing - 0.253 mm Tolerance - 0.254 mm
90R Differential Pair Width - 0.147mm Spacing - 0.253 mm Tolerance - 0.254 mm
Pick & Place file updated , Gerber Updated, PCB file updated
11/27/2025 - Removed the Via In Pad
Removed the via in pad and updated the repo with new gerber and pcb file. And still the manufacture price is same. JLCPCB Charges the standard 6 layer price.

11/29/2025 - Updated BOM
I have 2013 Excel software. That's Why It made everything unicoded. But luckily online cs editor exists and I fixed my BOM file properly. Hope this is the end!
