Help figuring out D3806 DC-DC Buck-Boost UART Commands

[ric96]

Hi All,

I just bought the MingHe D3806 for experimenting in my solar shed, so nothing mission critical. I was intrigued by the presence of what looks like a UART Debug header, which made me go down the rabbit hole and finally end up here.

Since I last posted in 2015, a lot has changed so here's a bit of an update on where my level of understanding lies, I work at a firmware level, so C (and friends) and digital circuits are a strong point, as for the analogue/power electronics side of thing I'm a noob who can google stuff and get a generalized idea of whats happening.

About the D3806:
Its a DC-DC Boost-Buck SMPS, and it does that in Boost first and Buck order. As shown in this video by Julian:



Also, seems to be very noisy as per many reviews.

Relation to the B3603:
Both the units have identical control boards (see attachment for images of the D3806). However functionality is very different.
I'd like to make use of the UART header to control it using a bigger micro for various tricks and automation purposes, however I can't seem to get ANY response form it.
I've tried the following protocols:
1: https://benjames.io/2018/06/29/secret-uart-on-chinese-dcdc-converters/
2: https://www.droking.com/cs/support/topic/200220-dc-dc-buck-converter-uart/
Also tried them at various bauds, attached a Logic analyzer just in case and the Tx line was silent the whole time.

Which brings me to the possibility that uart is completely disabled, or more likely uses a completely different protocol.

The finally the questions that keep me awake are:
1. Does anyone know the D3806's protocol ?
2. Can the same alternative firmware from B3603 be used here and all the Buck-ing and Boosting control is done passively in the bottom board?
I guess we can diff stock firmware image from both the units, and to that, I ask if anyone has tried firmware dl/ul using something other than a stlink? nvm. i have one.

Closest thing to a datasheet (unverified): https://easyeda.com/subbbash/D3806-a3308e2429d04687bd07515b512cb7e9

[ric96]

I'll probe the control board pin-out tomorrow and see if it like up with https://github.com/baruch/b3603

Anyone can get me a original firmware dump of B3603 for a comparison?

[geo999]

having the same problem I tried to do a dump of the STM8 and see if there are any interesting strings (uart commands) but,
unfortunately, the device is protected.

Anyone had any luck in controlling this using UART commands ? , custom firmware maybe ?

[geo999]

I did a little reverse engineering of the motherboard and it's very similar to d3603.
In addition to that it has a boost converter that brings the voltage up to 38V before XL4016 brings it down to your specified value.
The only difference I found on the control side was the missing on/off (pin 6), it's generated by the daughter card but not connected on the motherboard.
Besides this the d3603 open source firmware should work with the mention that some adjustment might be needed in the measurement side due to few different resistors values.
what I do not like on the d3806 is this boost-buck configuration that leads to +38V at the XL4016 input that in turn leads to a lot of heat on the XL4016 even for a small 12V 5W load. Not sure if this is the intended behavior or it was damaged in the process ... anyways I removed the NCE8580 mosfet and got rid of the boost feature.

[rin67630]

I did not waste my time in trying to reprogram the control board, i replaced it with an ESP32 TTGO.
Please take a look at my github dev for a couple of DC-DC converters of the Juntek-Drok family:
https://github.com/rin67630/Drok-Juntek-on-steroids