#112 A look at the Relatively new RISC-V IC

Since I have been using the ATtiny series of IC from microchip using the AVR instruction set I kept an eye out for similar chips.

After a few google searches I found 2 potential competitors:

Puya with their PY32F003F14P6T 32 Bit MCU 32bit ARM ® Cortex ® -M0+ Microcontroller Integrated Circuit


WCH (WinChipHead) with their CH32V003F4P6 series based on the 32-bit QingKe RISC-V2A core.

After some research I decided to purchase some of the WCH IC’s for testing. I setup a platform IO development environment and starter using the ch32v003fun open source software development stack for the CH32V003.


There is an Arduino core from the official WCH guys but it is still limited and slow for power users.

Once I finished setting up I found it very easy and intuitive to use the ch32v003fun development stack. There is a learning curve but I was able to test all my WCH chips within minutes. I purchased the CH32V003J4M6 4-pin, CH32V003A4M6 16-pin, CH32V003F4P6 20-pin, CH32V203C8T6 48-pin and the QFN20 CH32V003F4U6 20-pin.



#111 An affordable value packed multimeter

Measuring micro amps (I am not affiliated with Jacobs coffee)

While I was looking for a great value for money multimeter I came across the ZOYI ZT109 (Aneng 8009 clone??)

This multimeter is identical to the Aneng 8009 and is in many ways far superior than the older Aneng 8002 and 8008 models. I highly recommend it.

The biggest downside is the miniscule fuses but I was able to solder a standards 50x20mm fuse in a somewhat hacky way…

Also the CAT ratings are super dodgy but good enough for low voltage electronics.

My requirements were:

  • measure µA
  • auto volt measure
  • small no of turns
  • palm sized
  • large display + light
  • decent price
  • easy battery replacement
  • accurate

I can say that this multimeter checked all the boxes but there are some issues… firstly the tiny fuses are painful… I could find some replacements at DIY electronics. However once stock is finished the 3mmx5mm fuse market looks drier then a dead dingo’s donger what an epic fail.

10A fuse soldered diagonally.

Now I know soldering and hacking a different fuse is possible see my image… still it’s a royal pain in the place that’s darker then a coal miners rectum. There’s so much space for a 5mmx20mm fuse. Just why whoever added this tiny fuse must have been a real poepoltjie eh..

Another issue is true RMS but meh I don’t need it that much.

So in conclusion this multimeter is a great compact companion for any low DC voltage electronics hobbyist. I have had mine for a little over 2 years and love it.

I would recommend buying 2 if you can.


#110 Reviving an old 100A 12v lead acid battery

After the first zap the battery was around 4V…

Why do some South Africans have to scrape the bottom of the barrel reconditioning old batteries?

Opening themselves up to potential health concerns and pollution of the environment?

Maybe I can explain.

We’ve had load shedding for years now and with the recent shenanigans for almost half a decade it’s got exponentially worse with even the ultra wealthy feeling it a bit.

You’d think these clever a wealthy men would have come up with a solution by now… but it’s seems as if they have either found a way to get comfortable or they just don’t care… As long as they are making money from their diabolical cadres and corrupt hand shakers why should they care? After all they can run their water, washing machine, stove and medical equipment because they either don’t get loadshedding in their public servant mansions or they have installed million rand solar systems using taxpayers monies it’s a win-win “we fail upwards in life”.

All this while endorsing terrorist activities blatantly with no recourse or accountability using South Africa’s past to manipulate the current population into submission for the “election year” the audacity is unbelievable.. but yeah with the 30% pass rate these guys have dumbed us down and are extremely comfortable in the current climate they have designed. They really set an honourable bar to pass. I could go on but this is article is about a battery.

Seems bleak and it’s hard to ignore or be patriotic and loving towards your county and fellow citizens when there’s so much negative energy being pumped in by the guys we are supposed to trust in looking after us.. giving our data to, trusting their banks… No wonder there’s so much crime and hate.. these guys flourish in it like bacteria fueled by glucose eating a tooth. Even when the tooth is rotten the bacteria continuous to eat and will.. if not treated get into the bloodstream infecting the entire body. The bacteria doesn’t take into consideration that in future it will die along with the tooth it just consumes indiscriminately.

With that being said lets get into battery reconditioning.

Recently I got hold of an old 12v lead acid 100A battery. This battery was bought 20 years ago and stored in a corner for a rainy day.

The battery was never charged and never used.

Upon inspection the battery was at around a measly 1.95V.. this did not look good but luckily I have a DC MIG/MMA welder and decided to use the good old crude welding trick on this big boy.

I removed the MIG setup and installed the stick clamp to the + terminal and the ground to the – terminal. see the photo.

Welder settings:

I set the welder volts to about 21V and the amps to around 25A. Make sure you are using a DC welder AC will NOT work.

Make sure you do this outside or in a well ventilated room. It’s VERY important.. battery acid is no joke to organic materials.

First I did 21v at 25A for 5 minutes then let the battery rest for a whole day to observe it.

Once I concluded it seemed like it was fine my formula was 21v at 25A for 5 minutes then a 10 minute cooldown in between.

I did this 6 times and measured the battery in between times.


cycle: 19.17 V
cycle: 29.50 V
cycle: 310.97 V
cycle: 411.41 V
cycle: 511.75 V
cycle: 611.77 V

Once this was done I let the battery rest for a day.

Battery at 9V

Now comes the patience part… The battery could hold a charge but ever so slightly would drain and it was supper thirsty.

So I setup a dumb charger at 5A and let the battery charge up for a few days checking intermittently.

Next I setup my recondition charger and let it do it’s thing for a week and what do you know the recondition charger reported great values.

However the battery was still thirsty so I switched back to the 5A dumb charger and let it run for another week..

Fast forward about 3 weeks of low current and recondition charging and the battery seems to be doing fine

Holding a rock solid 12.6V and running my LED lights.

So my conclusion is that it is possible to desulfate and recondition a 100A lead acid battery that has never been used. The initial welder zapping was only the start I needed at least 3 weeks after that to “recondition” the battery to a useable state and I still don’t know the long term potential issues.

It really was just a patience game and also don’t do anything like messing around with the acid weights.

I would still like to figure out how to balance acid and water plus all the battery chemistry stuff but for now this welder trick is good enough.


About a week later the battery began acting up again seems the internal resistance is high and there is a constant draw bringing the battery voltage down.

Overall I can say that this was a temporary solution and at the moment I don’t have all the fancy battery tools or chemistry knowledge to experiment further.

Also though the battery has issues it can still be somewhat used for low voltage applications now. So I guess I’ll view this as a feature instead of a bug 🙂 cheers


#109 Types of 433Mhz RF modules in ZA

FS1000A module at 5v no attenuator just using antenna.

Recently I have been using wireless technologies for a few projects.

While looking for a balance between price, functionality and disposability I decided to focus on the 433Mhz RF modules.

These use a free spectrum and have been around for a long time. There’s is a few different types and kinds, with LORA being kind of new and better in almost every way but this comes at a high price compared with the standard 433 RF modules.

So I purchased a few receivers and transmitters from electronics suppliers located in South Africa.

All my tests consisted of running the 4 receivers at 5v and a single 17.3cm straight LAN cable strand as an antenna. The signal sent was a 23bit ASK signal with a pulse length of 1200ms.

All 3 transmitters were tested at 3.3v with a single 17.3cm straight LAN cable strand as an antenna.

The transmitters testes were the FS1000A, CYT1 and the WL102-341.

The crude module actually has more power and range at 5V but I am using them at 3.3v for super low power applications so In this case the module loses.

The Tests were done on farm land.

All transmitters could trigger the receivers at 400m line of sight but only a few could penetrate foliage and a galvanised steel shed.

I only needed MAX 400m which is why I stopped there but some sources claim up to 600m – 800m + for these superheterodyne modules. Not as good as LoRa but for the price what reason do I need not to use them?

*Sidenote Using RF or LoRa in conjunction with a 2.4G Wifi module like the ESP32 or even 3G/4G modules can create multi dimensional divers systems. where we are leveraging the long range and penetration + power output of 433Mhz and 868Mhz but also allowing packets of data to connect over the internet to be stored on a server for data analysis and the creation of graphs to make the data more visually appealing.

Currently I do have some pilot devices and hope to one day make some good quality sensors in 3 different tiers:

  1. Cheap and disposable sensors
  2. Affordable long term sensors
  3. High end sensors

These will be focused on use within rural outdoor areas and I will have a version with Gerber files and schematics etc. available for anyone to download and make for themselves. However the more refined version with a nice enclosure and style will be sold commercially since I do want to be paid for my work.

Back to the modules..

The transmitters that support 5v could penetrate a little better sometimes.

The position of the transmitter/receiver could also greatly affect the received signal especially at range.

Also during summer and during rain the signal was worse with the foliage and water most likely absorbing and/or reflecting the signal

All receivers were superheterodyne with a crystal and I did not use any counterpoise though it would help in some circumstances it makes the receiver unpractical and large.

From worst to best

Some people may wonder why I am using these modules instead of the fashionable LoRa modules. This is simply due to cost and availability.

Designing a good circuit cost time and money. Inserting said circuit into an extremely hostile environment like for example.. rural South Africa is an even more costly exercise

I have had devices damaged by the sun, damaged by water, damaged by ants, damaged by cows, damaged by some kind of rabid animal (assuming jackal) The list goes on.

AND I have not even mentioned the human element… devices damaged by criminals some even STOLEN… for what? You telling me that criminal is sitting in the bush conspiring to reverse engineer my simple circuit and RF protocol and some how will be able to defeat Microchips code protection? I highly doubt it but it is possible…

So now I hope you can understand why these cheap modules do work and are very useful + inexpensive for my purposes.

I also have LoRa versions but for now I only use those when distance and extreme sensitivity is needed.


#108 Low current standby fix for IP5306 MH-CD42

BC547B NPN Transistor
IP5306 Module

Over the past few years I have been using the IP5306 chip and specifically the module shown in the image above.

The module is a great all in one solution for LiPo battery powered projects: charge, discharge, protection, 5v step up etc.

That being said there is 1 massively annoying caveat:

If the load current drops below 45mA during 32 seconds, the IP5306 will go into standby mode…

For low power battery operations this is simply unacceptable.. and I will not simply increase the current draw to keep it on.

There is an I2C version which allows us to change a few settings like standby mode in the IP5306 but for this fix I will focus on the “dumb version”.


An easy solution is to create a simple heartbeat circuit.

Since there is a button which will prevent the IC from going into standby mode, if pressed it will reset the *32 seconds 45mA* timer.

The module I have also has a solder pad where I can easily solder a wire to control this button via an MCU.

Using a *BC457 NPN* transistor we can create a simple switch to “press the button” at least once within the 32 seconds within a loop.

In this way we can constantly keep the module powered.


– NPN transistor (I used the BC547B)

– resistor (1k is fine)

– hookup wires


The Base connects to the resistor and then your MCU pin of choice.

Emitter gets connected to GND.

Collector gets solddered to the button pad.


Once everything is soldered and double checked you can then add the code for the heartbeat.

In this case I use the millis() function and a simple repeating timer all written in a sketch .ino


#107 Identifying fake IP5306 MH-CD42 modules

Over the past 4 years I have been purchasing an all in one LiPo charge, protect and step up module for my LiPo battery projects.

This module has worked quite well but I have noticed some fake modules starting to creep in the market again…

Now I can live with fake chips that work close to spec.. but in this case the IC would power on once.. and then die completely.

The Module uses the IP5306 all in one power bank IC. There is a version which uses the MH-CD42 IC but I have always received a module containing the IP5306.

Therefore I will focus on this IC.

The module charges via 5v and steps up a non protected 3.7V battery to 5v, it also offers protection to the battery.

Over current protection (OCP), over-voltage protection (OVP) short circuit protection (SCP) and over temperature protection (OTP)

2.1A of current can be supplied which is a great reservoir for DIY projects and sometimes even an overkill.

After stating the most attractive traits above you can see why this module is much loved.

Fake on the left with a few distinctions on the IC.

Unfortunately once a module becomes extremely successful fakes start appearing out of the wood work trying to steal some glory.

And as usual the consumer suffers the brunt of the con job..

Thankfully in my case I only came across 2 modules that were fake and I was able to alert my local supplier.

Hopefully they will do something about it… and in case anyone has a similar issue I will do a small breakdown of the tattle tail signs these con boards display.

So you too will be able to identify and maybe save yourself some annoyance and time but most importantly save yourself some money.

Now I was able to compare 2 modules.

The fake has external circuitry which works and is laid out the same way as the original.

The fake IP5306 IC is the culprit here.

Soldering an original to the fake module actually can bring the module back to life again!

So if you are able to get a few working IP5306 chips you may be able to get your modules working again.

The tattle tail singes are:

Fake: text on PCB is faded

Fake: text on inductor is faded

Fake: text on IC looks elongated

Fake: text on IC is also slightly faded

Fake: The pin one identification is a small circular flat indent (original has a smaller concave ident like a ball)

Fake: the Infineon logo is close to the centre left of the IC and the thickness of the logo is very thin (original has a thick logo and is located upper left on the IC)

Fake: the inductor is completely flat (original has an ident all around the edge of the inductor)

The face capacitors, resistors, LED’s, button and inductor seem to be the correct values but I cannot speak to the quality of them

My theory is that the chip has a dye which is far inferior to that of the original and thus it failed.

It could also be a complete fake in some cases.

Faded fake on the right.


#106 AJ-SR04M ultrasonic distance sensor for water

looks like a STM8S003F3 MCU, unmarked crystal and unmarked TTL IC
Underside has a 2 pin socket and is quite dirty…

Recently I have had an old mildly annoying problem snowball into a new serious problem…

Every few months the clean water supply from uThukela Water (Pty) Ltd has been switched off for multiple reasons… striking, damaged electric motors due to Eskom, sabotage and other issues to name a few very serious reasons.

So two large 2500L water tanks were installed in series as a backup which worked well for small water issues that would last maybe a week or two.

However recently There has been no water from uThukela for over a month, and this is very serious.

This event triggered me to investigate water related problems and solutions specifically for my use case.

Order of importance:

  1. I need readily available clean drinking water
  2. Store this water for longer (get extra tanks)
  3. Keep water safely in the tank (no contaminates)
  4. Add sensors to monitor (water level sensor in this case)

For this article will be focusing on the 4th order of importance since this is a tutorial website mainly about electronics.

Therefore I will start by saying I searched for a suitable water level sensor and came across the JSN-SR04T and clones.

This sensor looks very promising and easy to use with 6 available sensor modes (adding increased diversity).

N.B the copy does not have 6 extra modes which was disappointing considering their price point…

2.2m wire with the sensor at the end.

The copy has 3 modes and is similar to the JSN-SR04T-2.0

Now my goal is to use the JSN-SR04T with an ESP8266 connected via WiFi to send readings to my server every 30s, this unit will be completely powered by solar.

The ESP8266 will also have a LAN dashboard to view the readings in real time connected to WiFi but with a connection to the internet not needed, just in case the internet goes down I can still read the water level values.

unfortunately finding a commonly available original JSN-SR04T Ultrasonic Distance Sensor has been quite difficult in South Africa.

I have only been able to find the AJ-SR04M (functions like the JSN-SR04T-2.0) which is a clone but works just like the original, however I see the price is equivalent and sometimes even more than the original which is quite strange. An of course the extra modes are missing…

The waterproof sensor
The sensor is epoxied and completely sealed looks easy enough to install

Mode 1: R27 = is open.

The sensor returns an analogue signal. The formula to calculate the distance from the data is:

Test distance = (high time * speed of sound (340M / s)) / 2;

Mode 2: R27 = A 47K resistor is soldered.

Every 100ms serial data will be sent in mm.

Serial baud rate: 9600, n, 8,1.

The frame format is: 0XFF + H_DATA + L_DATA + SUM
1.0XFF: for a frame to start the data, used to judge;
2.H_DATA: the upper 8 bits of the distance data;
3.L_DATA: the lower 8 bits of the distance data;
4.SUM: data and, for the effect of its 0XFF + H_DATA + L_DATA = SUM (only low 8)

Mode 3: R27 = A 120K resistor is soldered.

Good for low power applications.

After the module is powered on, the module enters standby mode.

If the module receives 0X55 it will send data over serial.

Serial baud rate: 9600, n, 8,1.

Datasheet for the stm8s003f3


#105 Custom integration sensors with custom receiver

V1.0 breadboard prototype with DIY EEPROM module
V1.0 stripboard soldered prototype with USB and Lipo battery
Testing 2x custom sensors (1x ATTINY85 and 1x ATTINY412) with 433 RF modules

Recently I wanted to integrate the RoboGuard system with some custom sensors on my farming property.

This motivated me to study the hardware and RF protocols used by the RoboGuard

I would like to also account for multiple RoboGuard transmitters scattered over the property each RoboGuard device has 2x pir sensors and sends an alarm signal once both are triggered.

They also send a heartbeat ping every 15min.

They have a range of roughly 400m from transmitter RoboGuard to receiver HQ.

Testing EEPROM data storage.

Now the RoboGuard system uses 433.92Mhz to send signals to the HQ however the HQ can only add up to 8 paired RoboGuards.

Once you reach this limit you will need to purchase more RoboGuard units.

For example if you had 12 RoboGuards, 2 HQ units would be required but if you wanted an HQ that can store more than 8 you would be out of luck.

luckily I had made my own custom RoboGuard receiver and was able to add my own DIY sensors to the RoboGuard device ecosystem

The protocol used is 433.92 ASK and each RoboGuard has 3 signals

  • alarm
  • tamper/learn
  • heartbeat ping
Testing penetration behind galvanised shed (using CY33 module)

Now my receiver needs to store the received device learn UID and this is done via EEPROM on my board

Now my custom device receives all signals just like the RoboGuard HQ.

Next is communicating with the TAK Server.

I could swap the 328P for an ESP8266 which allows WiFi connectivity to the internet

This then allows the device to connect wirelessly.

It still receives RF data from the RoboGuards and just ports these signals over the internet

In future I will make a device with an integrated WiFi connection but In this case all I wanted was more zones and an affordable extra device to keep in my laboratory permanently with the capability to receive 433mhz signals walking around the premises. If need be

Overall my unit contains

A speaker
6 push buttons
2000mAH Lipo battery
built in charger
ability to add clients 12 RoboGuards (more depending on EEPROM size)
433 MHz superheterodyne receiver only
logic to handle all these features

Front of the 433 Transmitter
Back of the 433 Transmitter

More info + datasheets and schematics etc. on my GitHub here


#104 Reverse engineering an old linear power supply

Back panel connectors Antronics made by TPW

Recently I came into possession of two working ups devices from 1993. both of them had old capacitors and old 12v7a lead acid batteries inside the devices.

First thing I did was clean the cases and the PCB boards. Once that was done I replaced the old capacitors and the 12v7a batteries, then I tested both devices. both work fine but the design is old and a bit dangerous.

Secondary side
Primary side

So I decided to reverse engineer the circuit in order to better understand the design and to see if I could make any improvements to a design I would like to make.

While reversing the PCB I noticed that the mains earth and the GND of the circuit were connected together. I also notices sone discoloration from what looks like heat between the regulator and the transformer. Also the 330 ohm resistor for the led appeared to be discoloured from what also looks like excessive heat.

Mirrored for reversing
The original schematic I reversed
Schematic after I implemented suggestions

With these issues in mind I also noticed that the heatsink for the LM317T was very small and close to the transformer and the mains 1A fuse was placed after the choke and varistors instead of before them.

In conclusion I decided to choose between a different regulator at a fixed voltage or a chain of 4 LM317Ts providing around 6A of peak current, Since I do not need to adjust my voltage like the original circuit I should be able to get 13.75v by using a fixed 1k and 10k resistor. I also wanted better heat dissipation and Amps so I will definitely install good heatsinks with thermal compound. Depending on the size of the enclosure I get for the project I may add a fan.

The project files and components list etc. can be found on my Github here.


#103 Dangerous BK-357 USB charger teardown

Listing on Takealot
Claimed output…

While looking for a new multi port USB charger I came across the Model BK-357 sold on Takealot by OQ Trading. This charger had many positive reviews with a 4.1 star rating at the time of writing this article and for the low price of R149.00 I had to give this device a try.

Caps could potentially be repurposed from e-waste…

Once I received my charger I noticed That the fast charge USB port was working flawlessly but the 3 normal charging USB ports seemed to have current divided between them.

So I decided to open the charger to investigate further. I noticed the bottom part was glued into place and could be pried open carefully with a small screwdriver and spudger.

Primary and secondary sides with a thin line in-between…

Once opened the PCB was in good condition and contains 2 small switching transformers and a nice fusible resistor that also acts as an inrush limiter but that’s where to positives end. The interference capacitor was skimped on also the electrolytic capacitors are all different colours and brands It’s possible that they have been taken off old junk and re-purposed which is okey but they may have a diminished quality which is almost as alarming as the gap between the primary and secondary sides of the transformers. The biggest gap is around 4.8mm which is not to bad but right in the middle the gap closes all the way to 1.3mm!!!

Largest gap is around 4.8mm and the smallest is a whopping 1.3mm!!

This is very dangerous since there is 1.3mm of PCB space between you and mains voltage!!!

I will be posting my findings as a review on Takealot.

When purchasing multiple socketed USB chargers go for the larger more expensive ones.. As you can see in this case the tiny transformers just can’t output enough current on the cheapies.

The article from Europe’s Safety Gate Alerts can be found here. They identified the problems and measured the charger all the way back in December 2021 in Ireland and here in SA we are happily selling these.

I have also included an article from a Russian review back in March 2021. where The charger was also analysed and determined to be a potential hazard the article can be found here.

1x fast charge 3x normal charge with current divided by 3

Passionate about technology!