Here is a story of one cheap HDMI digitizing device. I have wanted to connect devices with HDMI output Open Broadcast Studio video mixer/streaming software running on PC. Usually HDMI to USB adapters are expensive boxes (around $100 or more). But the I saw this really cheap (€16,15 / US$ 18.99) Mini Portable HD USB 2.0 Port HD 1080P 60fps Monitor Video Capture Card For PC device that looked like worth a try.

The specifications looked promising for the price:

I received it like this. Driver CD broken. It did not work without it – as plugging it to Windows 10 PC detected it as some video device, but trying to really use it led to blue screen of death crash. So it was not clearly plug&play.

The seller company gave money back and they did not want the trash back.

So I decided to take a look inside because I was curious what is inside this device.

Component markings on chip were sanded out on chips. The ended my first first step. The device was for some weeks on my misc electronics trash bin waiting if it was of any use for some application (like could drivers be found for download that make it work or something else).

Later just for testing decided to test Mini Portable HD USB 2.0 Port HD 1080P 60fps Monitor Video Capture Card For PC device on my older computer where I had earlier used this analogue video to USB converter device.

I was amazed that Mini Portable HD USB 2.0 Port HD 1080P 60fps Monitor Video Capture Card For PC device was recognized EasyCAP analogue video to USB converter. Why would this device look like analogue video interface when the video coming from HDMI connector is digital video signal. Weird.

I got video on same resolutions and quality as I did with analogue digitizer – not at 1080i HD quality I fed to HDMI input. The recorded video even seemed to have same type of noise and interference on the signal as analogue composite video signal has. Clearly this device was will not record in HD quality. Even if I tried adjusting settings, I could not get anything close to HD quality recording from this (capture maxes out at 720×576). So I consider it as bad HD capture card (at least without the original drivers if they are in any way better).

Based on the results I start to suspect that this device could internally somehow convert first the signal to analogue composhite video that is fed to analogue video to a cheap USB converter. Then some Google searches started to reveal that my what I suspect could be right:

1080p 60 FPS HDMI capture card for only $15? | UTV 007 HDMI USB Review concludes that some things ARE too good to be true. The capture maxes out at 720×576, interlaced 25 FPS, wrong aspect ratio, overscan crops out parts of capture, digital-analog-digital conversion means poor quality.

Mini Portable HD USB 2.0 Port HDMI 1080P 60fps How to install and capture using VirtualDub review video says that while the image quality might not be best, but it is only $15.00. The mentioned problems are limited resolution, it makes picture interlaced and flashes sometimes. Can be used with 720p or 1080p source when all you need is lower resolution video. Recommendation is to give a little more money and get uncompromising quality with capture devices around $100.

HD Capture – USB 2.0 HDMI capture card REVIEW & TEARDOWN video investigation reveals the dirty details I was suspecting that indeed this device seem to be converting first HDMI to analogue video and then fed the video signal to UTV 007 analoge video to USB chip. This reviewer thinks on video description “This thing should not exist”.

So it seems that using) Mini Portable HD USB 2.0 Port HD 1080P 60fps Monitor Video Capture Card For PC device is approximately same in quality than using cheap analogue video to USB converter connected to  HDMI to composite video downscaler. The result is quite poor SD resolution analogue TV video signal quality video. If you want to produce 640×480 or so resolution low quality video stream as cheaply as possible, then this might be an option. If you want any decent HD quality results, then it is maybe a better idea spend more of your money on some better HDMI to USB converter or capture card (whatever that might be).

Biltema sold this ground fault interrupter (RCD / GFCI) device few years ago (does not seem to sell anymore). It is designed to be connected to 230V AC 50 Hz electrical outlet.

The information on the back of the device tells that this will trip at 30 mA leakage current. The symbol tells that this device will trip with AC leakage and pulsating DC leakage. The device is designed to be able to handle up to 16A current.

This device was designed to that you have to press the green “RESET” always when you want to turn it on after you have connected it to mains power. The device will always reset to tripped/OFF position if mains power is cut out. Maybe it was designed to work in this so that it would also protect the user against fault if the neutral wire (one of the most cumbersome bugs that other RCD types do not always protect).

Let’s look inside

Quite normal here. Removing the button revealed some loose plastic parts dropping out. This tells why it did not work mechanically as it should. It seems that the main failure mode was a mechanical failure that prevented the device to be properly turned on so it was discarded as damaged device.

Here black thing with hole on the left the current transformer used to measure the current difference between incoming live and neutral wires. The third thin wire implements the “test” button simulated fault current. The electronics consists mainly of one 8 pin IC (with markings sanded off) and two components that look like transistors (could be transistors or thyristor). The big resistors and diode on the right seem to be as part of the transformer-less low voltage power supply used to generate the low operating voltage to the electronics.

Looking at the contacts closely reveal that there has been some serious sparking on the current carrying contacts. Maybe there has been a serious short circuit or big power turn-on surges (for example switch mode power supply with large startup surge connected to output).

Some of the sparks have been so strong that they have made their mark to the case as well.

I have had this cheap USB endoscope for few years. Then it suddenly stopped working.

So I opened it up to see what is inside and hopefully fix to work again.

The problems seems to be loose green wire. A bit of soldering and it works again.

Not much electronics. Just wiring from USB connector to camera cable and one potentiometer to adjust camera LED light brightness (black wire on cable).

Everything works again. Computer recognizes it as USB 2.0 camera. It also works with Android phone using USBC OTG adapter and CameraFi app.

I had a Huawei MediaPad T3 10″ (AGS-L09) tablet with broken glass. The repair at the official repair center costs more than buying a new tablet (both around 150 Euros), so repairing does not make much sense.

How about a DIY repair for broken screen? It seems to be doable according to several repair videos.
Huawei MediaPad 7 Youth Touchscreen Repair/Disassembly

HUAWEI MediaPad 10

The display was attached to plastic display panel frame with dual side tape. It took some work to remove it

The flat cable to screen can be removed by opening the connector “hinge” that hold the cable contacts in place. The flat cable was also glued to the back of the screen with tape like glue.

In this tablet the display was glued to glued to digitizer glass on this tablet model. So I can’t reasonably change just the digitizer. So I need to get the whole display+digitizer module. Found right new display+digitizer from ebay:

New display in

I needed also 2mm wide dual sided tape to re-attack display to plastic holding frame

Now rebuild the tablet

Everything works again.

For controlling stepper motors you need a suitable controller. It is possible to build controller using discrete components, controller ICs and using micro-controller. It can be fun to build our own stepper controller, but when you have done that few times, you might want to start looking or reasonably priced ready made easy to use modules for controlling stepper motors. I found TB6600 4A CNC 32 Segments 42V 42/57/86 Stepper Motor Driver Controller to be promising looking and reasonably priced (8.44  Euros).
Features:

Bipolar constant current chopper
High-speed optocoupler signal isolation input
Electrical input signal is TTL compatible
Stationary current is automatically reduced by half
Suitable for any small-and-medium automatic equipment with CNC controller, such as X-Y-Z tables, labeling machines, laser cutters, engraving machines, and pick-place devices
Specifications:

1. Supply voltage: 9~42VDC
2. Input voltage: DC 12 ~ 48V
3. Operating temperature: -10 to 45 ?C
4. Storage temperature: -40 to 70 ?C
5. Max drive current: 4A
6. Chopping frequency: 20KHZ

Notes:
Due to the different Production batch, some item will be TB67S109AFTG

This stepper motor worked as promised with small stepper motor I had (1A per phase current).
Driver did not generate much heat.
Worked well on whole voltage range and current limited motor current as promised

There are two indicator LEDS:

RED is turned on when the device is powered on.

GREEN is for step pulses: when no signal it is turned on, when pulse signal is on the led turns off.

Inputs are optoisolated.The signal inputs were opto-isolated and take around 17 mA current in operation (worked in voltage range 2-5V well).

EN input does not need to be connected to anything as it is turned on by default – if you feed 5V to it with turn stepper controller off.

PUL is pulse input – indicated with green LED (when signal active, LED is off)

DIR controls the direction

The picture on the product page seems to be showing that the case is made of metal, but the version I received the heat-sink part was metal, but the protective case around the circuit board was made of plastic.

My device with metal heat sink and plastic case top worked well.

Features of TB6600

• Support 8 kinds of current control
• Support 7 kinds of micro steps adjustable
• The interface adopts high-speed optocoupler isolation
• Automatic semi-flow to reduce heat
• Large area heat sink
• Anti-high-frequency interference ability
• Input anti-reverse protection
• Overheat, over current and short circuit protection

Look what is inside

B6600 4A CNC 32 Segments 42V 42/57/86 Stepper Motor Driver Controller worked well on my quick tests. It seems to be worth to consider it as stepper motor controller. There are also similar other controllers with pretty similar specifications, for example https://www.banggood.com/TB6600-4A-CNC-32-Segments-42V-425786-Stepper-Motor-Driver-Controller-p-1093437.html?p=27131452996820140438 and https://www.banggood.com/TB6600-Upgraded-Version-32-Segments-4A-42V-5786-Stepper-Motor-Driver-p-981997.html?p=27131452996820140438

There are also stepper motor controllers with same controller IC but in different form factor, for example https://www.banggood.com/TB6600-Upgraded-Version-32-Segments-4A-42V-5786-Stepper-Motor-Driver-p-981997.html?p=27131452996820140438

This is a teardown of Behringer B208D active PA speaker.

Inside

Here is input pre-amplifier

Look inside power electronics metal box

Inside here is swich mode power supply (mains to around +-27V) and amplifier board.

Amplifier board has 160W class D amplifier chip (below board and possibly broken) for driving low frequency speaker element, 40W power amplifier IC for high frequencies and power regulator for +-15V preamp power.

The wiring between amplifier and preamp carries ground, +-15v, unbalanced high and low frequency audio signal plus status pin indicating when power amplifier is overloaded.

This is a teardown of one broken Ethernet to 3G cellular network industrial gateway.

There is a miniPCI slots for communications module. It is a commonly used for cellular cards. MiniPCI slot provides both PCI and USB signals. This Option GTM382 cellular module has USB 2.0 host interface (uses AT commands).

There there seems to be main processor and Ethernet controller on the center. Memory is at bottom left corner.

On the backside of circuit board you can see SIM slot, one jumper wire, few ICs and some generic discretes.

Here is teardown of a broken Samsung J5 smart phone (SM-J510FN from 2016).

Let’s open it:

A

Closeups of circuit board

Not too much to see. Many parts were covered with metal cases soldered to circuit board. I did not see here any easy to fix problems (like loose connector) that would have allowed to cheaply fix the phone.

Broken display and almost dead battery maybe some other potential issues… not worth to fix this quite cheap two years old phone.

I ordered this cheap (around $10-15) 5.1 channel USB sound card from China few years ago. The device does not have markings on the box so hard to day which exact model it it was. It seems that there are several similar looking cards available at dx.com and banggood.com even today.

CM6206-LX based USB Sound Card (Cheap Chinese 5.1 + Optical in / out)

My sound card got damaged so that miniUSB connector was not where it was supposed to be – it was loose somewhere inside the case.

Let’s open the device to see what is inside and if it can be repaired easily.

Pretty much every funtionality seems to be made with this chip.

The main chip is Cmedia CM6206. Datasheet says that CM6206 is a highly integrated single chip USB audio solution. All essential analog modules are embedded in CM6206, including 8CH DAC and earphone buffer, 2CH ADC, microphone gain, PLL, regulator, and USB transceiver. Capacitors on audio inputs and outputs are all 100 uF 10V electrolytic capacitors.

CM6206 features:

USB spec. 2.0 full speed compliant
USB audio device class spec. 1.0 and USB HID class spec. 1.1 compliant
IEC60958 spec. compliant (consumer format S/PDIF input and output with loop-back support)
SCMS (Serial Copy Management System) compliant
Dolby® digital audio streaming via S/PDIF out
USB remote wake-up support
8 channel DAC output with 16 bit resolution: 3.1 Vpp (1.1 Vrms) biased at 2.25V output swing, volume control and mute function, earphone buffer
2X interpolator for digital playback data to improve quality
2 channel ADC input with 16 bit resolution: 3.2 Vpp (or 4.0 Vpp programmed by vendor driver) biased at 2.25V input swing, volume control and mute function
Additional headphone output with selectable source and phone jack sense

Then to repair task. The USB connector seems to have been ripped off from circuit boards pretty nicely without damage to board itself. For repair task this is very good news.

I suspect that the original soldering was somewhat weak because it broke down like this before tracks on board were damaged. My quess is cold solder joint – common manufacturing quality issue I have seen (bigger components like connector need more heat than tiny SMD parts).

Some soldering under microscope fixed the board.

The sound card worked nicely after resoldering. Hopefully it will last better with proper handmade soldering work.

Here is tear-down of cheap USB 2.0 to CAT5E CAT6E RJ45 LAN Modem Extension Extender Adaptor for Webcam XBOX adapter. This adaptor is designed to allow your USB cameras, printers, webcams, keyboard, or any other USB device to be situated some longer than normal USB cable (few meters) distance away from PC. This converts USB 2.o signal to format that can be transported over UTP wiring with RJ-45 connectors (Ethernet cabling).

The extension adaptor product page claims that you can connect your USB device to your PC at a distance of up to 100ft with the help of a high quality cat 5/cat 6 patch cable. On my tests it worked well with 10 meters of CAT5 UTP. I have not yet had tried with longer cable.

Let’s open the device. It just takes opening four screws to see what is inside.

There was not much to see in this or another one. Just one “magic” chip blob, three capacitors and one resistor. I think this chip could be some USB signal amplifier / repeater.

The board marked with D was the receiving end and one with U the signal transmitter.

Not much here.

I figured out this pin-out for signals at RJ-45 connector:

1 USB signal

2 USB signal

3-6 Ground

7-8 +5V power

Related thoughts:

I think this circuit with active electronics will be better than simply wiring USB signals to RJ-45 connectors as shown in the following How-To: Make a pair of USB over Ethernet adaptors video. I assume that the active electronics make the signal to go more reliably over the longer distances, handle impedance mismatching (USB 90 ohms vs 100 ohms for UTP) and maybe protects USB ports better.

I wrote around two years ago a blog posting on RJ-45 Ethernet surge protection. Now it is time to check out what is inside one cheap Ethernet surge protector. The product here is very cheap (around $2) CY UT-013 Lightning Protection RJ45 UTP STP CAT6 CAT5e Female to Female Network Lan Adapter Extender. The specifications promised on the product page are:

RJ45 UTP STP CAT6 CAT5e Female to Female Network Lan Adapter Extender
With Lightning Protection & Shield
Used to connect two RJ45 cables together
Make two RJ45 cables work as one cable
Straight through configuration
8 Conductors w/ Gold-plated contacts

This seems to be a product that would work for general Ethernet over-voltage protection. The general construction looks OK:

Let’s take a look inside. It was pretty easy to open with some force because the case just seems to be hot-glued together.

There seems to be eight small signal protection components components here. Those protective components are wired all between signal wire and the ground that is on the RJ-45 connector shield. This kind of protection device needs to be wired to a well grounded shielded (STP) network wiring to work best. If it is connected to UTP wiring, the protection level it provides is very much limited. There is no connection for any separate ground.

The general construction of the device looks OK. The wiring on board looks decent as different pairs are nicely routed as pairs on the board. If the capacitance is low and the operation voltage of those protection devices this construction looks basically OK.

The question remains what type of components they are and at what voltage they operate. They do not seem to have any readable markings.

I did few measurements:

The capacitance on those protection devices looks very low. I had hard time to do reliable measurement of them while they are in the circuit (and decided not to remove them), so I don’t have very accurate numbers here. The capacitance looks to be in order of few pF or so. I would say that very low and I don’t thinks would have serious effect on Ethernet signals.

Then I tried to measure the voltage they operate. For that I used an insulation resistance meter (250/500/1000V test voltage ranges) and a multimeter.When I did insulation resistance test 250V DC I got open circuit reading. It’s good that they can handle some DC, so they do not start to conduct at voltage used by for example Power over Ethernet (48-56V DC). Ethernet cards are designed to handle more than 250V voltage (typically designed for 1500V isolation level).

When I did the insulation resistance test at 500D DC voltage, I started to see some flashing on the protection component. When I measured the voltage that is left over the protection device when insulation resistance tester tries to feed 500 V DC at quite low current, I got reading that was at around 150V DC.

The end result seems to be that protection component seems to be miniature gas discharge tube from pin to shield. This type of protection components have typically very small capacitance, quite fast operation and can handle reasonable surges well. This Alibaba page seems to have somewhat similar looking (ruilon RLM501-141M) 500A miniature gas discharge tubes (8/20us,1time) for sale – they promise very small electrostatic capacitance(<0.8pF) and great isolation(>100MΩ) and available at several voltages from 140V to 700V.

This was a quick look what is inside. To really see how effective those are more tests would be needed to be performed with suitable surge voltage source.

I did earlier a teardown of broken Behringer B208D active PA speaker. Now it was time to get ot fixed. The problem was that sound level before distortion was very low and bass was missing.

It seems that low frequency driver could output only 1-2V signal to speaker before distortion, and to get that the volume control needs to be turned to very high gain.

I pin-pointed the problem to be at power amplifier board between low frequency audio input and speaker element output. There was one loose ferrite on one coil and something else that was hard to figure out without circuit diagram.

I found finally (B210D bigger brother uses same electronics) circuit diagram at https://www.scribd.com/document/386893914/B210D-pdf

It was a bit of low quality on preview, but enought to do fault-finding. The culpit turned to be one failed 1 uF 50V bipolar electrolytic capacitor on the way from audio input to class D amplifier IC. The capacitor should have 1 uF capacitance, but had only few nanofarads. Replacing it with new bipolar electrolytic fixed the speaker.

I would have wanted to use better capacitor type (longer life and hopefully slightly better sound), but there was not enough space for putting easily physically bigger capacitor there.

Here is the closeup of amplifier board (first pic has one extra test capacitor soldered in parallel with broken capacitor):

Other side view (failed capacitor removed):

For more pictures of speaker internals take a look at my teardown of Behringer B208D active PA speaker.

Android smartphones and PC use TRRS headset connectors. You can find details on that interface on my Telephone handset to smart phone and laptop posting. Ealier PCs used two TRS plugs (mic and headphones) for headsets. Because both connection types are still in use, sometimes adapters are needed.

Here are some details of those adapters.

The first suspect is  not very good types https://www.banggood.com/3_5mm-1-Male-To-2-Dual-Female-Earphone-Microphone-Splitter-Audio-Cable-Adapter-For-iphone-Samsung-p-1169141.html?p=27131452996820140438
https://www.banggood.com/3_5mm-Stereo-Audio-Male-to-Earphone-Headset-Microphone-Adapter-PC-Cell-Phone-p-1011309.html?p=27131452996820140438

The problem with those has been that 3.5 mm jacks are so close to each other that many 3.5 mm jacks do not fit in well.  Other issue is that the 3.5 mm jacks and the cable are not very durable. The connectors became loose quite quickly and cable failed on one of the unit. I also had some compatibility problems with some headsets (not detected correctly by smart phone and/or no mic sound).

Second better one is this 3.5MM 1 Male To 2 Dual Female Earphone Microphone Y Splitter Audio Cable Adapter
https://www.banggood.com/1-Male-To-2-Dual-Female-Earphone-Microphone-Y-Splitter-Audio-Cable-Adapter-p-1029049.html?p=27131452996820140438

End

Here is a collection of interesting videos on microphones and microphone technologies:

Audio 101: Microphones

Dynamic vs Condenser Microphones, What’s the Difference?

Audio recording tutorial: The different microphone types | lynda.com

XLR vs USB Microphones, Which Should You Buy?

$22 MICROPHONE VS $3600 MICROPHONE | Andrew Huang

Voice-over Microphone || DIY or Buy

How a Neumann U87 microphone is manufactured

How to Make Your Own DIY U87 Vintage Condenser Mic

Soyuz Video #3 Capsules

The Soyuz SU-019 FET Microphone

MKH-416 counterfeit mic teardown

https://www.youtube.com/watch?v=Vdoeh7y0yeE

Disassembly – Taking apart the BM-100FX Microphone.

BEST Budget USB Microphone!? | FIFINE Honest Review

Why the Blue Yeti Sucks

Why the Blue Yeti Su cks | Pro Audio Engineer Responds

Do XLR to USB Cables Work?

Samson USB GO Mic Teardown

BM-800 Microphone Review / Teardown

Teardown of the Blue Yeti Microphone

Samson c01u usb microphone test + look inside

A quick teardown on the Samson Go Mic USB studio microphone for laptops and tablets

Samson C01U USB Mic Teardown (part of the article)

Random Teardown #001: Blue Snowball Ice Microphone

Here is tear-down of PT 2X2-24DC-ST
over voltage protector for 24V systems: PT protective connector with protective circuit for two 2-wire floating signal circuits. 24 V DC nominal voltage. HART-compatible.

What is insideComponents close-upWhat was found was two dual gas discharge tubes ant two other over voltage protectors. The gas discharge tubes were connected to pins at one end (two protectors for two signal wire pairs).The contacts on other and go to flexible circuit board that has just those surgectors/zeners in it. Look like a bit of complicated electro-mechanical construction because it looks like just directly soldering the components to pins on bottom had done the same.

Here is teardown of a ALADDIN HASP software copy protection module that was designed to be plugged to PC parallel port. Because modern PCs no longer have parallel ports, this is obsolete technology ready for teardown.

Not much to be, just few custom or custom marked ICs (ALHY1 and SPIN1) inside.

Here is a tear-down of ST-OV2- 12DC/ 60DC/1 – 2905022 solid state relay from Phoenix Contact. This product is old, obsolete and manufacturing has been discontinued.

Solid-state relay ST-OV2- 12DC/ 60DC/1 – 2905022 is a plug-in power solid-state relay (plug to certain Phoenix Contact wiring terminals), with LED and protective circuit in input and output circuits. Here is a simplified circuit diagram (from manufacturer data) of this solid state relay model:

This relay model has nominal input voltage of 12V DC. The relay output is designed to drive load at 12 – 60 V DC voltage range and maximum 1A current. More detailed data:

Input data
Nominal input voltage UN 12 V DC
Input voltage range in reference to UN 0.8 … 1.2
Input voltage range 9.6 V DC … 14.4 V DC
Switching threshold “0″ signal in reference to UN ≤ 0.4
Switching threshold “1″ signal in reference to UN ≥ 0.8
Typical input current at UN 7 mA
Typical response time 20 µs
Typical turn-off time 200 µs
Status display Yellow LED
Type of protection Reverse polarity protection
Surge protection
Protective circuit/component Polarity protection diode
Transmission frequency 500 Hz

Output data
Output voltage range 12 V DC … 60 V DC
Limiting continuous current 1 A (see derating curve)
Surge current 5 A (t = 1 s)
Peak offstate voltage 60 V DC (Collector-emitter reverse voltage)
Voltage drop at max. limiting continuous current 1.2 V
Output circuit 2-wire, floating
Type of protection Reverse polarity protection
Surge protection
Protective circuit/component Polarity protection diode

General
Test voltage input/output 2.5 kV AC
Mounting position any
Standards/regulations DIN VDE 0110b, Gr. C for 250 V DC
DIN VDE 0160 (in relevant parts)

Let’s take a closer look at the product:

Let’s open the case:

A look at electronics inside. This looks quite complicated construction that has many components and two circuit boards.

The product data also gives this kind of drawing that describes the circuit protection built into the relay:

Here is my hand drawing of the circuit inside SSR:

The signal input circuit is pretty basic consisting of polarity protection diode, current limiting resistor, zener diode and optocoupler LED. There is also a VDR between input pins for over voltage / surge protection. Those three zero ohms resistors make me wonder why they have used so many resistors. The reason for this could be that because there are variations of this relay also for other voltages (for example 24V DC input and 110V DC input), and on those higher voltage versions there could be need to use several resistors in series (to meet voltage and/or power handling ratings needed).

The output side is implemented with NPN power transistor that is driven with optoisolator output amplified with one PNP transistor. The all output transistors and optoisolator output have high enough rating to handle the full output voltage. On the output there is a surge protection component between output pins (striking voltage 65-71V). Theoretically the power transistor itself could handle much more current than 1A, but I think the output is limited to 1A because of thermal constraints (just transistor inside small case). Other limiting factor is that the IN4007 diode on the output is rated only for 1A current.

Component data of most important components:
2SD843 NPN power transistor (Vce 80V, Ic 5A , 40W, 10 MHz, hFE 70)

BC640 PNP transistor

SFH601-3 optoisolator

1.5K-68CA Transient Voltage Suppression Diode

Here is few years old SilverCrest USB power pack with solar charger that stopper charging a smart phone. Because changing the charging cable did not help and the charging connector felt a little loose, it seems that the power pack is to blame here. Ithad served several years well as portable power pack that had good capacity and pretty useless low power solar panel.

Because this device had passed the three year warranty that Lidl gave to it, it’s time for a tear-down and hopefully a successful repair with little effort.

To open it I had to find six screws (three per side) under plastic sticker.

Here is the power pack completely open. I need to be now careful because there the lithium battery and it’s electrical connections are now exposed – it can potentially explode if you mechanically damage it or cause short circuit on the circuit board. Do do not touch anything unless you know really what you are doing and how to do it safely.

Here is a close look to the electronics. As you can see the problem is that the soldering on the USB power output has failed. This is a quite common failure mode in this kind of products.

Repairing it just needed putting the connector back to the correct position and re-soldering the connections. I decided to add some mode solder to make to have better connections and more reliable securing of the connector shell. I also decided to re-solder the mounting holes of the microUSB connector as well just to be secure as there seemed to be quite little amount of solder in it originally. After soldering everything worked well.

Now everything is ready to put back the case.

For some more information on the electronics on this device, view this Lidl Silvercrest Solar USB Power Bank video.

I have earlier written about bad cheap Chinese wire connect that did not live up to specifications and was potentially dangerous for mains power at

http://www.epanorama.net/newepa/2018/01/24/cheap-quick-wire-connector-block/

Here is teardown of that

Let’s open open it

Here was all. The problem in construction that there is quite weak spring that pushes against wire. On the other side is just plastic that melts easily if there is heating. The contact area is small and spring is so this msterial that things start to heat up considerably at current more than few amperes!

I had some broked LED flashlights and I thought of using components from them.

Especially the compact battery holder that takes three AAA batteries looks useful in other small projects.

Those worked OK when lightly loaded. But when I started taking current for a small motor, I could not get acceptable power out of them.

Some measurements showed that when load current is increased, the output voltage drops considerably. It looks like there would be almost two ohms of series resistace inside the battery holder!

Investication revealed that each spring made of thin wire on the battery pack had around 0.5 ohms resistace. Three springs and some more thin metal makes almost two ohms.

Was this resistance because of extreme material cost saving? Or is it intentional distributed series resistance uses as part of simple LED current limited circuitry?

I can’t say for sure. But it can work as such series resistor. I could wire a 3W UV LED just directly to the battery pack output and get reasonable amout of light. The LED directly wired got slightly less than 1W power with the batteries I had as with somewhat used batteries current was around 200 mA. With brand new batteries current was around 300-350 mA, that is right for 1W LED.

End of test.