I needed a decent WiFi adaptor for my latest Pi LCD project, so after trawling eBay for cheapy USB adaptors, I found this one.
USB WiFi Dongle
Unlike most USB WiFi radios these days, it actually has a proper RP-SMA antenna connector, not the low-gain built in jobbies that never seem to work too well.
There are a few versions of this adaptor, all of which seem to use the same casing, there’s a button push cut into the plastic for a WPS button that doesn’t exist on this model. This is fine, as I don’t enable WPS on any of my network equipment anyway. (It’s insecure, and can be cracked in minutes).
MAC Address
Here’s the rest of the essential details, the model is BL-LW08-AR, rated at 300Mbit/s.
PCB Reverse
Here’s the PCB removed from the casing, there are a pair of PCB antennas on here, but they’re not connected to the RF circuitry in this model, the links are missing.
Chipset
The chipset used is a Realtek RTL8191SU, there isn’t much more in this device, as it’s all built into the silicon.
Here’s another torch from eBay, this time with 5 Cree XML-T6 LEDs.
Label
Having 5 Cree LEDs rated at up to 3A a piece, this light has the capacity to draw about 50W from it’s power supply. In this case though, current draw is about 1.5A at 12v input on the full brightness setting.
Cree LED Torch
Here’s the LEDs mounted into the reflector. Fitting this many high power LEDs into a small space requires some serious heatsinking. The casing is made of machined aluminium.
LED Module
Unscrewing the front bezel allows the internals to come out. The core frame & reflector is all cast alloy as well, for heatsinking the LEDs. The controller PCB is mounted into a recess in the back of the LED mount.
Controller
Here’s the controller itself. The usual small microcontroller is present, for the multiple modes, and handling the momentary power switch.
Switching Inductor
As all the LEDs on this torch are connected in series, their forward voltage is ~12-15v. The battery is an 8.4v Li-Ion pack, so some boost conversion is required. This is handled by the circuitry on the other side of the board, with this large power inductor.
Reflector
The reflector screws onto the front of the LED array, centered in place with some plastic grommets around the LEDs themselves.
LED Array
Finally for the torch, the LED array itself. This is attached to the frame with some thermal adhesive, and the LEDs themselves are mounted on an aluminium-core PCB for better heat transfer.
This module unsurprisingly generates quite some heat, so I have improved the thermal transfer to the outer case with some thermal grease around the outer edge.
Charger
The supplied charger is the usual Chinese cheapy affair, claiming an output current of 1A at 8.4v. I never use these chargers, so they get butchered instead.
Charger PCB
Here’s the main PCB. Overall the construction isn’t that bad, the input mains is full-wave rectified, but there is little in the way of RFI filtering. The supply is fused, but with an absolutely tiny glass affair that I seriously doubt has the ability to clear a large fault current.
Like many cheap supplies, the output wiring is very thin, it’s capacity to carry 1A is questionable.
PCB Reverse
On the reverse side, there’s a nice large gap between the mains side & the low voltage output. There’s even an anti-tracking slot under the optoisolator.
I almost forgot about this bit of kit, that came with one of my LED torches as a Lithium Ion charger. As I never plug in anything that comes from China via eBay, here’s the teardown & analysis.
Another Lethal Charger?
Here’s the unit itself. It’s very light, and is clearly intended for American NEMA power points.
Specs
Claimed specifications are 100-240v AC input, making it universal, and 4.2v DC out ±0.5v at 500mA.
Considering the size of the output wire, if this can actually output rated voltage at rated current I’ll be surprised.
Opened
Here’s the adaptor opened up. There’s no mains wiring to speak of, the mains pins simply push into tags on the PCB.
PCB Top
Top of the SMPS PCB. As usual with Chinese gear, it’s very simple, very cheap and likely very dangerous. There’s no real fusing on the mains input, only half-wave rectification & no EMI filtering.
PCB Bottom
Here’s the bottom of the PCB. At least there’s a fairly sized gap between the mains & the output for isolation. The wiggly bit of track next to one of the mains input tags is supposed to be a fuse – I somehow doubt that it has the required breaking characteristics to actually pass any safety standards. Obviously a proper fuse or fusible resistor was far too expensive for these.
The output wiring on the left is thinner than hair, I’d say at least 28AWG, and probably can’t carry 500mA without suffering extreme volt drop.
Since I have a fair few 750GB disks sat doing nothing, I figured I’d get some USB3 caddies for them. Back when USB -> IDE caddies appeared, they were hideously expensive. Not so much these days!
USB HDD
For £6 on eBay, you get a basic plastic box with the required bridge circuitry.
USB – SATA Bridge
Here’s the PCB – a very basic affair, with only 2 ICs. The large QFN IC on the left is the USB-SATA bridge. It’s a JMicron JMS567. Unfortunately JMicron are rather secretive about their bridge chips & I can’t find much information about it, nor a datasheet.
PCB Reverse
Here’s the other side of the bridge PCB – not much on here, the activity indicator LED is a bit of a bodge job, but it’s functional. The IC on the right is a Pm25LD512 512Kbit SPI EEPROM. This is used to store things like the USB device & vendor IDs, device name, type, etc. Here’s what dmesg spits out when the disk is connected on my standard Linux system:
[snippet id=”1769″]
Here’s some speed benchmarks:
USB2 Benchmark
First attached to a USB2 port, above
USB3 Benchmark
And finally attached to a USB3 port, above
Tests were done with a 320GB 5400RPM Samsung HM321HI drive, direct into the root hub, for the shortest possible signal length.
Alas, my old trusty Hameg HM303 30MHz oscilloscope has finally died. I’ve had this scope for many years, an eBay buy when I noticed they were going cheap.
It’s been replaced with a brand new Rigol DS1054Z, a 4-channel 50MHz DSO.
Scope
This is a big jump from the old analogue CRT scope I was using, it’s certainly going to be a steep learning curve!
System Info
I chose this scope through the help of the EEVBlog & it’s associated forums. Through this I discovered that I could upgrade the scope with a key to enable some extra features! In the above screenshot, the key has been applied, and the model number now shown is the DS1104Z.
This is the next scope up in the model chain, with many more triggering options, serial decoders, higher memory depth, recording & 100MHz bandwidth. While I rarely need to measure anything higher than in the kHz range, these options will definitely come in useful! The list of installed options is below:
Installed Options
And now for some sample waveforms, the scope has the option to save screenshots to USB flash disks, so when I make posts with waveforms in the future, the need to photo the screen of the scope is gone!
Having now tested the supply I wrote about in my previous post, I can now say that it’s nameplate rating far exceeds it’s actual capability.
On running the supply under load, at 6.5A the operating frequency drops into the audible range, a big sign of overload. (It makes an irritating continuous chirping noise). The output voltage also drops to 10.5v.
The temperature of the unit while it’s been running under such a load is also questionable, the external casing gets hot enough to cause burns, I haven’t yet been able to stick a thermocouple into the case to see what the internal temperature is.
I’m currently talking with the eBay seller (wwwstation) regarding this, however they are arguing that the supply is only for LEDs & CCTV cameras.
However those two loads are very different, and the supply has no internal regulation for supplying LEDs. As a simple switchmode supply, any load is suitable, providing it’s within the load rating of the supply.
I would estimate that the supply is only capable of 5A as an upper limit.
They are requesting that I return the supply, but I’m yet to find out if they’re going to cover return postage. The item as listed is not as described, and I will escalate things if required.
I will update this post when I hear more back from the eBay seller.
Following on from my recent power supply build, I’ve added on a couple of improvements:
Front Panel
I’ve added on my standard SpeakOn type 30A connector, a bank of push terminals for quick connecting test leads, and a 15A FSD ammeter.
Panel Rear
Due to the limited space inside the supply, I’ve had to improvise some insulation on the mains-side heatsink to prevent a nasty accident. The heatsinks are tied to the supply’s HVDC bus negative, so they are energized at -145v DC relative to mains earth. This fact has given me a nasty surprise! The insulation is several layers of Kapton tape, with a couple of layers of Duct Tape. This along with trirated wire to the SpeakOn & the panel meter should ensure safety.
The Ammeter itself was sourced from eBay, for £2. It seems pretty accurate so far!
Ammeter
The shunt is built into the rear of these meters, in an ultrasonically welded part of the case, so I can’t examine it. Hopefully it is indeed rated to 15A!
The only things left to make this supply complete are a mains power switch, and a fan speed control, as the fan I have used is a little noisy at full speed. It will be good to get the speed based from the internal temperature, so the fan only runs at full speed when the supply is under load.
A while ago I blogged about modifying the output voltage of some surplus Cisco switch power supplies to operate at 13.8v.
Since I was able to score a nice Hammond 1598DSGYPBK ABS project box on eBay, I’ve built one of the supplies into a nice bench unit.
Hammond ABS CaseSupply Unit
Above is the supply mounted into the box, I had to slightly trim one edge of the PCB to make everything fit, as it was just a couple of mm too wide. Luckily on the mains side of the board is some space without any copper tracks.
PSU Fan
These supplies are very high quality & very efficient, however they came from equipment that was force-air cooled. Running the PSU in this box with no cooling resulted in overheating. Because of this I have added a small 12v fan to move some air through the case. The unit runs much cooler now. To allow the air to flow straight through the case, I drilled a row of holes under the front edge as vents.
Output Side
Here is the output side of the supply, it uses standard banana jacks for the terminals. I have used crimp terminals here, but they are soldered on instead of crimped to allow for higher current draw. The negative return side of the output is mains earth referenced.
I have tried to measure output ripple on this supply, but with my 10X scope probe, and the scope set to 5mV/Div, the trace barely moves. The output is a very nice & stable DC.
This supply is now running my main radio in the shack, and is small enough to be easily portable when I move my station.
Now the final bits have arrived for the SWR Meter module, I can do the final assembly.
SMA Connectors
Here the SMA connectors are installed on the side of the eBay meter, for forward & reverse power tap.
These are simply tee’d off the wiring inside the meter where it connects to the switch.
Uncalibrated
The meter is connected to the module via a pair of RG58 SMA leads, above is a readout before calibration, using one of my Baofeng UV-5Rs.
I’m using my GY561 eBay Power Meter as a calibration source, and as this isn’t perfect, the readings will be slightly off. If I can get my hands on an accurate power meter & dummy load I can always recalibrate.
Tools are only as accurate as the standard they were calibrated from!
After calibration, here’s the readings on 2m & 70cm. These readings coincide nicely with the readings the GY561 produce, to within a couple tenths of a watt. SWR is more than 1:1 as the dummy load in the GY561 isn’t exactly 50Ω.
High Power VHFLow Power VHFHigh Power UHFLow Power UHF
Shortly I’ll calibrate against 6m & 10m so I can use it on every band I have access to 🙂
I recently posted about a small analog SWR/Power meter I got from eBay, and figured it needed some improvement.
After some web searching I located a project by ON7EQ, an Arduino sketch to read SWR & RF power from any SWR bridge.
The Arduino code is on the original author’s page above, his copyright restrictions forbid me to reproduce it here.
I have also noticed a small glitch in the code when it is flashed to a blank arduino: The display will show scrambled characters as if it has crashed. However pushing the buttons a few times & rebooting the Arduino seems to fix this. I think it’s related to the EEPROM being blank on a new Arduino board.
I have run a board up in Eagle for testing, shown below is the layout:
SWR Meter SCH
The Schematic is the same as is given on ON7EQ’s site. Update: ON7EQ has kindly let me know I’ve mixed up R6 & R7, so make sure they’re switched round when the board is built ;). Fitting the resistors the wrong way around may damage the µC with overvoltage.
SWR Meter PCB
Here’s the PCB layout. I’ve kept it as simple as possible with only a single link on the top side of the board.
PCB Top
Here’s the freshly completed PCB ready to rock. Arduino Pro mini sits in the center doing all the work.
The link over to A5 on the arduino can be seen here, this allows the code to detect the supply voltage, useful for battery operation.
On the right hand edge of the PCB are the pair of SMA connectors to interface with the SWR bridge. Some RF filtering is provided on the inputs.
PCB Bottom
Trackside view of the PCB. This was etched using my tweaked toner transfer method.
LCD Fitted
Here the board has it’s 16×2 LCD module.
Online
Board powered & working. Here it’s set to the 70cm band. The pair of buttons on the bottom edge of the board change bands & operating modes.
As usual, the Eagle layout files are available below, along with the libraries I use.
[download id=”5585″]
[download id=”5573″]
More to come on this when some components arrive to interface this board with the SWR bridge in the eBay meter.
After watching a video over at Scullcom Hobby Electronics on YouTube, I figured I’d build one of these precision references to calibrate my multimeters.
It’s based around a REF102P 10v precision reference & an INA105P precision unity gain differential amplifier.
For full information, check out the video, I won’t go into the details here, just my particular circuit & PCB layout.
In the video, Veroboard is used. I’m not too fond of the stuff personally. I find it far too easy to make mistakes & it never quite looks good enough. To this end I have spun a board in Eagle, as usual.
Precision Ref SCH – Click to Embiggen
Here’s the schematic layout, the same as is in the video.
Precision Ref BRD
As usual, the Eagle CAD layout files can be found at the bottom of the post.
And the associated PCB layout. I have added the option to be able to tweak the output, to get a more accurate calibration, which can be added by connecting JP1 on the PCB.
As in the original build, this unit uses pre-built DC-DC converter & Li-Ion charger modules. A handy Eagle library can be found online for these parts.
I have however left off the battery monitor section of the circuit, since I plan to use a protected lithium cell for power. This also allowed me to keep the board size down, & use a single sided layout.
Toner Transfer Paper
Here’s the track layout ready to iron onto the copper clad board. I use the popular toner transfer system with special paper from eBay, this stuff has a coating that allows the toner to easily be transferred to the PCB without having to mess about with soaking in water & scraping paper off.
Ironed On
Here’s the paper having just been ironed onto the copper. After waiting for the board to cool off the paper is peeled off, leaving just the toner on the PCB.
Etched PCB
PCB just out of the etch tank, drilled & with the solder pins for the modules installed. Only one issue with the transfer, in the bottom left corner of the board is visible, a very small section of copper was over etched.
This is easily fixed with a small piece of wire.
Components Populated
Main components populated. The DC-DC converter is set at 24v output, which the linear regulator then drops down to the +15v rail for the reference IC. The linear section of the regulator, along with the LC filter on the output of the switching regulator produce a low-ripple supply.
SMPS Ripple
Here’s the scope reading the AC ripple on the output of the DC-DC converter. Scale is 100mV/Div. Roughly 150mV of ripple is riding on top of the DC rail.
Linear PSU Ripple
And here’s the output from the linear regulator, scale of 50mV/Div. Ripple has been reduced to ~15mV for the reference IC.
In total the circuit as built has a power consumption of ~0.5W, most of which is being dissipated as heat in the linear part of the PSU.
I recently managed to score a 3″ B&W portable TV on eBay, a Panasonic TR-3000G. As these old units are now useless, thanks to the switch off of analogue TV signalling, I figured I could find a composite signal internally & drive the CRT with an external source.
Panasonic TR-3000G
Here’s the TV in it’s native state. Running from 9v DC, or 6 D size cells. I’m guessing from somewhere around the 1970’s. Here is the CRT & associated drive circuitry, removed from the casing:
CRT Module
After dissecting the loom wiring between the CRT board & the RF/tuner board, I figured out I had to short out Pins 1,2 & 5 on the H header to get the CRT to operate straight from the power switch. This board also generates the required voltages & signals to drive the RF tuner section. I have removed the loom from this, as the PCB operates fine without. It doesn’t seem to be fussy about power input either: it’s specified at 9v, but seems to operate fine between 7.5v & 14.5v DC without issue.
Video Connections
Tracing the wiring from the tuner PCB revealed a length of coax snaking off to the section marked Video/Sync. I successfully found the composite input!
Running OSMC
A quick bit of wiring to a Raspberry Pi, & we have stable video! For such an old unit, the picture quality is brilliant, very sharp focus.
Matsushita 85VB4 CRT
Closeup of the CRT itself. I haven’t been able to find much data on this unit, but I’m guessing it’s similar to many commercial viewfinder CRTs.
Electron Gun Closeup
Amazingly, there isn’t a single IC in the video circuitry, it’s all discrete components. This probably accounts for the large overall size of the control PCB. Viewfinder CRTs from a few years later on are usually driven with a single IC & a few passives that provide all the same functions.
After seeing these on eBay for £8.99 I thought it might be a good deal – interfacing with the RasPi’s GPIO & it has built in power supplies.
As a kit, it was very easy to assemble, the PCB quality is high, and is a fairly good design. It worked first time, the regulators hold the rails at the right voltages.
However there are some issues with this board that bug me.
The documentation for the kit is *AWFUL*. No mention of the regulators on the parts list & which goes where – I had to carefully examine the schematics to find out those details.
The 4x 1N1007 diodes required weren’t even included in the kit! Luckily I had some 1N4148 high speed diodes lying around & even though they’re rated for 200mA continuous rather than the specified part’s 1A rating, the lack of heatsinking on the regulators wouldn’t allow use anywhere near 1A, so this isn’t much of a problem.
Component numbering on the silkscreen isn’t consistent – it jumps from R3 straight to R6! These issues could be slightly confusing for the novice builder, and considering the demographic of the RasPi, could be seen as big issues.
On the far left of the board are the 5v & 3.3v regulators, well placed on the edge of the board in case a heatsink may be required in the future. However the LM317 adjustable regulator is stuck right in the middle of the PCB – no chance of being able to fit a heatsink, & the device itself seems incredibly cheap – the heatsink tab on the back of the TO-220 is the thinnest I have ever seen. Not the usual 2-3mm thick copper of the 5v & 3.3v parts – but barely more than a mm thick, so it’s not going to be able to cope with much power dissipation without overheating quickly.
As the adjustable rail can go between ~2.5v – 10v, at the low end of the range the power dissipation is going to shoot through the roof.
The GPIO connector – this could have been done the other way, at the moment the ribbon cable has to be twisted to get both the Pi & the GPIO board the same way up. Just a slight fail there. See the image below
Plugged In
The power rails are not isolated out of the box – there is no connection between the 5v & 3.3v rails & the Pi’s GPIO, but the GND connections are linked together on the board.
Getting the ribbon cable through the hole in the ModMyPi case was a bit of a faff – the connector is too big! I had to squeeze the connector through at a 45° angle. The case is also remarkably tight around the connector once it’s fitted to the board – clearly the designers of the case didn’t test the an IDC connector in the case before making them!
Everything does fit though, after a little modification.
All Cased Up
Here is the unit all built up with the case. The top cover just about fits with the IDC connector on the GPIO header.
More to come once I get some time to do some interfacing!
I bought one of these cheap HID kits from eBay to build a high-brightness work light that I could run from my central 12v supply.
At £14.99 I certainly wasn’t expecting anything more than the usual cheap Chinese construction. And that’s definitely what I got 😀
Potted PCB
The casing is screwed together with the cheapest of screws, with heads that are deformed enough to present a problem with removal.
As can be seen here, the inside of the unit is potted in rubber compound, mostly to provide moisture resistance, as these are for automotive use.
The ballast generates a 23kV pulse to strike the arc in the bulb, then supplies a steady 85v AC at 3A, 400Hz to maintain the discharge.
This module could quite easily be depotted as the silicone material used is fairly soft & can be removed with a pointed tool.
Hi-Lo Bulb Assembly
Here is the bulb removed from it’s mount. Under the bulb itself is a solenoid, which tilts the bulb by a few degrees, presumably to provide dim/dip operation for a headlight. This functionality is superfluous to my requirements.
This unit was bought from eBay to experiment with Magnetic Stripe cards, for little money. This unit is capable of reading & writing all 3 tracks, & both Hi-Co & Lo-Co card types.
Interfaced to a PC through USB, this has a built in PL2303 USB-Serial IC & requires 3A at 9v DC to operate.
The 3 Indicator LEDs on the top of the unit can be toggled by the included software for Power/OK/Fault condition signalling.
Unit Bottom
Bottom of the unit with the model labels.
Model Label
Closeup of the model label & serial number.
PCB Bottom
Here the bottom cover has been removed, showing the main PCB. The pair of large ICs bottom center interface with the magnetic heads. The IC above them has had the markings sanded off.
USB-Serial Interface
Closeup of the Prolific PL-2303 USB-Serial converter IC.
PCB Top
Here the connections to the R/W heads are visible, current limiting resistors at the left for the write head, a pair of signal relays, a pair of optoisolators & a LM7805 linear voltage regulator.
LEDs
Here is the trio of indicator LEDs on a small sub-board.
Frame Bottom
The PCB has been removed from the main frame here, the only component visible is the rotary encoder.
Rotary Encoder
The rotary encoder has a rubber wheel fitted, which reads the speed of the card as it is being swiped for writing. This allows the control logic to write the data to the stripe at the correct rate for the speed of the card. This allows the unit to write cards from 5-50 inches per second speed.
The Write head is directly behind the rubber pressure roller.
Read/Write Heads
Here you can see the R/W head assembly. The write head is on the right, read on the left. When a card is written to, it immediately gets read by the second head for verification.
If you’ve found my content useful, please consider leaving a donation by clicking the Tip Jar below!
All collected funds go towards new content & the costs of keeping the server online.
