Today we’ll be taking a look at the Power Shield PSO-650 650VA Powerboard UPS, a 3+3 output UPS with surge/battery protection, phone line protection and USB for monitoring it on your PC. This particular unit failed because the 12V battery dropped to around 2 volts.
By looking at the power cord is going to tell us that this UPS is going to be built down to a price because the cord comes straight out of the UPS itself.
We’ve got our main PCB on the bottom and a huge transformer on the top which isn’t glued or bolted down, it just sits there due to the pressure the case puts on either end when closed. We have a 250V 5A circuit breaker on one side of the case. The surge protected outputs have some big MOVs on them. There are a few things which could be a little better, the heat shrink may have been too big that it’s pretty easy to move around and they left in an extra wire like it was meant to go somewhere else (not that it matter too much as the terminals screws are exposed anyway). The transformer is a 1065SP-2-V8/ E186x40.
We recently purchased a Toshiba Satellite Pro L50 for a client, took it out of the box and started setting it up. About 2 hours later, I hear a bang and the circuit breaker trips. After smelling the power adapter; it smelt bad so it was likely blown, it was sitting about 1 metre away from me. Luckily the laptop survived and to Toshiba’s credit they sent a power adapter which arrived the next day (I left it running for 24 hours without issues).
The power adapter is a 19V 3.42 (model PA5178E-1AC3, part PA-1650-60).
After taking a dremel to the plastic casing, we’re in. The metal shielding is connected to earth via a 1K resistor. So far no signs of damage.
Now we see lots of charring on the bottom of the board and most of it is pointing to around the front where the mains is connected near the fuse. Just above the black connector we have the diode rectifier, notice the resistor chain to the left and the PCB’s silkscreen has good markings where the pads, rubber or glue should go.
After desoldering the fuse, we can see that it’s blown pretty badly, it’s a 250V 3.15A. There must be some reason it’s blown or not?
At work, we use Snom phones and we take quite a number of calls every day, when you use the handset and need to type it’s a bit difficult. There are some headset/headphones options available, the Snom uses 4P4C / RJ9/RJ10 connectors but I’ve always got my iPhone headphones on, so I’d like be able to use them on the Snom phone.
The first step was to build a Snom 4P4C to TRRS cable, I used a 220 ohm resistor for protection to find the left/right speakers and then used a 1.5K resistor for the microphone.
It turns out that the microphone was a little hard to hear so after removing the 1.5K resistor it sounded better but it wasn’t loud enough unless you place the mic close to your mouth.
After speaking with a colleague, the idea came about of being able to switch between an iPhone and the work desk phone easily, it shouldn’t be a problem since we can an analog switch to do this. I had the Ti SN74LVC1G3157 SPDT switch on hand and it should do the trick, just hook up 3 of these for the mic and speakers.
Now it’s back to making a boost circuit for the microphone, I looked around for a low cost op-amp and found the LMV358 which is just like the LM358 and it’s rail to rail. I found an LM358 op-amp inverting circuit which we’ll use to start off with (shown above). I firstly tried a non-inverting opamp configuration but it didn’t turn out to be as good as the inverting one above. The configuration above boosts the microphone’s output (which is AC coupled) by 100x and the 10K/10K divider on the non-inverting side keeps the opamp’s output at half of VCC so that when sounds are produced there will be an AC waveform (e.g with VCC/2 being 2.5V, we could get a waveform that goes 1.5V to 3.5V) just before the 220uF capacitor.
Today we’ll be taking a look at the Dell Power Connect 2124, a 24 Port 10/100 Switch with 1Gbit uplink. It’s got the capability of a redundant power supply and has 2 fans on the back, it’s a fairly old switch dated back to 2003.
A few screws later and we’re in.
It’s a pretty dusty unit, we’ve got our power supply on the left which gives out 12V at 4.5A that connects to our main board and from the main board we have some cables running to the front panel. They’ve cut the board a little bit around fans to allow for better air flow.
We’ve got a pretty large Broadcom chip in the middle that was under a heatsink – probably one of the largest I’ve ever seen and there’s heatsinks on each Broadcom PHY chip. If you look closely you can see the capacitor on the bottom right hasn’t been soldered down properly and the inductors nearby are a little like that too. There’s a whole heap of 74 series logic on the left which go off to the front panel. On the bottom we’ve got a fair few passives and a nice looking line of them too.
As this is my 200th post (excludes some posts which I didn’t count), I thought I would do something just for fun/just because I can
I acquired have 6 loud fans from a server a little while back and have been trying to figure out what to do with them so I’ve built a kind of wind tunnel by using a cardboard box to mount the fans, bubble wrap with some cardboard for support and some bubble wrap on the top to seal most of it off (would have been nice if it was all see through from the outside).
Now add some small pieces of paper and we have a little tornado, it works but the papers do settle down the bottom most of the time.
Over the years I have collected hard drives, most of them a failing and were going to be thrown out but I kept them and thought what would it sound like for all of them to power up at once?
Well I’ve got 40 hard drives on 5 ATX power supplies (it was 4 before but 1 was too overloaded when starting up), so now we know!
Today we’ll be taking a look at the TP-Link 200Mbps Mini Powerline Adapter (TL-PA211), an ethernet over power (EOP) device which won’t power up – this is the second one of these that has had power up issues.
One screw later under the label and we’re in.
We’ve got the AC inputs on the left which are soldered on to the tabs, after unsoldering them the board pops out. There are a few heatpads on the top cover and the date code is 2012/11th week.
For the input power side of things we’ve got a Power Integrations LNK623DG AC to DC converter which can output up to 6.5W and this IC eliminates the need for an opto-isolator.
From our previous post, we made a small number of changes and started building acrylic cases/mounts for some of our boards. This time the cases/mounts have been build, a few more smaller changes and add in the SMS capability.
Here’s how everything looks at the moment – ready for deployment. I’ve got cases for the PIR, door sensors and the alarm system server which I’m planning to mount to the wall, connect the leads to the battery/adapter on the ground and the siren near the ceiling.
One problem that I’ve been able to find a simple solution to is how to mount the PIR on the corner of the wall like my current alarm system does. I’ve found that a 0.5mm aluminum sheet is flexible enough to cut with tin snips, mold into a U shape and it will keep it’s shape.
Today we’ll be taking a look at the MG-350HD Media Centre which is as it says, is a media center with Wifi/LAN, a 3.5″ IDE hard drive, USB with lots of outputs including component/composite/DVI/audio/optical so it’s pretty packed.
At first I wondered why it was so heavy, after opening it up there is a 750GB 3.5″ hard drive – so that’s why, it’s an IDE drive.
A few screws later and one screw under the wireless card, we’re in. We’ve got a fair amount of passives on the top with our Mini-PCI wireless card and on the bottom we’ve got chips with some regulators – 3x AZ1117 LDOs, 2x AP1605 SMPS and AP1501 SMPS. There’s flat flex cable connecting the optical/S-video and the front panel. PCB date code is 22nd week of 2007.
For the IDE cable, it looks like they have an ferrite core in heat shrink for EMI and it could be the same thing they are doing on the wireless antenna cable too. It’s interesting to see the video from the PCB to the connectors, the connecting conductors are pretty large.
A year or so ago I decided to make a smaller version of the Standalone Temperature/Voltage Logger which would only do temperature called the A25TTL which measured 17mm x 12mm. In this post, we’ll be updating the A25TTL to v1.1 so we can use any EEPROM, v2.0 so we can now also use an ATtiny13 and briefly show my new programming method for SMD ATtinys.
At the time, I used the cheapest 512K EEPROM I could find (the ST M24512) however I found that not all EEPROM’s SCL/SDA lines acted the same and I had run the ADC a few times to determine if the logger was connected to the reader – not the best way to do this. I recently found that the M24512 EEPROM doubled in price so this was the main reason to update to v1.1 which makes the logger to set it’s pull up resistors on the SCL/SDA lines at boot which means that we can now use any EEPROM.
watchdogSleep(T1S);// Wait a little bit for things to settle
// Check if we powered up from the USB port
system_sleep();// Sleep forever
This not only helps some EEPROMs reducing the current they draw but puts the ATtiny’s pins as inputs thus allowing the reader to pull SCL low. The logger waits 1 second, reads the SCL pin, if it’s low it goes to sleep otherwise it starts logging.
We have the original Xbox unit which is used infrequently and all of a sudden it wouldn’t power on, unplugged everything and measured no voltage on the outputs.
Naturally the problem would have to be on the power supply board, at quick inspection found black marks around a diode and resistor.
If you look more closely at the resistor you can see a slight mark which could mean it’s been damaged, I measured it in circuit and I was reading into the megaohms. I removed it from the circuit and it read open circuit, it looks to be 22 ohms.