News

Teardown Tuesday: USB Power Meter

December 06, 2016 by Robin Mitchell

In this Teardown Tuesday, we will split, smash, crack, and sand our way into a USB power meter to see how it's made.

In this Teardown Tuesday, we will split, smash, crack, and sand our way into a USB power meter to see how it's made.

The USB Power Meter

The power meter is housed in a transparent plastic case which is held together using internal tabs. These internal tabs can make it problematic to get inside as once the case is squeezed together, the tabs cannot be accessed from the outside and therefore rely on brute force to pry the case open.


The USB power meter

 

The display used on the power meter appears to be an OLED with readings for voltage, current, mAh, and time. There are no IDs or numbers on the display so from the outside the actual product code for the display cannot be determined. Shown here is the display when connected to a USB port from an old Dell computer (2006), using a two meter USB extender cable.

 

The display when plugged in

 

The underside of the USB power meter shows some of the internal workings including a chip-on-board (COB) method for the main controller. Other components that are barely visible include a large SMD resistor, various analog parts, and an SMD device.


The underside of the USB power meter

 

To restart the logging, a small button on the side is used (RESET). The button, itself, consists of a piece of plastic that makes contact with a small tactile switch and has flanges to keep it place.

 

The reset button found on the side of the USB power meter

Removing the Casing and the Inner PCB

With the outer casing removed from the USB power meter, the internal components can be seen. The LCD display has 24 pins with a backlight that sits underneath. The reset tactile button can also be seen, as well as some of the construction techniques with the PCB.

For one, the PCB has a black solder mask (which is becoming increasingly popular) but this has some drawbacks, including that it makes it difficult to see traces. The gold plating on this PCB is done through complete submersion of the entire PCB instead of trying to only plate specific parts (for example, a PCI card may have only the edge connector plated with gold while everything else uses a tin plate).

 


The topside of the PCB

 

The underside of the PCB shows the components more clearly. A large R050 resistor can be seen, which is clearly for current sensing on either the power or ground line (most likely high-side sensing as placing a resistor on the return ground path can lead to problems with circuit grounds).

The use of both through-hole parts and SMD components suggest that this device is made in several stages. One method that was not used was wave soldering. I say this because there are two through-hole pads that have no solder on them at all (connector on the far right). If wave soldering had been used, those holes would have solder. Therefore, this board was soldered by hand after the PCB was placed through a reflow oven.

 

The underside of the PCB

 

The two main ICs on the USB power meter show no identification numbers which suggests that either the parts were really cheap (e.g., the use of the chip-on-board) or that the exact workings of the device are intended to be kept a secret from other manufacturers.

The chip-on-board method is surprisingly cheaper than ICs with packages when dealing in large volumes and this method of construction is done by hand, as well. In fact, these chips arrive on trays and a person adheres them to the PCB and then places them into a machine which bonds the pads on the PCB to the pads on the chip using tiny wires.

The SMD IC is most likely for current sensing (such as a quad op-amp or a specialized current-sensing IC). The chip-on-board is most likely to be the main controller with many capabilities. The large number of traces going from the display to the chip suggest that the IC not only handles the display but may also handle the output of the current sense and all main GPIO processing.


The first IC with no identification
 
The second IC with no identification

Putting the “Tear” into Teardown

With almost none of the parts having any form of identification, it was clear that the USB power meter needed to really be torn apart. The first part on the kill list was the display. So, armed with a pair of wire cutters, the display was removed from the PCB, revealing the backlight.

 


The backlight for the display

 

Of course, this is not good enough for me. So, with a bit of prying, the back-light was removed revealing an identification for the display! According to the PCB, the display is a KWS-V5.0 with a date showing 2015 – 10 – 31. However, this could indicate a revision number of the PCB instead of the actual identification code as this number returns no results on Google.

 

The possible identification code for the OLED
 

The second IC to be unforgivably ripped apart was the SMD IC that had no part numbers on it. In hope of finding a number on the PCB, the pins were torn off the PCB and the part bent over. While the PCB had no ID number, the chip had a number on its underside with the ID JZY042916CV. Again, this number does not return anything on Google so that, too, was a dead end!

 


The unidentifiable SMD IC


A desperate attempt to see the IC under the epoxy was made by using a belt sander. Unfortunately, the sanding went a bit too far but the chip, itself, (along with a few bond wires) can be seen. The correct method of seeing the IC would have been an acidic mixture of hydrofluoric acid and nitric acid.

 


The IC chip and some bonding wires. Note that the other IC also got sanded!

Summary

This USB power meter (which is now in an electronic grave) shows many manufacturing techniques needed to keep costs low for products made in the hundreds of thousands (if not millions). Chip-on-board designs keep IC package costs low, SMD parts utilize pick-and-place machines, and molded plastic cases remove the need for screws, which further reduces the price.

Thanks for checking out this Teardown Tuesday! Stop in next week for another teardown!

 

Next Teardown: Robotic Floor Cleaner

7 Comments
  • edwardholmes91 December 06, 2016

    Nice work Robin, it’s always fun to take things appart and try to see how they work, shame there weren’t more useful part numbers… I guess they want to keep the design secret.

    You mention about the board not being flow soldered as there are two holes with no solder in.  I’d tend to agree, although the holes may have been plugged before? I’ve seen some flow soldered boards that have a rubbery resist protecting holes which are intended for the user to use and peeled afterwards, so they may have used that?  Unlikely though 😊

    Like. Reply
    • Robin Mitchell December 08, 2016
      Edward long time no see! Interesting about the resist protecting the holes and you learn something every day!
      Like. Reply
  • T
    tmig December 07, 2016

    Interesting teardown. You say at the end “this… shows many manufacturing techniques needed to keep costs low for products made in the hundreds of thousands (if not millions). Chip-on-board designs keep IC package costs low, SMD parts utilize pick-and-place machines, and molded plastic cases remove the need for screws, which further reduces the price.” All very true, but unfortunately for your teardown it also clearly shows many of the techniques used to keep people from tampering with/reverse engineering their product! I’m sure you were as frustrated as I was with their protection schemes as they frustrate our curiosity; but as you pointed out there is still something to be learned!
    Another side note - the display is not an OLED, as an OLED wouldn’t have a backlight - each of the pixels in an OLED are active elements that light up, so no need for a backlight. With all the pins and the way they are attached It looks like a custom LCD.

    Like. Reply