29 October 2014

Deterministic error in temperature measurement (BAR206) - results

As announced, I opened my Oregon Scientific BAR206 station. The back cover comes out with a little help to unlock a small hook on both longer sides: just be careful. The circuit inside mine look like this:

Inside an Oregon Scientific BAR206

Humidity and temperature sensors are easily spotted, as highlighted in the picture above. With a bit of surprise, the temperature sensor is an analogue thermocouple. I found its tip sitting on the resistor, which could be the cause of the +0.5 °C constant error, added with an insufficient air flow.


I lifted the thermocouple and began the experimental comparison with an external sensor, one next to the other, BAR206 being without back cover (higher air flow).



Much to my disappointment, but without too much surprise, BAR206 body continued measuring 0.5 °C more than the external sensor. This confirms the observation at my parents' place with other three Oregon Scientific thermometers (one internal sensor in another model, two external sensors).

So this must be a firmware error or a design issue of analog circuitry inside BAR206. Even though some solder joints look awful (look at the huge blob under the thermocouple tip), the reason must be somewhere else. I may write to Oregon Scientific, but their specs say the accuracy is +/- one °C.

From now on I will mentally deduct 0.5 °C from the internal reading.

You may ask: why bother? I am an Engineer, I love accurate measurements and, most important, I hate recurrent (design) errors.

27 October 2014

Deterministic error in temperature measurement (BAR206)

I had doubts, but not a second device to do some tests, until my father bought himself an Oregon Scientific BAR206 radio-controlled clock and weather station, like mine. He noticed a difference of +0.5°C in the measurement of indoor temperature between BAR206 (production code 11A13 ?) and his old station.

So we gathered all his external sensors in one place and waited for readings to settle: BAR206 internal sensor was 0.5 °C higher. One value against three values.

I repeated the test at home with my BAR206 (production code 05A11 ?) and I could confirm the extra half-degree.

I think the reading error is due either to a firmware bug or the tight case that doesn't let air to flow inside.

I will now remove the back cover and see what happens. Check back for results.

16 October 2014

Lithium balloon!

Remember the inflated Li-poly battery I extracted from the family tablet? I had it laying on the balcony for two weeks, where it could release dangerous gas or catch fire without much damage.

Since it looked more inflated than ever, I checked it out. The inner pressure had now reached a level where I could not press the outer film: it felt like a properly inflated bike tire!

I am not planning yet a visit to the recycling centre, so I decided to vent the Lithium cell in order to avoid an explosion. A pin easily drilled a few holes in it (in open air!!) and internal pressure was released.

02 October 2014

Inflated tablet = inflated lithium battery!

Having an inflated tablet with a suspect inflated lithium battery inside the house is not a situation that should last long. Failing Li-xx batteries can release gas or catch fire, something I do not want to explain to my family and neighbours.

So I opened the Mediacom 930i tablet, that was half-open anyway, and finally saw the battery pack inside these devices. My tablet has two 4000 mAh cells in parallel and as expected one of them was puffy:


One of these Li-po cells is 3 mm thick or so: they could make an interesting battery pack for radio use, i.e. to fit inside existing equipment (FT817 and others). The cell on the right, incidentally where the liquid blob and backlight anomalies were, was swollen.



To the touch it felt like a partially inflated baloon.

I measured the voltage across red-black terminals and it read 3.65V: the two cells are in parallel.

What I did next was to carefully remove the blue tape in order to separate the puffy cell, cut its terminals to get it loose and lay it outside in an area where a casual flame would not set something on fire. Interestingly the swollen cell still held the charge. Inside the "battery pack" there was some sort of circuitry that I have not investigated, probably to select the most charged cell.

I was left with one good cell inside a working tablet device. It has been recharged overnight and now we have a tablet 80 grams lighter :-)





01 October 2014

Inflated Mediacom Tablet

Last time I picked up the tablet at home (a Mediacom Smart Pad 930i) I noticed it wasn't flat anymore. Looking at it through the short side I saw this:


The back cover has been pushed away and the display was bent too, as if someone had inflated a balloon inside it. I immediately suspected a faulty battery, but the tablet was working and holding the charge.

Actually a first sign of failure appeared one month ago on the display in the form of a liquid blob stuck under the glass:


But I though it was because of mishandling during holidays or just a failure after two years from purchase (warranty is over). The tablet never fell off. Another sign I have not recorded in pictures was a difference in backlight intensity in the border and one brighter spot above the apparent liquid leak.

These two alterations could be an early sign of upcoming battery failure in a tablet.

Next step: opening it up and getting rid of the battery, as soon as possible!

TBC

(For searches in Italian language: Batteria gonfia nel tablet Mediacom Smart Pad 930i. Continua nel post successivo.)

29 September 2014

DIY component tester (kit) - reading the microcontroller

Since the component tester kit was shipped with firmware version 1.05k, I would like to flash the newest 1.11k code available online from the original developer.

I do not have a spare ATmega328P at home, only ATmega168P's, so first I tried to dump the original 1.05k firmware: the chip is fully locked. I cannot even read current fuse settings without erasing it.


Too bad. I will now study limits of the code running on ATmega168 which has half the program space available of an ATmega328. Then, in case, buy (online, sigh) another microcontroller to experiment with.

27 September 2014

DIY component tester (kit) - assembled

The aforementioned component tester kit can be assembled in 60-120 minutes, depending on your soldering skills.

Not much to say about it. Components must be inserted on the side with solder mask ("top") and soldered at the other side. Just keep transistors close to the PCB and mount both the socket for IC and the pin-header for display.

There is no specific assembly order since the board is pretty wide. Sit down and have fun.