Teardown: Failed Asus eeePC 701 Netbook Power Supply (AD59230)

I’m not sure how many people today will remember the Asus eeePC 701, arguably, one of the earliest netbooks which drove the concept to its eventual boom. The 701 featured a stripped down specification, originally intending to be reminiscent of the one-laptop-per-child type device but utilizing Intel instead. They missed the price mark by quite a margin, and instead, it ended up being a new “toy” for developed countries instead.

I purchased my 701 for AU$415, overpaying slightly compared to the eventual $300 price point, but much cheaper than launch prices of AU$599. The device featured a 7″ screen with just 800×480 resolution, a Celeron M353 CPU underclocked at 630Mhz, and 512Mb RAM (shared with graphics). The storage was a tiny 4Gb, based around a CompactFlash controller and NAND flash, and it could run Linux or Windows XP. It had a tiny keyboard and trackpad to match, and patient users could find that it was capable of more than they would have expected.

It wasn’t perfect – it was a little on the slow side, and the battery life wasn’t spectacular, being about 3-4 hours. Charging the battery while using the netbook meant a 9-10 hour wait due to the power supply limitations, and the 4Gb SSD wasn’t fast (it stuttered) or capacious enough, so many users needed to augment it with SD card storage which was slow due to the on-board reader.

Regardless, I loved my eeePC, using it mostly for its first year, then it served three years as a dial-up modem server for the home, followed by a two year stint as a webcam server, and then finally helping me along with my PhD as a terminal to be used with a home-brew Arduino-based device.

Unfortunately, returning to my lab the other day, I discovered the laptop to be completely dead. No sign of life. Taking it home, and probing the power supply with a multimeter confirmed that the cause was a dead power supply.

Teardown and Post-Mortem

20150319-2027-3990

20150319-2027-3991The eeePC 701 power supply is a relatively compact brick. The unit puts out 9.5v at 2.315A, with a positive centre tip, although later units seem to up the amperage slightly to 2.5A. The model number is AD59230, and it is manufactured by Pi Electronics of Hong Kong/China.

The power supply comes with a two parallel blade fold-out configuration, which is localized to the Australian plug by a clip-on adapter. Fearing that there was a contact issue, I tried the power supply without the adapter but it was still not functional. It was likely something had failed somewhere inside.

 

20150319-2029-3995

Opening the power adapter was rather difficult due to a single Torx screw, embedded at the end of a thin screw-hole where none of my screwdrivers would fit. I decided to go ahead and pry at it until it broke, seeing it as a write-off. The first PCB is on the back-side, and has a date code of Week 45 of 2007, making it almost 8 years old. The fuse, full-wave rectifier, filtering inductor, and soft-start thermistor can be seen on this side. The casing itself already shows sign of issues – notice the slight liquid residue. This residue smelt a bit like methylated spirits.

20150319-2030-3996

As it turns out, the primary 1A fuse had blown, with some very nice melting, to render the supply safe. Externally, there was no sign of distress, which means that the design worked as intended.

20150319-2030-3997

The rectified DC passes to the other side where the rest of the SMPS is placed. A close look shows two main primary-voltage caps which have bulged, one (top right) more than the other.

Extracting the board from the casing proved to be difficult and ultimately destructive due to the use of adhesive to glue the board to the case.

20150319-2033-3998

The same liquid residue can be seen in the back of the casing.

20150319-2034-3999

The whole board ended up snapping along the primary-to-secondary isolation line. There is some mild sign of heat stress and electrolyte leakage from the bottom of the capacitor. There is a spark gap at the top right, which seems interesting.

20150319-2034-4000

The offending primary caps are both Ltec branded, a known bad cap brand. The design squeezes a vertical PCB in for thermal monitoring and shutdown, to save space.

20150319-2035-4003

The bulging is more obvious from this view. The capacitors on the secondary, which are Taicon, were both perfectly fine though.

20150319-2035-4004

Conclusion

As it turns out, this is another case of bad caps claiming another victim. If the board were extracted without breaking the PCB, it might be possible to re-cap and re-use the power supply and revive it.

As a person that doesn’t give up that easily, I’ve ended up reviving the eeePC 701 by using a power supply from the “upgraded” eeePC 701SD which I was given by a friend. This keeps my existing netbook going, whereas I salvaged the cable from this power supply to hook up to a lab benchtop supply to keep the eeePC 701SD going if I’m desperate enough.

About lui_gough

I’m a bit of a nut for electronics, computing, photography, radio, satellite and other technical hobbies. Click for more about me!

This entry was posted in Computing, Electronics and tagged , , , , , . Bookmark the permalink.

5 Responses to Teardown: Failed Asus eeePC 701 Netbook Power Supply (AD59230)

  1. sparcie says:

    I remember seeing the odd one of these for repair back in the day. Whilst they weren’t the first netbook, they certainly popularised it and began the demand for processors in the Atom line.

    I also saw a fujitsu machine that could be called a netbook that came out before the term became common. It was the u101 if I remember correctly. It was more expensive and more advanced than the early netbooks, but was quickly eclipsed of course. Very few people bought them.

    • lui_gough says:

      Hi Hi! Long time no see. Yeah, I’m sure they weren’t the first, but they were the first popular one to hit the shelves. The 701 pre-dates the Atom, which got introduced when the later eeePC 901’s started coming in. The Celeron M353 was a ULV CPU underclocked to 630Mhz, but I did note that it was possible with eeeCtl to overclock it back to 900Mhz or a bit further. I ended up running mine at an FSB of 110Mhz giving me 990Mhz effective core clock rate, and mine has been pure reliability (until the power supply gave way).

      Sure it wasn’t powerful enough for real work, but it wasn’t useless as a hobby machine for running single-tasking. I definitely liked mine, probably a little too much … and then I saw the ultrabook concept roll in, kill off the netbooks, while OEMs learnt not to race to the bottom meaning the prices remained mostly north of AU$650 for anything compact weighing <1.8kg.

      Ironically, even when the 701 and 701SD go, I still have another BenQ Joybook Lite netbook in stock ... one of my MSI Wind U100's let go earlier this year with a fan failure (too noisy) and inconsistent performance.

      - Gough

      • sparcie says:

        Yeah, they weren’t powerful, but later Atom processors actually did ok compared to the first ones which were about as fast as a Celeron half the speed. Your netbook being a Celeron and being clocked up to 990Mhz probably out-performed many newer models.

        There can’t have been much margin on Netbooks which would have been another nail in the coffin for them, that and ipads/tablets sorta took their market. There was so little margin that dedicated computer retailers didn’t sell them (that I knew of anyway), you could only get them from supermarket style stores, or general consumer electronics places like Dick Smith.

        I think the ultrabooks are probably worth the extra money, they are generally built better with a nicer chassis, and are as powerful as other portable computers. They are probably a better option than many if not most laptops.

        Sparice

  2. Wester547 says:

    In all fairness, from your description, the thing must have had 40,000 hours on it at least. That’s a good life for LTEC capacitors in a hot, cramped environment. And those Taicon may or may not be okay – only an ESR meter would tell you. Those lower ESR Taicon capacitors were known to wreak havoc upon older motherboards. Electrolytic leakage at the bottom is intriguing – that either denotes that the rubber bung is starting to decompose or that one of the leads is compromised.

  3. Satan 666 says:

    my eeepc 701 (running Arch Linux) :
    – torrent web client : transmission
    – mpd web client : ympd
    – web client file manager : file manager (yes that’s his name…)
    – sharing usb devices over my network (in my case it’s only for keyboard and mouse so i can use one combo of keyboard and mouse for all computers on my network) : usbip
    – I also using it every morning as alarm clock, playing some music… (using rtcwake and stuff)

    for security reasons all web interfaces are only accessible on my private (wired) network

    this baby has a lot of hours of work and never let me down so far! now I thought to change this baby for some odroid… where I’m working we have one odroid C2, I now rtc is not working with those devices. The eee pc 701 has (in my opinion) few problems, lack of space and the fan is noisy… I mean the major problem is the fan, because for the space I use some usb key to move the data…

    anyway it’s always funny to read about people using those *shit*,

    regards or cheers to aussies

Error: Comment is Missing!