2014-11-09

Lenovo R500

Fingerprint reader (back of the palmrest)

Quite solid metal chassis

BCM5787 (LAN controller), MAX6622 (temperature monitor), Turbo Memory slot

Intel 5100 WiFi






Matsushita (Matshita) / Panasonic optical drive

Back of the palmrest - IBM/Lenovo/Foxconn

Touchpad

Touchpad
Pros:
  • solid construction; slightly weaker points are display twisting resistance and palmrest plastic (although metal chassis is directly under palmrest, so this is only minor crackling when pushing at right side)
  • mat finish and probably very high wear resistance (plain black plastics with no coating of any kind)
  • still perfectly fine performance for most home/office uses; inexpensive to upgrade with DDR3
  • low temperature (exhaust at both back and left side) and fan noise (although fan seems to be active all the time)
  • relatively small touchpad (I've disabled it completely in BIOS as it often generated random clicks for me when typing)
  • display proportions: 16:10
  • decent WiFi range / antenna set
  • great trackpoint (comparing with my previous Dell D600)
Cons:
  • poor USB port placement - all three are placed on left side near the front and also susceptible to mechanical damage - there are numerous reports of broken ports and they are quite expensive to replace
  • some noises from switching on-board voltage converters (quite disturbing at first, although I stopped noticing it after some time); surprisingly I haven't hear them when working on second R500 with T5870 processor - it might be individual issue
  • bad speaker sound quality at higher volume (resonating at some frequencies, especially right one - fixed partially by putting foam layer between speaker and optical drive)
  • quite fragile when it comes to fluids - that's how my first R500 died; it was just a small amount of water spilled at left bottom keyboard area, but there is nothing to prevent fluids from entering through the crack between palmrest and top bevel and DC/DC converters are very close to this area

Update. While R500 is definitely not resistant to water it turns out mine has survived. At first, after days of drying, laptop was reacting to power switch but screen was blank with inactive backlight. It started working again after resetting CMOS by removing backup cell (well, pity I got this idea very, late buying T500 in meantime).

2014-09-20

TP-LINK TD-W8951ND

WiFi/cable router with ADSL modem from late 2012.
Power consumption looks significantly lower than with W8950.

RaLink RT5390RL
RT63087N AFE (Analog Front-End) for VDSL2
EN25QH16-104HIP, 16Mbit NOR Flash.
LM1117.
34063-based voltage converter

Ralink RT63365E

TP-LINK TD-W8950ND

WiFi/cable router with ADSL modem. Manufactured in 2011, 802.11n support.



Quite unusual: Broadcom ADSL chip in SO8 package.
cFeon EN25F32-100HIP, 4MB SPI Flash memory: 64 blocks, 1024 4kB sectors, 256B per page. Block- and sector-erasable.
Two SMPSs based on 34063.

Second 4MB Flash chip and LM1117.
Atheros AR9285 with 2 antennas: external + PCB.


Pentagram P6391

Router with 4 LAN + 1 WAN ethernet ports and WiFi based on Atheros AR2318.

On the right: second antenna.
HY57V641620: 8MB SDRAM, 3.3V.
Marvell 88E6060 - low power ethernet switch with 5 independent PHYs, 6 independent MACs, auto crossover, lookup memory for 1024 MAC addresses. Interestingly, each PHY supports technology called Virtual Circuit Tester,  according to datasheet able to locate cable problems (open/short cable) with 1 meter precision, based on Time Domain Reflectometry. Chip can be configured with MDIO interface or with external serial EEPROM.
Classic DC/DC converter with 34063 and some external transistor in SO-8 package. There is rectifying bridge on board, router is intended to be powered with AC voltage.
25L1605 SPI Flash memory, 2MB, 32 sectors 64kB each, 10k P/E cycles, 3.3V.

Planet WAP-1963A

Planet WAP-1963A is a wireless access point compliant with 802.11b (11Mbps max) WiFi. Since original firmware supports only WEP it is pretty obsolete today.
AP is based on RTL8181 SoC, integrating Lexra LX5280 32bit RISC CPU, 802.11b MAC, 2 ethernet MACs and PCI bridge. Since same SoC is used by more popular devices, e.g. Ovislink WL-1120AP there are some homebrew linux distributions able to run on it (http://sourceforge.net/projects/shortlinux/, http://rtl8181.sourceforge.net/). See also: http://www.linux-mips.org/wiki/Realtek_SOC.
RAM: 2 x IS42S16800A (128Mb Synchronous DRAM, 3.3V from ISSI)
Philips SA2400A transceiver, applicable also for ISM purposes.
IP101 Single Port 10/100 Fast Ethernet Transceiver with MII interface.
Macronix 29LV160ABTC-90 16Mbit FLASH, interestingly switchable between 8 and 16-bit modes. Operating at 3.3V.
9435A (FDS9435A) P-Channel MOSFET and AIC1578 DC/DC converter.

2014-09-06

FX-350MS clone

 This very cheap (~$2.70) calculator was supposed to be my second/spare scientific calculator. Unfortunately it was disappointing in terms of both quality and function set.
This calculator is as generic as it could be. Under the sticker from local distributor there is just "Scientific Calculator" text, no brand name or anything like it anywhere (that may be a plus).
 Visually and functionally this looks like exact copy of Casio FX-350MS, although user manual was written poorly and quite incomplete.
 Device was dead on arrival with few of the LCD segments not working (exp part of the display).
 Another problem is very poor build quality, pressing buttons was causing severe distortions on LCD.
 My attempt to search for broken LCD connection and fix was proven unsuccessful as even careful manipulations were causing problems with other connections. Basically I believe that these connections are doomed to fail even on piece that is initially working. I have doubts if this thing passed DEKRA tests or these tests are worth a dime - when trying to search DEKRA seal database I couldn't get any result on ANY search term I've tried and this is supposed to be full-text search.

Another disappointment (same as with original Casio model) was lack of HEX mode considering it has advertised 240 functions.