Type Cover with Fingerprint Reader Misadventure

According to my records, I purchased my Surface Pro 3 in October 2014, along with the Type Cover available at the time. Mostly, I’ve been pretty happy with that hybrid tablet PC, but its Type Cover has always driven me to distraction. The touchpad, in particular, was trying, and the keys were closely spaced, so my typing accuracy suffered when using it. But what really kept me away from direct use of the device as a “laptop replacement” was my tendency to accidentally brush the touchpad while typing, often resulting in highlighting and deleting whatever it was I was trying to input at any given moment. For somebody whose keystrokes are literally his living, this was simply unacceptable. My wife and son love the machine, though, and it proved an able travelling device when a simple plug adapter was all that proved necessary to use the charger (and the device) for their extended trip to Germany this summer.

sp3-typecover sp4-type-fing

SP3 (old) Type Cover left/top; SP4 (new) Type Cover right/bottom.

When I learned that MS was releasing a new Type Cover along with the Surface Pro 4 earlier this fall I was mildly interested. But when I learned the following bits of information, my interest changed to a fully-formed “buy” decision:

  • The new Type Cover is backward compatible with the Surface Pro 3 even though built for the Surface Pro 4.
  • The touchpad is much larger, made of smooth, strong glass, and easier/more ergonomic to use.
  • The keys on the keyboard, though slightly smaller, are more widely spaced and offer a better feel while in use, all intended to promote faster, more accurate keyboard input.
  • A fingerprint reader version that works with Windows Hello is also available, as an added-cost ($40) option.

As fingerprint readers go, $40 is kind of pricy: most laptop makers charge $10 to $20 for these devices when they’re available as added-cost options, and USB models are readily available for $11 to $15, as this Google Shopping search shows.

While we were out and about on the weekend before Thanksgiving, visiting the mall at The Woodlands, TX, I dropped into the Microsoft Store and plunked down $160 of my hard-earned cash for the new Type Cover with fingerprint reader, even though it is currently available only in black (other colors are in production but probably won’t be out until late December or early January). I unplugged the old Type Cover and plugged in the new one when we got to our host’s home that evening and sure enough, it was quite a bit easier to use, enabled more accurate typing, and I didn’t have any touchpad problems or issues like those I experienced when using the previous version. I also quickly realized I had no idea how to set up the fingerprint reader, and a cursory pass through control panel widgets and settings options didn’t lead me anywhere useful, either. Realizing I had to do a little research to figure out which way to proceed, I shelved the fingerprint stuff until after I returned home late Monday (11/23) afternoon.

Once I went looking for the proper set-up, I found it almost immediately: Settings, Accounts includes a panel called Sign-in options. On a properly-equipped PC running Windows 10, you’ll see a Windows Hello item in that panel (it often requires a bit of scrolling down to find) where you can enroll or manage fingerprint data suitable for login (and other authentication) purposes. I’ve been working with fingerprint readers for years, and both of my Lenovo notebooks and my Fujitsu Q704 tablet all have them installed (I even wrote a mini-review for the Microsoft Fingerprint Reader for the 2006 Tom’s Guide Holiday Buyer’s piece). But though I was able to enroll my fingerprints — or so I thought — the reader was not able to read any of them. I found myself wondering if I didn’t have the 10/26 Firmware Update that added support for the Type Cover with Fingerprint Reader to the system (I did) or if the fingerprint reader might be somehow defective (it wasn’t).

What was defective, it turns out, was my understanding. If you take a close look at the image for the Type Cover with Fingerprint Reader above (right/bottom) you’ll see that the sensor occupies a pretty sizable surface area (it’s square, with dimensions of just under 0.5″x0.5″ or 1.27cmx1.27cm). Having been trained on fingerprint sensors that are about the same width but only 0.125″ or 0.32cm high, which require dragging a finger across them to enroll that data for the device to use later on, I was using the same finger dragging technique on this fingerprint reader as well. However, the device simply wanted me to lay my finger down on the sensor without moving it for registration and subsequent recognition. Until I stumbled onto this information (the registration prompt says “Press your finger against the fingerprint sensor, and then lift it,” which is what ultimately clued me into what I was doing wrong) the fingerprint reader wouldn’t recognize my fingerprints. Given a proper fingerprint registration, of course the device worked like a charm.

Duh! Sometimes what we already know causes us to ignore what we need to learn. Suitably chastened by my experience here, I’m sharing it in hopes of preventing others from falling into the same trap. Live and learn — if we can — has to be the watchword for this cautionary tale. Consider yourselves warned, and please benefit from my mistake! No need for concern: I’ll eventually recover from the embarrassment…

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

Samsung UFDs Muddle Through the Middle

In early September, I agreed to live with and review some Samsung USB 3.0 flash drives (to which I’ll refer from here on out as UFDs). Around the second week of the month, they arrived, and I was allowed until early December to play with and keep them for purposes of sharing my impressions and observations. Alas, I have dithered long enough that these items have already been reviewed at a number of sites, including one of my personal faves: Les Tokar’s The SSD Review which featured some pretty comprehensive coverage on September 18 in a story entitled “Samsung’s Newest Flash Drives Reviewed.”

Samsung-USB-3.0-Flash-Drives

Samsung FIT at front center; Samsung Bar at right rear.

My own observations of these Samsung devices are consistent with his, but not quite as flattering. These drives are neither exceptionally cheap (for that kind of thing, check out the Mushkin Atom drives in capacities from 8 to 64 GB at Newegg, at $7 to $25) nor exceptionally fast (there are many options of this kind available, including devices that actually use SSD controllers to access their Flash RAM, but I am fond of the Mushkin Ventura Pro line, at Newegg for $25 for 32 GB and $34 for 64 GB).

What I have from Samsung is a 32GB Fit drive ($16 for 32GB, $27 for 64 GB, Newegg) and a 16 GB Bar drive ($13 for 16 GB, $28 for 64 GB, Newegg). The performance is somewhat better than the Atom models, but not enough to offset the price difference between them. And the performance difference between the Samsung and the Ventura Pro models inclines me to to favor spending somewhat more to get a nice jump in performance (particularly on the write site of the read/write action that is file I/O).

This is a tough product niche with a lot of different products from which to choose, many at each rung of the price performance ladder. I will say that Samsung’s 5-year-warranty on its UFDs is better than most other vendors’ warranties (the Atom and Ventura Pro drives get 2 years, by way of contrast, and 1-2 years seems typical for most such products I’ve seen). It seems to me that Samsung will either need to lower its prices further, or find some way to bump its UFDs’ performance, if it wants to grab major market share in this crowded field.

My final assessment of these products is: if you can get a better-than-Newegg deal (especially if it’s cheaper than for Mushkin Atom items of similar capacity), jump on it. If not, you may be better served by looking around for something faster at the same or a slightly higher price (Nir Sofer’s USB Flash Drive Speed Tests are a great source for such data, BTW).

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

Unraveling a USB Mystery

About two months ago, I finally got a break in my schedule and used part of that time to replace the motherboard in my ailing production PC with a newer, more capable one. The “new” mobo is actually about two-three years old in technology, if still brand-new for the unit I rebuilt my system upon. It’s a Gigabyte Z77X-UD3H and it features more USB 3 ports than its predecessor, and support for an mSATA slot that now houses a dandy Samsung EVO 512 GB SSD. I was all ready to go off to the races, and my initial impression of the system was overall pretty positive.

z77x-ud3h

The specs looked good and I saw uniformly positive reviews, so I bought it.

About 5-6 days after the rebuild, right when I started moving my 1 TB of music files from older smaller drives to the newer bigger ones, I started losing USB 3 devices: first a couple of big drives (2 TB and 3 TB) connected to an Eagle dual-drive caddy, various plug-in ports on the case (from interior on-board USB3 header), and once my keyboard went away and didn’t come back. This motherboard features 4 Intel USB 3 ports, and an equal number of VIA USB 3 ports. Somewhere in those ports (and it looks like the Intel chipset is the culprit) one of those ports seems to jump in and out of failure mode. The big drives in my caddy seem particularly likely to go off-line, the more so when I’m doing huge file transfers over an extended period of time, which makes me suspect something heat related might be involved, or perhaps some of the circuitry that supplies those ports with power is flaking out under higher loads.

It took me quite a while to get to the bottom of this problem, not because the problem was especially difficult: the symptoms were obvious and the devices involved rather more consistently failing than mysteriously flaking out only every now and then. I just haven’t had a lot of time for troubleshooting research lately, and it took me a long time to find the right search string to get confirmation that this is a known problem (I ended up using the mobo designator and “USB port fail” and finally found reports from others of the same problem I’ve been having).

At long last, I finally stumbled upon some forum posts up on Toms Hardware where other buyers reported similar USB 3 problems and seemed to indicate that this is a known defect of this particular motherboard. I don’t want to replace it once again, so it looks like I will install one or two (I have a LOT of USB 3 devices on this machine, mostly for external storage) 4 port USB 3 cards that connect via PCI-Express x4 lanes and feature a separate controller chip for each connection so as to support max bandwidth for each one. The make and model is HighPoint RocketU 1144C 4-port USB 3 Controller card (here’s a Newegg link for this somewhat pricey item, that gets good reviews, on the site and at AnandTech). At $108 and change, I’m going to try one first and see how it goes. If I like the results, I’ll consider buying another one.

Something about the potential for less-than-optimal results, in the wake of this modest but nevertheless frustrating motherboard flake-out has me feeling more wary and chary of new technology purchases than usual right now. Wish me luck!

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

New Tablet/Notebook Storage A-Comin’

 

Now that I’m the proud owner of a Microsoft Surface Pro 3 (SP3), I’ve taken the plunge to buy into the largest microSDXC component currently available. These days, that’s a whopping 128 GB, which is an amazing amount of storage to cram into a device that’s about the same size as my pinky nail. I plunked down about $120 at Newegg for a SanDisk Ultra model that’s rated as a UHS-1/Class 10 device, which is supposed to make it the fastest device of its type around. Given that the device I’m currently using in the SP3 is rated the same — it’s “only” 64 GB in size, however, from the same maker — I’m not expecting blazing fast performance by any stretch of the imagination. This CrystalDiskMark output tells why, in its own inimitable way:

64sdxc

These speeds are on par with a 5,400 RPM conventional hard disk. Sigh.

But I can use the extra 64 GB of space, primarily because I’m using File History and my Outlook files weigh in at around 11 GB all by themselves nowadays. My total file history space is running at about 13 GB at the moment, per snapshot. You don’t need much history at that rate to suck up 100 GB (or more) in a hurry. I’ve set the file history space cap at 20% of the drive’s space (the most File History will allow), which means I can go two versions back, at most.

I’ve also purchased a couple of 128 GB conventional sized SDXC cards from PNY as well (for $70 a pop, somewhat cheaper than the $120 that the micro format costs right now, owing to “newness and scarcity”). These will replace the 32 GB SDXC cards in my Lenovo X220 and T520 laptops, likewise to extend storage there as well. The performance results for that media will probably be even slower than that for the SanDisk media on display in the preceding screenshot, but that’s life in the storage game right now. A USB3 flash drive is undoubtedly faster, but they still don’t have the 64 GB (or bigger sizes) in the teeny-tiny “Atom-style” format familiar to those using Bluetooth or minimalist 802.11n (or older networking technologies) in their notebooks nowadays. When they become available, and I can leave them plugged in all the time, that’ll probably be my next flash storage purchase for these hard-working portable PCs.

[New material added 11/12/2014 late afternoon]

The mail lady dropped off the new 128 GB microSDXC late this morning, and I had a chance to run some more benchmarks on the Surface later today. Here are some resulting screenshots, which I follow with some commentary:

surface-ssdsurface-128

The Surface SSD on the left; 128 GB microSDXC on the right

The results show that the 128 GB SDXC device is a tad faster than its 64GB counterpart, and that both are significantly slower than the built-in Samsung MZMTE256 SSD. I’ll report on the full-size SDXC devices when they show up in the next few days as well.

[New material added 11/12/2015, one year after this post went up]

Out of curiosity I decided to check on prices for 256 GB micro SDXC cards, one year after the first instances made their appearance, and one year after this post went live. I’m pleased to report that most 128GB devices of this kind now go for under $50, and you can find 256 GB [putative, it’s actually more like 240GB of actual storage inside Windows] devices for upwards of $65 to around $100 for all but the fastest such cards. Good news all around on the price front, then, especially those seeking to extend the capabilities of tablets with limited built-in storage (which I define as 64 GB or less, as is typical for most low-cost devices).

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

A Tale of Tablet Storage: USB 3 vs micro-SDXC

One thing I’ve noticed lately when it comes to breathless hype about Windows tablet storage is the fairly blithe assumption that by adding a micro-SDXC memory card, problems can be staved off or avoided. With a 64 GB (fairly affordable at around $35 to 60, depending on speed) or 128 GB (not yet affordable, with only one SanDisk model at around $120 currently available) card one can indeed extend the storage available to most Windows tablets (including the recently-announced Surface Pro 3 from Microsoft, plus countless others). But there’s something else about this form of storage that savvy buyers will wish to ponder as well, as illustrated in this side-by-side set of screen caps from the CrystalDiskMark disk speed analysis tool

tablet-storage

USB 3 on the left, microSD card on the right: which would you prefer?

It doesn’t take a rocket scientist to observe that speeds from the Mushkin 32 GB USB 3 Atom flash drive on the left vastly outstrip those from the SanDisk Ultra 64 micro-SDXC USH-1 card on the right, even though the SDXC card is rated at about as fast as such cards get, while the Mushkin Atom is by no means near the top of USB 3.0 UFD speed ratings. Methinks the interface for the card reader plays a significant role here on the Fujitsu Q704 tablet where I ran the test, and on other tablets as well.

There are two interesting take-aways from these speed measurements — namely

1. If you are going to use a micro-SDXC card to extend storage on a tablet, you’d better not expect it to run anywhere near as fast as the built-in SSD does, nor even as fast as plugging in an external USB 3 UFD. This also suggests that spending extra $$$ on faster SDXC cards may not be worth the added expense (though I’d want to test slower models to see how much slower they wind up going to be doubly-darn sure about that).

tablet-ssd

But then again, the SSD is oodles faster than either USB or micro-SD.

2. If you have to choose between a bigger UFD (size-wise that is, as anything over 32 GB requires a bigger “stick” in which to store the necessary memory chips) and a micro-SDXC card, the UFD clearly comes out ahead on performance, while the micro-SDXC card comes out ahead in compactness and lack of added external protrusions to carry and keep track of. Only you can decide which of those trade-offs is most worth making in your particular circumstances.

As for me, I’m opting for performance and capacity by using fast UFDs like the Mushkin Ventura Pro or the ADATA S102. Sure, they stick out from the side of the tablet, but they rock the speed and provide plenty of capacity. Then again, something like the newly-minted Corsair Flash Voyager GTX (just announced this week at Computex in Taipei) which includes an SSD-level flash controller and high-speed flash circuitry, might be just the ticket (for other options in this class of expensive peripherals check out Les Tokar’s USB 3.0 Archives at The SSD Review).

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

What I Don’t Like About the Fujitsu Stylistic Q704 Tablet Convertible

OK, it’s been a couple of months since my brand-new Fujitsu Q704 tablet with oodles of add-ons and appurtenances showed up at my door. I’m learning to live with the $165 monthly payment that my “let’s experiment with a business lease” ends up costing me. But I’m not as enamored of the machine as I’d hoped to be, and it hasn’t yet graduated to the status of “production traveling machine.” In fact, I’ve been using my older Lenovo i7 X220 Tablet and T420 notebooks more or less interchangeably on the road for the last two years, and am still happy to work with them.

FujQ704-stuff

As this image from the Fujitsu site shows, the Q704 is available with a battery-powered keyboard dock (center) and a plain dock with video and three USB 3.0 ports.

But now that I’ve been working with and using the Q704 unit for a while, here are my gripes:

1. The battery-powered keyboard dock is heavy, and turns the unit into a 4.75 lb notebook (it’s not even really an ultrabook, at that weight). Battery life is under 8 hours, too. By contrast my older Dell XPS-12 weighs 3.375 lbs, and gets 5 hours from a single battery.
2. When detaching the tablet from the keyboard dock, I’ve learned that it’s best to shut down. Mode switching from tablet to notebook by docking the unit, and from notebook to tablet by undocking same, is not as easy or trouble-free as I would like it to be.
3. As you might expect, an i7 (even a Haswell U4600) gets pretty warm, if not hot, when busy in the tablet, which can make it a little uncomfortable to hold in the hand or on the lap.
4. The touchpad on the keyboard dock, while sizable, is nowhere as nice to use as the one on the XPS-12, and the bottom buttons only click when pushed toward the center of the touchpad, not on the outside edges (as my fingers seem to want to do, to avoid pushing the right button when seeking the left, and vice-versa). For the kind of money Fujitsu charges for its gear, I’d expect them to use absolute top-of-the-line components, including the touchpad.
5. Given the presence of a U4600, the performance is less than you’d expect, and lower than even some higher-end i5 machines. See the Geekbench results at Tablet PC Review for more details. Read the whole thread to get the overall “nerd-view” of the Q704: it’s not pretty. Apparently, waterproofing perforce means less efficient ventilation and lower performance because of thermal inefficiencies. Sigh.

All in all, I’m hopeful that a combination of BIOS fixes and “best use practices” will help me extract a more acceptable experience from my Q704. Should be interesting to see how it all turns out.

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

Machine Crash Prompts Lots of Learning, Especially for Intel Rapid Start

OK, so I’m on the phone last Friday, and I’ve just installed Service Pack 1 for MS Office 2013. After the obligatory post-install reboot, I discover the system won’t boot because it can’t find a boot drive. A long bout of troubleshooting ensues, after which I reluctantly conclude my motherboard’s gone south. My immediate efforts concentrate on moving over to my back-up desktop system, and I put an order into Newegg for a new Z87 motherboard, with a i7 4770K processor.

By the time I got my standby system promoted to production status, and prepared my T520 notebook for backup status (which means it’s got all the apps I use installed, and is ready to play host to my Outlook PSTs and all my usual data files and stuff), the parts arrived from Newegg. Wednesday, I started putting pieces together, and managed to get all the way through the build process for my first shot at a new backup system. But that’s where my first lesson of this adventure came into play: I struggled mightily to get the Corsair CWCH cooler to mate properly with the MSI Z87-G45 mobo I’d selected (good price, with mSATA drive slot), but ended up having to switch for the stock Intel cooler because I had to keep fiddling with the CPU to get it to work.

Turns out my #10 Torx screwdriver came in really handy, because torqueing down the front screw to clamp down more tightly onto the processor was the only way to get things working. This necessitated three tries before I was able to boot into the BIOS. Getting the Corsair cooler in place ONCE was more than enough for me, so I’m using the stock cooler now, watching temps, and plan to drop in a Zalman CNPS95 (had good luck with these before in various earlier builds) if the stock cooler doesn’t cut it. At present, CPU temps mostly fall in the range from 40-50 degrees Celsius (about 10 degrees hotter than on the i7 930 CPU I just replaced, but I think that’s normal when scaling down from 45 to 22 nm technology). The Antec 902 case in which the build is housed is well ventilated, with four 120 mm fans, plus a 200mm venting warm air out the top.

My next big challenge came from a Windows licensing snafu. MS didn’t inform me until one day after the initial install that my key was a duplicate (why couldn’t they check that at initial validation, I wonder?). I wasn’t able to get tech support to issue me a new key (even though I’ve got an MSDN subscription and two unused Windows 8.1 keys, the activation utility would accept neither one of them). After trying every trick I know of, I elected to re-install, as much because I wanted to switch to RAID drivers and try out the Intel Rapid Start Technology, as because I got tired of trying to fix the licensing snafu I’d caught myself in.

rapstar

The GUI doesn’t show much, but it takes several contortions to get it working.

That’s where my next big learning adventure began. I quickly learned that I needed an SSD “hibernation partition” in which to shadow memory contents to make Rapid Start work, and I chewed through several sets of instructions before I figured out how to make it work on my MSI-based system:

0. I number this step zero because it occurs when Windows isn’t running (yet): you must get into your system BIOS to enable Intel Rapid Start Technology (and set the hibernation value to “Immediate” if present) before you can get the Rapid Start installer to work properly. You’ll have to catch and set this at an opportune reboot before attempting to install the software.

1. I set up the Intel mSATA 80GB SSD (nominal size; 76.29 GB actual) with a 42 GB partition, to leave 34 GB for the hibernation partition that Rapid Start uses to snapshot or copy memory contents. I used Disk Manager to set up a 42 GB GPT partition, leaving 34 GB unformatted for the series of diskpart commands I figured out I would need. Any SSD will do for the snapshot, and you should leave slightly more space in the hibernation partition than you have memory installed on your PC (mine has 32 GB; hence, a 34 GB partition). Interestingly, while the hibernation partition is visible in Disk Management, it is invisible to Windows Explorer (aka “File Explorer” or explorer.exe in Windows 8.*).

2. Next, I fired up diskpart following the instructions in the Intel Rapid Start Technology Guide for UEFI Mode, with special emphasis on the section entitled “Create a Primary Store Partition on a non-OS drive SSD.”

3. I skipped the convert gpt step, because the format I created in diskmgmt.msc was already GPT-formatted (in Step 12). I followed the create partition step as shown, using 34816 as the size of the hibernation partition in megabytes (34 GB * 1024 (megabytes per gigabyte) = 34816).

4. I jumped over into Intel Rapid Storage Technology, clicked the Performance tab, and made sure the acceleration features were enabled (this turns out to be an essential step in the process, though not very well documented).

5. I cut-and-pasted the exact value for the set id command from the documentation, namely “D3BFE2DE-3DAF-11DF-BA40-E3A556D89593” as shown in the Intel User Guide, then exited diskpart as directed.

6. Before you can get the Intel Rapid Start installer to work, you need to reboot the PC. After that, provided your chipset (Intel 8-series chipsets or better, with spotty 7-series coverage) and setup are copascetic, the program should install nicely and do its thing. One more observation: I had to reboot my machine two more times after installing Rapid Start to start observing the effects that the software promises — that is, a more rapid start. Apparently, it takes two reboots before the hibernation file gets set up and starts working properly, so be patient, please!

My PC now boots to the login prompt in under 10 seconds, which is at least 10 seconds faster than it used to boot before installing the Rapid Start stuff. Is it worth it? The real answer is that it depends on how often you reboot your PC. I mess around with mine all the time, so I think it’s worth giving up 34 GB of SSD space for this purpose. If you don’t reboot often, or you don’t have the SSD space to spare, you may not feel the same way.

But now, at least, I’ve figured out how to install and use Rapid Start on a home-brew PC, and I’ve got a RAID-based disk setup going for this UEFI Windows 8.1 install. The machine currently clocks at 8.1 in the WinAero version of WEI (Windows Experience Index, with the Intel 530 SSD the slowest link in the collection of subsystems measured) so I can’t be too unhappy with the results. By a slim margin (8.10 to 8.05 for my other desktop’s SSD), it’s now the fastest system I’ve got.

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

Interesting USB Access Issue on Fujitsu Q704 Worked Around

I’m still breaking in — or rather, getting to know in depth — my latest Windows 8.1 tablet. Somewhat annoyingly, the Fujitsu Q704 stops “seeing” a USB flash drive (UFD for short)  plugged into the keyboard dock once the machine has been idle for half an hour or longer. Continue reading

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin

Interesting Adventures with New Fujitsu Q704 Tablet/Convertible

Having secured permission from the Chancellor of the Exchequer (aka, “The Boss” — namely, my wife Dina) I recently purchased a new Fujitsu Windows tablet convertible, model Stylistic Q704 Hybrid Tablet PC with the keyboard dock/extra battery option. The price came in at over $2K, which is kinda painful for a 12.5″ tablet, but when the Boss said I could go for it, go for it I did. Now I’m learning to live with it. Here’s a snazzy publicity still:

fujitsu-q704jpg Continue reading

facebookgoogle_pluslinkedin
facebookgoogle_pluslinkedin