LongRunner's Mini-review Collection

Everything goes... within reason!

LongRunner's Mini-review Collection

Postby LongRunner » April 25th, 2016, 11:51 pm

In order to pass the time I'd otherwise spend doing pretty much nothing meaningful…

Firstly, a brief rundown on the definitions (at least, as I use them):
  • A micro-review is something that amounts to little more than a summary of (some of) the item's properties. These can be multiple per post and will not have pictures unless they're easiest way to convey something important.
  • A mini-review is more detailed than a micro-review, but not enough to qualify as a full review (at least by my standards, which require proper test methodology to make the cut).
There will be no particular standard of "worthiness" for an item to qualify for this thread, beyond having some characteristic to mention that isn't a dead giveaway.

Here are the mini-reviews so far:
  1. Jaycar "Digitech" QP-2000 power point & earth leakage tester (Australia/NZ only) (this post, scroll down) – marginally useful, very cheaply (and nastily) constructed.
  2. Official Nintendo DSi/3DS AC adapter (WAP-002) (Australia/NZ version) – modest, but well-made.
  3. Allocacoc PowerCube Extended and PowerCube Extended USB (Australia/NZ version) – elegant, but with a few faults.
  4. Omniwheel's 3DStick+ – nice enough, provided you can tolerate the somewhat rough 3D-printed finish.
  5. Nintendo (3)DS 16-game case (not first-party, but officially licensed nonetheless) – looks alright, until you actually try to fit the games in.
    (And I wouldn't have complained if it was unlicensed. :rolleyes:)
  6. Stylux T-869 "Trax" desk lamp (Australia only) – alright looking, but a bit lazy on the wiring front.
  7. Kambrook KFH660 – a basic but competent 2.4kW fan heater.
    (See also: KFH600 addendum, with internal differences)
  8. Trent & Steele TS5 – a simple rice cooker, of middling build quality.
  9. Goldair GSFH110/FH-07A (also available under other brands) – a lesson in how not to build a fan heater.
  10. Projecta SB200SP automotive booster cable – a typical hardware-store piece of crap.
  11. A surge protector…? A detailed look at the reality (as opposed to popular belief) of these things.
  12. Laser PW-4ADAPT set - if this is legal, then something's wrong with Australian electrical safety standards.
  13. De'Longhi HVS3032 - a fan heater that's relatively expensive to buy, but built like a tank to justify it (unlike far too many pretentious heaters).
  • Round 1
  • Round 2
  • Round 3 - did you know there was a way to make a more decisive thermal fuse, even with 1997 technology?
  • Round 4 including a rant about just how much trouble you can get people into by butchering semantics.
If you want to discuss the mini-reviews, go here.

Jaycar "Digitech" QP-2000 power point & earth leakage tester (Australia/NZ only)
Front.jpg (122.29 KiB) Viewed 42067 times
So, this is obviously one of those three-neon devices (with all their limitations - U.S. site, but the concept is unchanged), combined with an ELCB/RCD tester with four test currents (10/15/30/100mA). I got this one years ago, so the orange casing has faded slightly, but Jaycar still has them. And already there's a typo: "Correct wring"? :silly: (Although they appear to have fixed that, going by the photo they now provide.) The labels indicating the wires each neon is connected between are also wrong; their correct order, from top to bottom, is: L-N, L-E, N-E. (Also fixed in the revision, they are no longer specifically marked as such.) It's also indicated (in broken English) that the instructions are on the back:
Back.jpg (133.07 KiB) Viewed 42067 times
To its credit, it does work as described, more-or-less. Except for the open neutral condition:
Open neutral response.jpg
Why is the N-E (wrongly labelled as "N-L") light glowing dimly? (More on that later.)
Open neutral response.jpg (96.06 KiB) Viewed 42067 times
This post would, of course, be pretty boring without actually going inside this thing. So here it is:
Innards.jpg (122.05 KiB) Viewed 42067 times
Soldering.jpg (122.01 KiB) Viewed 42067 times
Right away we get to see the cheap circuit board, with horrendous soldering. <insert barf smilie here>

The way the polarity-indicating neon lamps are connected is nothing unusual - just series 100k±5% 0.25W carbon film resistors (which actually manage to look cheap-and-nasty, although they at least measure within tolerance) limiting the current to ≈1.5mA through each. A bit close to their rating for comfort, but not too bad. The more interesting part is the ELCB/RCD test.

The highest test current of 100mA uses a series pair of 1.2k±1% 1W (going by the physical dimensions) metal film resistors (R4 and R5). The resistance itself is spot on (for 240V), but when testing that current they will dissipate 12W each until the protective device trips or the operator releases the button. :omg: Oddly enough, the PCB actually provides the option of installing 2W resistors here.
For 30mA a 5.6k±5% 0.25W carbon film resistor (R3) is added in series with the pair. Again, the correct ohms, but now this resistor dissipates 5.04W!! :runaway: Sure enough, plugging the unit into an unprotected outlet and holding down the button with the switch set to 30mA burned it out:
R3 burned.jpg
Well, it could have been worse. At least it stopped short of toasting the whole board.
R3 burned.jpg (117.78 KiB) Viewed 42067 times
(I see enough room on the PCB that they could have designed it to accommodate a 2W resistor there if they cared to…)
For 15mA an 8.2k±5% 0.25W resistor (R2) is added in series with those. Now the total resistance is slightly higher than "correct" (but to be fair, RCDs are meant to respond a bit below their "nominal" trip current, and 8.2k is the closest E24 value). This resistor has to survive 1.8W, which isn't quite as bad but still 7.2× more than it's rated for (unless it's a "mini" 0.5-watter, but I can't tell that from looking at it). Finally, the 10mA setting adds an identical 8.2k resistor (R1) in series with the other four. Then R1 and R2 take "only" 793mW each.

As for the green LEDs that indicate which leakage current the slide switch is set to (as if you're using it in the dark or whatever…), they are lit with a simple half-wave rectifier (1N4004 in combination with a 47k±1% 0.5W metal film resistor, which works at 613mW), connected between L and N to avoid an errant leakage current (to earth, anyway). This part also explains what's up with the unit's response to interrupting the neutral connection.

Even the choice of cord strikes me as odd. It's (allegedly) 1.0mm², which is a bit excessive given that this thing would catch on fire long before maxing out even 0.5mm² (the smallest allowed, and 0.75mm² is much more usual). Actually, even hair-thin 36AWG (≈0.013mm², even smaller than a single strand of a typical mains cord) wire could theoretically sustain 100mA - although that would break way too easily. And look at this:
Why is that last zero unevenly spaced? And given that the unit was made after 2004, shouldn't it read AS/NZS3112:2004?
Plug.jpg (116.92 KiB) Viewed 42067 times
It seems to be a counterfeit cord (even if the dimensions are about right). And when flexed, it makes noise a bit like that expected of a foil-shielded cable. :wtf:
Overall, I give it a D−; it's mostly functional at its intended purpose, but that's about it.

Stay tuned for more mini- and micro-reviews. :cool:
Last edited by LongRunner on June 30th, 2020, 12:50 am, edited 10 times in total.
Reason: Added micro-reviews round 4
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

LongRunner's Mini-review Collection: NDSi/3DS official P/S

Postby LongRunner » April 30th, 2016, 2:04 am

Official Nintendo DSi/3DS AC adapter (WAP-002) (Australia/NZ version)
Front view.jpg
Front view.jpg (70.19 KiB) Viewed 42001 times
Back view.jpg
Back view.jpg (77.63 KiB) Viewed 42001 times
Manufactured by Mitsumi. The internal circuit was redesigned at some point between 2010 and 2012, although the outside appearance remains the same (well, more-or-less; the label background is slightly lighter on my newer unit). And it has that annoying horizontal form that blocks off adjacent sockets; I'm sure these (and similar) designs have helped sell a few more extension cords. :silly:

The casing is held together with two security T8 screws. Remove them and then pry open (ideally with a spudger, although I don't have one) to get inside:
Overview (new).jpg
Overview (new).jpg (119 KiB) Viewed 42001 times
Solder side (new).jpg
Solder side (new).jpg (149.5 KiB) Viewed 42001 times
Overview (old).jpg
Overview (old).jpg (128.44 KiB) Viewed 42001 times
Solder side (old).jpg
Solder side (old).jpg (139.46 KiB) Viewed 42001 times
For a plug-pack supply it's pretty well-made, right? Electrolytics are all from Rubycon.

Starting at the input, both versions have a 500mA slow-blow fuse (oddly enough, the lower-rated DS Lite AC adapter was fused at 1A), and a (relatively) large 22Ω thermally fused (at 145°C) resistor for inrush limiting. The older version has a DF06MA (labelled as DFA06; 600PRV, 1A at 40°C, 30A surge) DIP-style bridge rectifier; the newer one has an SMD bridge that I can't identify.

Both versions have a pi filter (although the coil used will only have a noticeable effect at the switching-frequency range) after the bridge, with the coil in the positive side (I mention this explicitly because some primary-side pi filters have the coil in the negative side). The older version has a 4.7μF 400V 10mm of some mysterious custom series before the coil and a somewhat taller 400V 10mm (which I can't read the value or series of in-situ, due to the glue) after. The newer version has a 4.7μF 400V 8mm AX (a series which, according to the datasheet, is made especially for this application) before the ferrite coil and a custom 6.8μF 400V 10mm WXA after. What's more interesting, though, is that before the pi filter in the newer version, there's a 1μF 400V 6.3mm YXA (C13) followed by a 1N4006 (1A 800PRV) separating it from the pi filter. Two resistor networks draw from C13 (one apparently for start-up, the other for voltage sensing).

In both versions, a 115°C thermal fuse is mounted in the source circuit of the switching transistor, to disable the supply if the transistor gets dangerously hot. (Identifying the transistor itself is out of the question short of unsoldering stuff.) The newer version also has parallel 5R6 and 6R2 resistors there for current sensing. Both versions have some fairly complex snubbers.

The newer version has a noticeably smaller transformer. The old version had a clear plastic "fence" around it, which is apparently no longer needed in the newer version. Both versions have a contact connecting the transformer core to the "common" node of the primary side. The secondary windings in both versions are the "triple-insulated" type of wire, as providing adequate internal spacing to use ordinary magnet wire (the coating of which is only considered to be "functional" insulation) would be impractical in a transformer this small.

The solder side of the board has three small 6-pin ICs (2 on the primary, 1 on the secondary) in the older version and a single 8-pin SOIC in the primary circuit of the newer version (I couldn't identify any of them). The older version used an NEC PS2561A (W rank) optoisolator for feedback, but this was ditched in the newer version. The newer version has an additional 22μF 50V 5mm Rubycon YXF in the primary circuit.

The secondary rectifier in the older version is a parallel 2×SB240 (2A 40PRV Schottky), and in the newer version a parallel pair of Taiwan Semiconductor SR206 (2A 60PRV Schottky - which are somehow the same size as the DO-41 form of most 1A diodes, even though the datasheet claims the usual DO-15 form for 2A diodes). The capacitor at the start of the secondary pi filter is (in both versions) a 680μF 10V 8mm MBZ (which, with a 1.14A ripple current rating, should be more than adequate for the task). The cap after the coil is a mysterious 220μF 10V in the older version, and a custom-made 470μF 8V YXG in the newer version (which also has fewer turns on the coil); both are 6.3mm diameter. The newer version adds a snubber across the secondary rectifier (which was absent in the old version), and a 680Ω minimum load resistor.

Creepage spacing between the primary and secondary circuits is up to standard (at 7mm+), but within the primary circuits (from the DC side of the bridge onwards) is very narrow (in some places, less than 0.5mm between tracks over 300VDC apart). This may reduce reliability of the design, although it's hard to say by how much. I could see a whisker growing from pin 3 of IC1 in the older version (you can just about see it in the photo above), so I still don't totally trust lead-free solder in switching supplies. If I recall right, it was PeteS in CA (on Badcaps.net) who wished for the envirocrats responsible for RoHS to find their lives dependent on RoHS compliant medical equipment… The standard of the soldering itself is mostly OK (for lead-free), although pin 3 of the older version's optoisolator has an oversized blob.

The plug wires are UL AWM style 3386 (600V, 105°C XLPE) in 24AWG (which, while smaller than allowed for actual mains cords, is fine for the tiny current requirement of a charger, given appropriate fusing). The output lead is a single-core shielded 24AWG UL AWM 1571, which over its length of ≈1.9m will drop about 250-300mV at the full rated load of 900mA. And, of course, there's the annoying proprietary Nintendo connector (but at least they've kept it the same since the DSi).

Cracks.jpg (99.97 KiB) Viewed 42001 times
If anything, the weakest part might be the casing; there is evidence of multiple hair-line cracks in my older unit, which suggests that the plastic they chose (clearly marked as PPE) isn't very forgiving of flexing. Given that the casing stands between the user's hand and the full mains voltage, I think they should be made to be virtually unbreakable, short of using a hammer (I should probably add such a requirement in the next revision of LRS 002 :D).

The unloaded output measures close to 5V on both versions. Not having a 5.1Ω 5W (or higher rated) resistor on hand (or, for that matter, an oscilloscope), I can't do a load test. Both versions have working short-circuit protection.

The older version weighs 58g excluding the output lead, the newer one 54g.

Overall, I give a B−; while good-quality components are used and with better-than-average attention to fail-safety, the awkward form is irritating and the very small creepages within the primary circuits don't inspire total confidence in the overall design. Personally, I would rather have it in a rectangular form, a tad bulkier but with proper creepage spacing throughout, and using an IEC C8 inlet (although Nintendo never seems to have used them to date). And there's no MOV…
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

LongRunner's Mini-reviews: Allocacoc PowerCubes

Postby LongRunner » October 23rd, 2016, 12:14 am

Allocacoc PowerCube Extended and PowerCube Extended USB (Australia/NZ version)

The top and front have an outlet each, as do the back and underside (hidden from the camera).
PowerCube.jpg (86.21 KiB) Viewed 41839 times
One of those items that's "love at first sight", and winner of multiple awards, this Dutch re-invention of the humble power strip (or power-board as referred to here in Australia) makes a great antidote to awkwardly shaped plug-pack supplies like the above example. Having got some (4 PowerCube Extended + 1 PowerCube Extended USB) recently, I figured it was about time for a deeper look into them.

The USB ports, opposite the cord entrance. Non-USB versions instead have a fifth mains outlet here.
USB.jpg (66.9 KiB) Viewed 41839 times
As the name suggests, they come in the form of a cube measuring 64mm (or about 2.5″ for the metrically challenged) on a side, in contrast to the traditional bar/strip/board shapes. The Cube weighs about 180g excluding the cord (although it's hard to get precise measurements without actually removing the cord from the unit). The outlets (4 on USB versions, 5 on non-USB – though one will be blocked off unless you suspend the Cube in mid-air by its cord) are in various orientations, so that you can (hopefully) find a way to fit whatever combination of plug-pack supplies you may have. A mounting base (which I haven't bothered photographing) is also provided, which can be fixed to a surface (either with the provided adhesive strips, or two small screws spaced 30mm apart) and the Cube then twist-locks onto it. (I should note that on my two green Cubes, the twist-locking action required excessive force; if you encounter this issue, judicious filing of the responsible parts may be needed.) I couldn't find a hint as to how much weight a mounted Cube is allowed to support, but I would suggest that 2kg is a sensible maximum.

The cords are standard fare, H05VV-F3G1.0; the only thing that stands out, apart from the restraining piece that you can slide along the the cord (and fix in place with a small screw), is some additional text identifying the cord as belonging to a PowerCube:
PowerCube Extended – designed by | allocacoc | in the netherlands – protected intellectual property

The claimed length is 1.5m, I measure about 1.4m.

That slot is fully insulated, so there's no need to unplug the Cube from the wall before resetting (just the offending loads from the Cube).
Nameplate.jpg (81.47 KiB) Viewed 41839 times
A resettable thermal circuit breaker is included for overload protection, as with all standard Australian power-boards, only it's concealed, requiring the use of a small bladed screwdriver to press the reset button. I tested it on one unit to be sure and it tripped in 24 seconds with two 2.4kW heaters connected (and 6 seconds with a total 6.8kW connected), which is a reasonable response time at double the rated loading. Early PowerCubes had no surge suppressor and copped some flak for it, but now they do. We'll see how they've implemented it in due course (given the compact size of the unit, we can't expect much more than a few MOVs in there).

Nameplate off.jpg
Although not clear in this photo, the cover actually has the same "RESET FUSE" and arrow markings from the nameplate (albeit moulded on rather than printed), just in case the nameplate is ever lost.
Nameplate off.jpg (70.18 KiB) Viewed 41839 times
To get inside the PowerCube, the first step is to remove the nameplate (made of 1.0mm polycarbonate, and flexible enough that there should be no risk of accidentally breaking it), by releasing the eight tabs holding it in place. Then there are four quite long screws with T10 heads; after removing them, you can lift the cover off.

Under the cover (1).jpg
Under the cover (1).jpg (87.48 KiB) Viewed 41839 times
Cord clamp inside.jpg
Cord clamp inside.jpg (102.28 KiB) Viewed 41839 times
Cracked cord clamp.jpg
(This is in another, non-USB Cube. Hopefully, someone will 3D print replacements or something…)
Cracked cord clamp.jpg (57.37 KiB) Viewed 41839 times
The cord clamp is a two-piece arrangement, which in addition to keeping the cord secure also serves as a spacer to keep the outlet assembly firmly in position. Separating the two parts (by removing the screws) reveals a series of ridges on the inner surfaces, to keep the cord from slipping (although I should note that some of the ridges are rather "soft", looking almost as though the plastic melted slightly). Unfortunately, there are cracks in some. :silly:

Under the cover (2).jpg
Under the cover (2).jpg (120.73 KiB) Viewed 41839 times
The circuit breaker is the common type, albeit with shortened terminal tabs to save a bit of space. (The wires are consequently soldered right up against the body of the breaker, but it is made of a hard thermosetting plastic.) The short red wire from the breaker to the outlets is silicone insulated, curiously enough. For some reason, one of the screws securing the breaker housing was missing in my USB Cube.

As basic as a surge suppressor can get.
MOV.jpg (73.12 KiB) Viewed 41839 times
There is a reasonable-sized MOV (BrightKing part # 471KD14J) connected between active and neutral, but with no thermal fuse. :rolleyes: At least it's not an “all-modes” arrangement – those are truly the work of the devil.

Look into socket contacts.jpg
Look into socket contacts.jpg (43.37 KiB) Viewed 41839 times
Spacer removed.jpg
Spacer removed.jpg (45.57 KiB) Viewed 41839 times
The socket contacts are formed from 0.4mm phosphor-bronze and grip the plug pins well enough. If you ever feel the need to tighten them, the insulating spacer can be pulled off and the contacts given a squeeze.

IMG_1039-edit.JPG (100.64 KiB) Viewed 27764 times
IMG_1040-edit.JPG (80.34 KiB) Viewed 27764 times
IMG_1041-edit.JPG (86.23 KiB) Viewed 27764 times
IMG_1042-edit.JPG (78.2 KiB) Viewed 27764 times
Here's the shutter mechanism (it will be different for other regions, of course).

USB supply.jpg
Unimpressive, but at least it doesn't use the obsolete 2-transistor arrangement.
USB supply.jpg (182.67 KiB) Viewed 41839 times
The USB power supply is built on a double-sided epoxy-fibreglass PCB (1.0mm thick) and starts with a 10Ω 1W fusible resistor (better than nothing, but far from ideal), followed by a bridge made of four M7 diodes (SMD equivalent to 1N4007). The four electrolytic capacitors are all BH (allegedly) CD288L series. A 10μF 400V 10mm follows the bridge, then a small ferrite choke (sleeved in heatshrink, which is good) paralleled with a 5.1k resistor, then a 4.7μF 400V 8mm (both primary electrolytics are 12.5mm tall). The switching IC is an OB2538AP, which has an integral MOSFET with 4.4Ω "typical" on resistance (the maximum isn't specified so make of that what you will :rolleyes:) and suggested output power up to 20W at 230VAC±15%, or 15W full range (85–264VAC), which is ample for the rated output of the unit. A 4.7μF 50V 5mm (by 11mm tall) electrolytic smooths the supply to it. An RS1J (1A 600V fast recovery) diode rectifies that supply, and another (presumably identical, but buried under the silicone caulk) diode forms part of the primary snubber. The transformer looks decent enough for the purpose. There is a 1nF Y1 capacitor from the negative side of the primary rectifier to the secondary "ground", which is standard practice but the charge in these capacitors is capable of ruining small-signal circuits with great gusto if given the chance; with the earth already available at the outlet block, they might as well extend it to the USB supply (or, if possible, design the supply to not need the Y cap). An SR540 (5A 40V Schottky) is the secondary rectifier (theoretically adequate, but will get quite hot at the full rated loading) and the lone output cap is 1500μF 10V 8mm (by 20mm tall). This doesn't appear in the datasheet for the series (the only one that does is the 4.7μF 50V, for which Chemi-con KY/Nichicon HE is a suitable replacement), but from the values that are listed it's clear that the rating won't be more than 1.4A – for a nominal 2.1A flyback converter, this is marginal. (Indeed, there are quite few options that will fit there and be adequately rated apart from a polymer cap.) A 560Ω load resistor is also included. The USB sockets are mounted on a separate PCB (1.2mm thick), with both sharing a common resistor network:

USB board.jpg
USB board.jpg (71.64 KiB) Viewed 41839 times
+5V – 10k – D− – 6.2k – GND
+5V – 5.6k – D+ – 6.2k – GND

Connection to the power supply board is via a short red silicone-insulated wire (+) and a loop of orange UL AWM 1672 (300V, 105°C PVC, jacketed) in 22AWG (−). Two longer orange UL AWM 1672 22AWG wires provide the input connections. The casing also has a plastic barrier, to prevent any unsafe displacement of the USB ports.

Solder side.jpg
(This was after removing and resoldering the SR540, so ignore the flux residue around its pads.)
Solder side.jpg (115.32 KiB) Viewed 41839 times
Creepage distance between the primary and secondary circuits is up to standard at 7mm; creepage distances within the primary circuits are a bit skimpier, down to 1mm between the traces connecting to the 4.7μF 400V cap (and, for that matter, between the drain and "ground" pins of the OB2538AP). There is no coating visible (apart from the solder mask).

Size comparison.jpg
Even allowing for the fewer outlets on the Cube, that's quite a difference. (Although the SPB 1 is admittedly a bit longer than average for 6 outlets.)
Size comparison.jpg (187.56 KiB) Viewed 41839 times
For comparison, here is an Arista SPB 1 from probably some 15~20 years ago (it predates the insulated-pin mandate, at any rate). A boring 6-outlet horizontal strip, but a solid version. Its cord is 0.9m long (and again the normal type) and it has a neon illuminated combination switch/CB (which still glows at a reasonable brightness after what must be >100k hours, although it does shimmer a bit). The bus-bars inside are made of robust 0.8mm brass, and overall construction is considerably sturdier than present-day versions. Weight is around 260g excluding the cord; a flimsy "modern" equivalent weighs about the same as the PowerCube. Cord retention is OK, though could be a bit better (I can budge it a few millimetres with enough force).

Convenience: A. Quite simply, they are the most elegant way I've yet seen of dealing with oversized plug-packs. The one slight downside in this department might be the concealed circuit breaker, but you shouldn't be tripping it on a regular basis anyhow.
Style: A−. I'm not a huge fan of gloss, but at least it's white so the effect is less obtrusive than with glossy black. The colour options for the outlet faces are also a nice touch.
Build quality (outlet block): B−. The socket contacts are a definite step above average, but the shutters aren't too great and the cracks in a few of the cord grips are also somewhat concerning.
Surge protector (not present in older versions): D−. Personally, I would rather have had nothing here than an unfused MOV, inside a plastic housing. I've added a salvaged thermal fuse to my USB Cube, although I only have one more so can't do all of them at the moment. (I might also upgrade the MOVs to Panasonic ERZ-E series, which have one of the highest energy ratings for their size.)
USB supply: D+. The semiconductors seem OK, but the bare-bones output filter and minimal safety provisions fall short of what I would like. To be fair, it is rather cramped in there (it has to supply more than twice the power of the Nintendo supply above, from a PCB no larger), but there are still a few tricks they could have used to "borrow" a bit more PCB area, such as inserting the fusible resistor (or, preferably, a proper fuse) in one of the AC leads, using a SIL bridge rectifier (e.g. KBP or GBP), mounting the output diode upright…

Overall: C. While nifty, and sturdier than an average modern power-board, the USB supply is a let-down and the very basic surge protector isn't entirely safe.

Altogether, this might be fair justification for a "PowerCube 2.0". (The original with no surge suppressor being 1.0, with the current version 1.1, I suppose.)
Last edited by LongRunner on August 4th, 2019, 5:33 pm, edited 2 times in total.
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

LongRunner's Mini-review Collection: 3DStick+

Postby LongRunner » November 17th, 2016, 11:00 pm

And now for something a bit smaller and simpler:

Omniwheel's 3DStick+
Broken the Circle Pad on a Nintendo 3DS? Don't despair – replacements are out there, courtesy of Omniwheel, who has designed the compatible – and arguably superior – 3DStick+. It is available via Shapeways in nylon or metal versions (plus another metal version that can be made-to-order with a custom emblem), with several colour/alloy options on offer. I got two of the nylon versions – one white and one blue (both "polished"). The 3DStick+'s main claim to fame, though, is its ridges (apparently modelled after the old GameCube controller) added to prevent slipping (which Nintendo ditched from their official analog sticks many years ago, for whatever reason).

Installing it (in any model of the 3DS family; though I don't know if anyone's tried it in a Circle Pad Pro) isn't very difficult, provided due care is taken (in particular, not to tear ribbon cables). Instructions can be found elsewhere on the web, though some may include extraneous steps; or you may be able to work it out yourself. (You will have to reset the clock after the repair, though, as the systems don't have a backup battery.)

Anyway, here's a (slightly dirty) white 3DStick+ on a white 3DS XL:
3DStick+ (white).jpg
3DStick+ (white).jpg (63.83 KiB) Viewed 41731 times

With a bit of cleaning, it looks quite nice:
3DStick+ (white, cleaned).jpg
3DStick+ (white, cleaned).jpg (54.86 KiB) Viewed 41731 times

And here's a blue 3DStick+ on a black original 3DS:
3DStick+ (blue).jpg
3DStick+ (blue).jpg (72.08 KiB) Viewed 41731 times

Control: A. With the 3DStick+'s ridges, my thumb doesn't slip, however long I hold any one direction.
Comfort: B+. Like the originals, the 3DStick+ has a nice concave shape, though some people might not love the feel of the ridges. (That's just a compromise that has to be made, though.) For those people, though, a plain 3DStick (no +) is available.
Appearance: B−. They look neat enough to start with, and the colour options are nice, but the combination of the ridges and the rougher finish (even on the "polished" versions) compared to factory parts results in greater dirt accumulation – so, especially on the white version, more frequent cleaning may be necessary to keep it looking good. Fortunately, the control doesn't seem to suffer much.
Durability: B? I haven't used it long enough for a definitive answer, but it ought to hold up better than the original stick; indeed, the potentiometer tracks in the Circle Pad mechanism may wear out before the 3DStick+ gives way.

Overall: B+. A straightforward and cost-effective repair, and perhaps a worthwhile upgrade for regular 3DS players even if the original stick isn't yet broken.
Last edited by LongRunner on November 19th, 2016, 9:55 pm, edited 1 time in total.
Reason: Added links to 3DStick Custom and plain 3DStick (which I forgot initially)
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

LongRunner's Mini-reviews: 3DS 16-game case (licensed)

Postby LongRunner » March 15th, 2017, 10:21 pm

This one is out of frustration, really. The item is a case (like this one, only purple/pink instead of clear/black and strangely enough, that one didn't have the "3" in the logo even though it did accommodate the extra tab on the 3DS cards) that holds up to 16 Nintendo (3)DS game cards, with a removable adapter also included so you can use one of the spaces for an SD card instead. Although not actually made by Nintendo themselves, it does have their seal, so it would seem reasonable to expect it to at least work as it should. Yeah, you wish…

Overview.jpg (139 KiB) Viewed 40976 times
Latch end.jpg
Latch end.jpg (53.36 KiB) Viewed 40976 times
At first glance, it looks decent enough: Tough polycarbonate outer casing, reasonably robust latch, accurate Nintendo 3DS logo. So, what could possibly go wrong?

Ridges.jpg (95.08 KiB) Viewed 40976 times
These ridges, which are about 0.55mm thick and cause the game cards to jam and pop-out (I had to shave them off with a utility knife, as shown, to salvage it):

Poor fit 1.jpg
Poor fit 1.jpg (150.01 KiB) Viewed 40976 times
Poor fit 2.jpg
See also, the SD card adapter on the left.
Poor fit 2.jpg (155.84 KiB) Viewed 40976 times
Poor fit 3.jpg
(Yes, I did get those lower two games in the "wrong" way around.)
Poor fit 3.jpg (151.72 KiB) Viewed 40976 times
The trays are also prone to popping out of the hinge at the left, to add insult to injury:

Lousy hinge.jpg
Lousy hinge.jpg (44.19 KiB) Viewed 40976 times
You'd just think Nintendo would have checked that the games fit in properly, before giving their endorsement. :silly:
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

Stylux T-869 "Trax" desk lamp (Australia only)

Postby LongRunner » May 16th, 2017, 9:19 pm

I wanted a decent lamp at my bedside for reading, and a quick shop left me with this Stylux branded desk lamp (alongside which I got a 5W warm white LED bulb to fit in it); while far from the best review candidate (an integral LED luminaire would be quite interesting, but that'll have to wait for another day), I was still keen to check for any problems with it. Stylux promote "European Style Lighting", whatever that means (I didn't care about that in the least).

I'm not bothering with an overview photo, as you can get a good enough idea of what it looks like from the product page. There are 9 plain paint colours (black/white/red/orange/yellow/green/blue/purple/pink) and 3 metallic finishes ("Brushed Chrome", "Gun Metal", and "Antique Brass" which I have and is also the only one with a differently finished gooseneck; all others have a chromed gooseneck) to choose from.

Appearances aside, an E27 lampholder is used (standard fare in Europe, I'll give them that), and the cord itself is H03VVH2-F2X0.75 (as is customary for small Class II appliances) and measured 1.6m "net" (from the grommet at the base unit to the strain relief at the plug). Nothing out of the ordinary - or is there? More on that below…

Well, that spoils it…
Magnet.jpg (118.39 KiB) Viewed 40698 times
It's clearly not solid brass, unfortunately, as magnets stick to it strongly, hinting at ordinary steel construction with just a brassy finish (over-coated with a thin film of plastic). (There was a little bit of rust visible on the inside of the metal base, too.)

Electrical safety
Here comes a more serious issue: This luminaire only marginally complies with the double-insulation standards.
Insulation issue (1).jpg
Now, that's a lazy heat-shrinking job if ever I saw one!
Insulation issue (1).jpg (95.69 KiB) Viewed 40698 times
Insulation issue (2).jpg
I used my heat-gun before taking this picture, but the incoming active/line remains partly exposed. Clearly, they should have used somewhat longer sleeves over the individual wires…
Insulation issue (2).jpg (100.78 KiB) Viewed 40698 times
As can also be seen, from the way the neutral wires don't go even near it, the switch is only a single-pole type. If the neutral remains connected correctly, this is OK; but if neutral and active/line are swapped anywhere upstream, or the neutral is broken upstream with any other loads switched on downstream of the break, then the outer contact of the lampholder will become live, even while the switch is off. For the retail prices of these (below), the cost increase to use a double-pole switch would have been trivial. I would definitely not recommend using it in continental Europe (or any other regions that use non-polarised plugs), either, whether by way of a plug-in "travel" adapter, or a permanent plug change (not that there's any compelling reason to bother taking this unit to another country).

Odd, this. Both pins are out-of-tolerance at the short end, yet the manufacture quality otherwise looks too good to be a counterfeit…
Plug.jpg (145.81 KiB) Viewed 40698 times
The plug pins measured about a millimetre shorter than usual (compared to an assortment of other AS/NZS 3112 plugs I had to hand, both moulded-on and rewireable), which is actually (to my knowledge) out of tolerance; their width and thickness were fine, but in any case, I cut off (and destroyed) the original plug to make way for a side-entry type (Deta 6333, which includes a grommet to help grip H03VVH2-F and the likes thereof; white, grey, and transparent versions are also available if desired, although the grommet remains black for all), which fits behind my bed better. Incidentally, the product page for the plug shown in the photo is here (unfortunately only in Chinese).

I checked behind the lampholder and the appropriate fibreglass sleeving is present, and there's a suitable piece of plastic film between the wiring and the metal stud in the lampholder back cover (into which the mounting screw threads). Still, when it comes to safety, it's the weakest link that matters most.

In conclusion, it doesn't justify AU$79, or even the AU$55 that the plain colours sell (sold?) for, especially with only borderline compliance with safety standards. The grade, then, is a D.

For now, it'll have to do, but when the time comes for a replacement, a native LED luminaire is the only thing I'll be interested in. (Even if it means importing from Europe and changing the plug - which, as noted above, I had to do anyway to fit it behind my bed without kinking the cord…)
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

LongRunner's Mini-reviews: Kambrook KFH660

Postby LongRunner » July 3rd, 2017, 6:26 pm

Kambrook KFH660 fan heater (Australia/NZ)

For a while now, I've thought electric (resistance) space heaters to be “over-hated” by many people – mostly economists/environmentalists (for somewhat valid reasons, at least), and people concerned with “safety” (for reasons which aren't all valid). And for whatever reason, fan heaters seem to get the worst of the hate. But rather than going into every which detail about those, I'd better catch up on my review plan (which I've had in the making for nearly 2 years); while I need more test equipment before I can make a full review, I can probably still do far better than most (largely superficial) “reviewers” of these (including consumer magazines with their nominal – but not always true – “experts”; no doubt, the Dunning–Kruger effect befalls many of them).

So I figure, best to start with a common model, and here I have one – the Kambrook KFH660, a 2.4kW (240V × 10A, the full rating of normal Australia/New Zealand outlets) fan heater with a room thermostat and oscillating base (although I don't generally use oscillation on fans…); lower-priced variants include the KFH610 (without oscillation, but now discontinued), and KFH600 (without oscillation or a thermostat; still available, see here for internal differences). (Kambrook previously made the mistake of implementing a uselessly crude fixed "thermostat" in the model KFH200 – by way of setting the thermal protector to a ridiculously low temperature – but rest assured, they've learned from that.) Recommended retail prices are respectively AU$69.95, AU$59.95, and AU$49.95 (GST included); my unit was AU$68.00 at the local RetraVision, bought 2017-05-19. (They also offered the KFH310, which looks different but has basically the same function as the KFH610, and is listed at the same price too.)

Like most (if not all) Kambrook appliances, the KFH6x0 has (according to Kambrook, anyway) been endurance tested – albeit for a less than impressive 1,000 hours. There are a few unexplained differences between the unit in the “hero” shots and the real product (namely, a slightly different switch knob, and the label for the SWING switch and the DO NOT COVER notice are both missing from the promotional picture).

When I originally wrote this post, this statement on the fan heater selection page stuck out like the proverbial sore thumb; however, it has been removed in the site redesign:
Kambrook fans keep your home cool without making a huge dent on your wallet.

Well, a fan alone might not make too much of a dent, but a 2.4kW heater can rack up several hundred dollars per year if used frequently. (Fortunately, I don't need that much heating myself – though my sister is a different matter…)
And if you want to be pedantic, fans inside a room can do nothing to cool the room – all they can do is facilitate heat exchange with objects at a different temperature from the room (hotter items will be cooled down, colder items will be warmed up). Cooling the room (albeit while warming the outdoors) is what air conditioners are for.

But enough picking nits from that – let's get onto the review proper.

As is expected of a “budget” appliance, the casing is plastic (polypropylene according to an embossed mark on the base) – albeit with a slightly metallic paint finish. (I don't know about you, but I'd rather have just plain grey plastic; the money then saved could be put towards build quality, or a longer warranty.) I noticed a bit of crunchy noise when trimming bits of flash off the molding, suggesting some glass-fibre reinforcement. The “face” has a slight cylindrical curve (albeit angled back a bit, as usual). The overall feel is reasonably solid, and it weighs about 1.8kg.

Yeah, that paint is already coming off… (Honestly, does anyone other than manufacturers actually like metallic-painted plastic?)
Controls.jpg (88.42 KiB) Viewed 40202 times
Like most basic fan heaters, a 4-position rotary switch and rotary adjustable thermostat are used – though someone at Kambrook decided to reverse the rotation direction of the switch (normally, they switch through off-fan-low-high as the knob is turned clockwise). “Difference for difference's sake”, eh? :s

The knobs themselves are no ergonomic wonder, either, as one ProductReview.com.au member complained about.

Lower frontal view.jpg
Normally in Australia (like the UK), our switches are arranged down=on/up=off. But I suppose it's not really a problem, as long as the labelling is present (and some people reckon the American up=on/down=off is more logical)…
Lower frontal view.jpg (126.42 KiB) Viewed 40202 times
The oscillation/swing is controlled by a separate rocker switch (circular in this case) at the base, which is also typical enough.

Base view.jpg
Base view.jpg (292.73 KiB) Viewed 40202 times
The tip-over switch is implemented in the simplest (and cheapest) way – a peg at the base, which has to be held in to power the heater. This is decidedly less than ideal, as such an arrangement is trivial to override (just tape it flat); that said, this is probably no catastrophe here. (A radiant heater, on the other hand…)

The base has four hollow cylindrical pillars apparently designed to hold silicone feet, although none have been installed. (The KFH600/610 don't have that possibility.)

Annoying tag.jpg
It escapes me how a heater like this one, with its plastic grille, could deliver dangerously hot air without melting itself down…
(Presumably, these tags are tacked on by the legal department, without actually consulting the engineers.)
Annoying tag.jpg (95.72 KiB) Viewed 40202 times
The cord is type H05VV-F2X1.0 (as expected) and has a useful length just short of 2m (196cm to be precise). (Of course, I'd prefer a C18 inlet – though using a counterfeit cord with this would be mortally dangerous…) A hyperbolic warning tag was attached, which I promptly cut off after photographing (and cleaned off the adhesive residue with IPA). (Tags with valid warnings survive by me, but this one doesn't exactly pass that test…)
The cord grip is not the strongest I've seen, but it's passable at least. The cord itself (branded Mainland, plug model was “ML-240” although their site lists no Australian standard plugs :huh:) is neither the most flexible I've seen, nor horribly stiff.
Unfortunately, as is typical of molded-on plugs these days, the wire-to-pin terminations developed increasing resistance (and thereby heating) fairly quickly; so I ended up fitting a rewireable plug (Clipsal 439S) after the warranty end (which is still fine another year later).

The manual (included with the unit in paper form, but also available as PDF here), in case you're wondering, says nothing in particular regarding extension cord use – though it does recommend a record 1.7m clearance in front of the heater!! :lol2:

The oscillating base and main housing are each held together by four T15 “security” screws (though the necessary torque is low enough that you may get away with using a T10 driver). Remove them and you can get inside; all internal screws have plain Phillips #2 heads.

The wires (white) to the neon lamp are UL AWM 1007 in 22AWG, while all the other internal wiring is UL AWM 1015. The main active (brown) and neutral (blue) wires are 16AWG, the switched wires to the individual heating elements (yellow and red) are 18AWG, the synchronous motor's leads (black) are 24AWG and the other black wires (mostly for the fan) are 22AWG; all of these are ample for their respective loads.

The neon bulb itself is quite small (measures 3.7mm diameter by 9.2mm long), but thankfully has a sensible 510kΩ (±5%, carbon film, 0.25W judging by size) series resistor. (I hate when manufacturers overdrive neon indicators with too-low resistances, especially in items such as power boards/strips that will be powered for a long time…)

Switch rear view.jpg
Switch rear view.jpg (70.94 KiB) Viewed 40202 times
Sleeved connectors.jpg
Each end of this sleeve has one connector in it.
Sleeved connectors.jpg (47.71 KiB) Viewed 40202 times
Inside base.jpg
The cord grip, incidentally, is integral to the base and cover, secured by the pair of screws either side of the cord…
Inside base.jpg (135.4 KiB) Viewed 40202 times
Crimp indent.jpg
I (temporarily) took the sleeve off the neutral connector to inspect it; it seems to be crimped satisfactorily.
Crimp indent.jpg (19.42 KiB) Viewed 40202 times
A majority of the internal "blade" receptacles have silicone insulating boots. The 3 “cup”-style crimp connectors (a pair behind the controls, feeding the neon+fan+oscillator; and one in the base, joining neutral from the cord to the "internal" neutral) are sheathed in braided fibreglass sleeves, which I think is a neat way of reducing the fire (or electrocution) risk if they go bad. (Even though only that latter one actually bears the full load of the heater!)

Controls rear view.jpg
Controls rear view.jpg (96.06 KiB) Viewed 40202 times
The rotary power switch is (apart from its oddly reversed rotation direction) the common type, as is the thermostat. (I'm not sure why Kambrook put the thermostat in the neutral; aside from leaving the elements live while the switch is on but thermostat off, this also appears to complicate the wiring harness, for no apparent benefit.)
Added washers.JPG
Added washers.JPG (51.98 KiB) Viewed 26234 times
The mounting screws for the thermostat are a tad over-long, so I ended up inserting a washer under each to compensate.

The heating resistance wires measure about 0.5mm diameter (24AWG) and stick to a magnet, indicating iron-chromium-aluminium (FeCrAl, also known by the trademark Kanthal) alloy. They are wound onto their mica “rings” in the usual fashion. Measured resistances (for respectively the front and back elements) are 50.6Ω and 51Ω (cold), giving true powers (at 240V) of ≈1138W and ≈1129W respectively (combined, ≈2268W which is about 5.5% short of the nominal); and at the actual working temperature, this can be expected to drop another 2~3% (as their resistance rises a little). (This is probably intentional to some extent, to allow for tolerances and avoid going over the rating; but does it really need to be that far down?) Their length then calculates to 7.0~8.3m each, depending on the exact alloy used (based on the official Kanthal resistivity figures; that data can be found here), but I'd guess at Kanthal AF, AE (either of which would take ≈7.3m) or D (which would take ≈7.5m) or similar.
The “surface load” (power density over the surface of the resistance wire) ultimately calculates to 9~11W/cm², which seems reasonable enough.
The front-most element (with the yellow feed wire) is the one used for half-heat, which seems to be the norm in this type of heater.

Thermal cut-outs.jpg
Yeah, it's a little rusty here too, suggesting that the plating (nickel?) has some weak points. (I've seen it before in a few other heaters that the steel terminals got rather rusty, even while the resistance wire itself is in good condition. Maybe stainless steel would hold up better, but that has much higher resistivity which might be a problem…)
Thermal cut-outs.jpg (94.32 KiB) Viewed 40202 times
The primary thermal protector is a Klixon TH10CA061 (which trips at 75±5°C and is rated for 30,000 switching cycles at 250VAC 13A), and the thermal fuse is set for 158°C. This is a wider margin between the two than I've seen in other heaters – perhaps they've done it to prevent “nuisance” failures?
While here, I'll note that Kambrook have made sure to connect the conductive parts of the heater assembly with separate rivets from those used to mount each part to the plastic frame; some cheapies I've seen aren't so careful about that, which may lead to connection failure if the plastic creeps under pressure.

Fan motor.JPG
As usual, specified to work up to 120°C even though the lead-ins are for 105°C. (Though that's not as bad as some others I've seen.)
Fan motor.JPG (139.28 KiB) Viewed 26234 times
The fan has just a basic 2-blade impeller, about 155mm diameter stamped out of 0.4mm steel with a black powder-coating (?). The motor is a conventional shaded-pole type, made by Shenzhen Zhaoli Motor (here's the nearest equivalent they list; they suggest a lifespan of 5,000–10,000 hours, although without stating the conditions).
Homemade bearing shield.JPG
(Plasticized PVC will be too soft; plastics other than PVC are most likely flammable unless specifically fire-retarded.)
Homemade bearing shield.JPG (79.23 KiB) Viewed 26234 times
I eventually (2018-12-05) made a dust shield for the back bearing, with a strip cut from PVC packaging. Hopefully this will extend the life a bit… (As of July 2019, bearing noise is still minimal. Both heating elements are still quite shiny and I expect them to stay that way for years to come. To be fair, I haven't given it a ton of use, but so far so good.)

Oscillation motor.JPG
Oscillation motor.JPG (225.46 KiB) Viewed 26234 times
The oscillating motor (also as usual) is a synchronous type with a speed of 2.5RPM at 50Hz/3RPM at 60Hz (1,200:1 reduction), and each back-and-forth cycle indeed takes 24 seconds (12 each way).

Functional testing
The fan noise level (after about a 5-second spin-up time) is roughly comparable to a Vornado 533 or 633 on medium speed. (Maybe slightly quieter, but at promoting heat exchange it falls well short of even their low speed. :silly:) So, consider the fan setting on this (or any other budget-oriented fan heater) as just a bonus; if you want a fan to circulate cool air, I'd get one of them (or if they're out of your budget, at least a larger plain fan).

As a heater, it does a perfectly competent job (not that there's a ton of room for difference there, given the fixed efficiency of any and all resistive heating elements); almost any small room will get quite toasty on full power, and even rather larger rooms are manageable (with decent insulation, or if you only need a modest temperature rise above the outdoors). The air coming out of the heater is lukewarm on half-heat, and on full-heat is quite warm but still not excessively hot. And despite that annoying “warning” tag, the thermal protection in this heater responds fine – in fact, it's the quickest-acting of any heater I've tested to date! (I've timed it within 1 second when fully covering either the front or back grille, on full-heat.)

The front grille remains quite rigid even on full-heat (which is more than I can say for some cheaper fan heaters).

Glow in dark.JPG
(If the presence of nearly red-hot wire elements bothers you, then PTC ceramic heaters are of course an alternative; though they do have some trade-offs… As long as nothing flammable gets INTO the heater, though, it's pretty safe.)
Glow in dark.JPG (58.11 KiB) Viewed 26234 times
There is a very faint red glow from the innermost section of the elements (inside the mica ring), but this isn't really noticeable in normal lighting (I took this photo at night, with the lights off).

Now for the grades:
Build quality: C. While it could be better, it could just as easily have been much worse. (Want to see inside a really cheap fan heater? I have two coming up…)
Heat distribution: B. Maybe not quite up there with Vornado's offerings, but better than most other heaters I've used.
Airflow (fan setting): C. While no Vornado, it's not too bad for a “budget” appliance.
Fan noise: D. Given the relatively fast fan, this was inevitable.
Warranty: D−. (1 year for “domestic” use, reduces to 3 months if used commercially)

In conclusion, while it's no miracle worker, you could very easily pay more for less.
Last edited by LongRunner on May 27th, 2020, 12:30 am, edited 12 times in total.
Reason: Fixed Klixon link
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

LongRunner's Mini-review Collection: Trent & Steele TS5

Postby LongRunner » October 21st, 2017, 4:05 am

Trent & Steele TS5 5-cup rice cooker (Australia/NZ) (a 10-cup version is also available)

If you eat rice on a regular basis, you probably don't want to keep buying the overpriced microwaveable cups and pouches (by one of my calculations earlier in the year, about 9 times more expensive than a cook-it-yourself equivalent from the same brand). (When making those comparisons, remember to allow for the fact that the microwave rice is pre-cooked, so already has the water absorbed into it; putting it in the microwave is merely re-heating it, which you can do just as well with home-cooked rice. For reference, properly cooked brown rice is roughly double the weight of the raw rice that went in, so about 50% water.)

Having bought this appliance last March, I might as well review it. I got it at Harvey Norman, where it retails for AU$29.95 (but mine may have been discounted, from memory). It's my first rice cooker, so I unfortunately can't provide any direct comparisons.

It looks much like any other basic rice cooker. Being at the cheaper end for a stainless-steel appliance, the casing (while acceptable) is not the most rigid.

Handle flash.jpg
(Trimming off those bits is one of my hobbies.)
Handle flash.jpg (91.02 KiB) Viewed 39296 times
I noticed a bit of flash on the handle(s); if you're like me, you can cut it off with a utility knife.

Both neons lit.jpg
(To light both indicators, I temporarily swapped two of the wires inside the unit – specifically, the blue and red leads to the PCB carrying them. This is as good as I could get them to look on-camera, but of course they're still more vivid in reality.)
Both neons lit.jpg (101.54 KiB) Viewed 39296 times
The front panel looks quite standard, with the two neon indicators (with a red filter in front of COOK, and yellow for WARM) and switch paddle below them (push down to start cooking, then it pops up when (nearly) finished; however, it won't latch if the internal button isn't pressed down). Strangely, the positions of the indicators have been swapped between my unit and that shown in the manual (and on their website).

On the left of the unit, below one of the handles, is a normal C14 inlet (though even the C6 would be fine at the given power rating), so you can use almost any cord. The included cord is 1.1m long, and is curiously 90°C rated (which I doubt actually matters much, given that the C14 inlet is only allowed to reach 70°C anyway), but otherwise nothing out of the ordinary. In any case, even 1.1m is longer than I need, so I divided an existing longer cord into a 1.0m (net) extension cord (which I use elsewhere in the home) and a 74cm cord to go with this. I might instead use the included cord with my PC (saving the longer cords for when they're needed)…
Oddly, the inlet is nowhere to be seen in the promotional photo.

Rating label.jpg
(That C-Tick looks a bit off…)
Rating label.jpg (79.37 KiB) Viewed 39296 times
Given the relatively low power rating of 400W @ 230V (scales to 436W @ 240V), you could safely operate up to 5 of these units through a standard 10A Australian power-board (or power-strip as they're called elsewhere).

Warning label.jpg
Warning label.jpg (71 KiB) Viewed 39296 times
The warning label on the inner wall (placed at a wonky angle, and noticeably left of center) has some mild Chinglish, so here it is if you want a quick laugh.

No exotic screws, fortunately – just plain Phillips #2 types. Once open, it looks like this:

Interior.jpg (152.19 KiB) Viewed 39296 times
The wiring is silicone insulated with a braided covering. The internal circuit could hardly be simpler for its function – in fact, here's a complete schematic:

TS5 circuit.png
TS5 circuit.png (25.47 KiB) Viewed 39296 times
The 40W element, for the record, is no fancy PTC – it's pretty much just a thin NiCr wire wrapped around a mica card, then sandwiched between two more mica cards for insulation and put into a steel casing (crimped shut, although it can be pried open if you're that curious to see inside).

Inlet bracket.jpg
(Setting the white balance with that wooden table in the background can be a real bitch…)
Inlet bracket.jpg (45.62 KiB) Viewed 39296 times
A simple plastic bracket is used for the inlet mounting. (I'd be a smidgen more comfortable if the bracket concealed the screw tips, but at least the wires don't have so much slack that they're likely to touch.)

Neon PCB (front).jpg
Neon PCB (front).jpg (89.23 KiB) Viewed 39296 times
Neon PCB (back).jpg
Neon PCB (back).jpg (92.58 KiB) Viewed 39296 times
A small PCB (mounted to the panel with a single screw, and two pegs for alignment) carries the neon indicators. Their series resistors are marked as 120kΩ (±5%), which may be a bit hard on the neon bulbs (4.8mm diameter by 11.2mm long) but if it's any consolation, the runtime of this appliance should be modest (not more than a few hundred hours per year, in regular use). The coating on the resistors is a bit poor, but they at least measure within tolerance. Inexplicably, the through-holes for the connecting wires are offset from the actual pads. :s

Thermal fuse.jpg
Thermal fuse.jpg (33.47 KiB) Viewed 39296 times
The set temperature for the thermal fuse of 192°C does seem very high, given that its only distant thermal connections to the heater are via a thin steel mounting bracket, and the wire from it to one heater terminal. Would be interesting to stick a thermocouple in there and see what temperature it actually reaches in normal operation…

Pinched heatshrink.jpg
Pinched heatshrink.jpg (19.64 KiB) Viewed 39296 times
Overheated heatshrink (1).jpg
Overheated heatshrink (1).jpg (38.38 KiB) Viewed 39296 times
Overheated heatshrink (2).jpg
Overheated heatshrink (2).jpg (29.08 KiB) Viewed 39296 times
Heat-shrink sleeving has been applied over the crimp connections, but this turns out to create two slight issues: Some of the eyelets were sleeved a bit too close to their business end, so the sleeving got pinched in the screw terminals (which I remedied by cutting off the excess). More noticeably, the main earth connection to the heater gets too hot for the sleeving (only rated to 125°C) to withstand, causing it to degrade in short order.

Earth connection to heater.jpg
This is where the sleeving deteriorated (I removed it before taking the picture, but it was present on all 3 terminals).
Earth connection to heater.jpg (47.58 KiB) Viewed 39296 times
I just got rid of those bits, but another thing bothers me: They shoved two wires into one of the crimp terminals. :dodgy: Were they really that desperate to save 1¢ – and if so, couldn't they do it in a more sensible manner, say by removing the considerable slack length of the wires? On that note, the eyelets themselves aren't the steel types recommended for high temperatures – just ordinary tinned (?) brass. I guess this isn't too horrible, though.

Also, two of the main metal parts have only indirect earth connections, but earth continuity is present so this isn't a total catastrophe.

Filling the cooker up is straightforward enough, not much to say there. (A plastic measuring cup is provided, but it's rather small and I haven't needed it.)

An annoying issue is that ricy water can spurt out of the steam vent (moreso with brown rice, which takes longer than white). One time I was testing it before, with the vent oriented to the left, it got onto the power connector; fortunately, it stopped short of the actual pins, but if it didn't, then it could trip the RCD (if it bridged active/line to earth) or burn the connector (if bridging active/line to neutral). :runaway: Wondering what to do about it, I eventually did a web search, and found a suggestion that adding a bit of cooking oil should help – and applying oil to the sides of the pot seems to do the trick. As for the cooked rice, I have no particular complaints.

Overall, I'll tentatively grade it a C; if you don't already own a rice cooker, this would seem like an OK one to try. Alternatively, the Kambrook KRC5 has the same 5-cup capacity and retails $7 cheaper (at Harvey Norman), in exchange for being white and having only a 1-year warranty.

Well, this post wouldn't be complete without it, so I've attached an OpenDocument (I'm not keen on perpetuating Microsoft's strangleholds any further than I can avoid) listing my results to date. I've also made a spreadsheet for the relevant calculations.
TS5 cooking data.odt
Currently at version 2
(33.16 KiB) Downloaded 855 times
TS5 cooking data calculator.ods
(If you want to use it for a different rice cooker, you may edit the equations accordingly – provided you understand the relevant mathematical relationships.)
(10.86 KiB) Downloaded 810 times
Last edited by LongRunner on December 16th, 2017, 10:51 pm, edited 1 time in total.
Reason: More data gathered
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

LongRunner's Mini-reviews: Cheapie fan heaters, part 1

Postby LongRunner » December 12th, 2017, 4:25 am

This time, I have two subjects. One of them is a Goldair GSFH110 (or FH-07A) bought 2016-07-02 (yeah, I slacked a bit :blush:), together with another 1.8kW ceramic fan heater (which is a bit higher priced, and noticeably better-built – though still far from perfect). The sticker price at Harvey Norman was AU$39, but a discount to AU$34 was given by the sales assistant. Already, it didn't make a good impression when the heater fell out of its own box while carrying it on the way home. :silly: The other is a Homemaker Everyday upright (without a visible identifying label), which looks suspiciously like this recalled model. It too is rated for about 2.0kW based on measured resistances. For the remainder of this post, I'll focus on the FH-07A (also sold under a few other brands).

It doesn't look like much, so I won't bother with an overview photo; but suffice to say that it's just a typical low-end plastic construction (although the front grille, at least, is steel). The slots in both the top (intake) and front (outflow) grilles are a little under 5mm wide. Weight is about 1.2kg IIRC.

Neon bezel.jpg
Neon bezel.jpg (27.13 KiB) Viewed 38554 times
The neon power indicator has a red bezel with vertical ridges, which isn't perfectly lined up with the casing but I can live with that. The switch and thermostat knobs (both identical in design) seem robust enough, although their shape is perhaps less than ideal (from an ergonomic standpoint).

Fan motor area.jpg
Fan motor area.jpg (115.77 KiB) Viewed 38554 times
Like most cheap plastic fan heaters, this one too is double-insulated…almost. There are meant to be two rubber plugs covering the motor mounting screws to provide supplementary insulation, but one of them fell out before I even got the unit (apparently). So strictly speaking, this is already a fail, but it's not the end of the world so I'll soldier on. You can also see a bit of melting (visible in more detail from the inside), I don't recall exactly when it happened but it's definitely bad. We don't get any non-slip silicone feet here, either.

Strange rating label.jpg
Strange rating label.jpg (39.64 KiB) Viewed 38554 times
A quick note on the power rating: 2000W nominal @ 240V (28.8R) should scale to ≈1681W at 220V; 1800W would be at ≈228V.
(I've only ever seen such inaccurate scaling on this cheap junk; after all it only takes a few seconds with a calculator or in a spreadsheet to do correctly.)

The right-most casing screw is a tri-wing type (as is quite common), just to try to block people from going inside (but I cheated it out anyway).

The cord seems OK, is 1.5m (net) of H05VV-F2X1.0 (in accordance with the Australian standard, and exceeding the IEC standards which allow 10A through 0.75mm² for 2m or shorter), surprisingly flexible for being attached to such a new (and cheap) appliance, and the plug kept its cool for the duration of the testing I did (and passed a quick pin-pull test).

Element glow.jpg
Element glow.jpg (81.07 KiB) Viewed 38554 times
Because of the uneven airflow, a part of the front element in this heater actually does glow a visible red. (The air temperature in this spot approaches a 150K rise :eek:.) Despite this, it still wouldn't ignite anything that I placed in front of it (short of poking it through the grille), so the thermal protection is at least somewhat functional. In fact, the trip temperature seems to be set slightly too low, as it cycled when I tried it on high heat (even with nothing covering it). :silly: On the low setting, it stays on.
Unlike the Kambrook heater, the protection in this unit only cuts off the heating elements; the fan is unaffected.


Interior.jpg (109.83 KiB) Viewed 38554 times
Breakage (1).jpg
Breakage (1).jpg (68.54 KiB) Viewed 38554 times
Breakage (2).jpg
Breakage (2).jpg (59.41 KiB) Viewed 38554 times
So, the casing is a real mess of cheap plastic (complete with two broken screw pillars), and unlikely to survive being stood on (certainly not by me) without breaking.

Motor underside.jpg
Notice also, melting of the case plastic (although the impeller is intact). I suppose that's what you get for housing a heater in ABS (?)…
Motor underside.jpg (138.62 KiB) Viewed 38554 times
Motor end bracket with flash.jpg
Motor end bracket with flash.jpg (33.79 KiB) Viewed 38554 times
The fan motor looks workable, although it is a sleeve bearing arrangement with the shaft vertical. It does have its own thermal fuse, as can be seen by turning it over after removing the impeller and the screws mounting it to the case. However, one of the (ZnAl?) die-cast brackets has a rather big bit of flash. :s

Most of the internal wires, from reading their markings, appear to be just the 90°C version of the individual 300/500V rated insulated cores (IEC 60227 type 08/H05V2-K) resembling those in the flexible cords; nothing special, but most likely fine given the modest temperatures they will actually experience in normal operation. The conductors are 1.0mm² for the common neutral to the elements and active/line up to the selector switch, and apparently 0.75mm² for the individual switched wires to the elements. Glass fibre braided sleeves (probably like Techflex type FGL or equivalent) are applied to the leads to the element, as well as to the individual cores from the cord (and quite oversized in diameter there). The exception is the leads from the fan motor which are UL AWM 1569 (300V, 105°C max.), and both are black as the motor doesn't care which way around they are connected. (I would have preferred XLPE or silicone insulated leads there, but what more do you expect in this age of cheapness…) The thinner red and blue wires to the neon lamp (held in place by hot-melt glue, which I then re-melted with my heat gun) are unmarked, and of a somewhat unusual (though seemingly still metric) size: 16 strands, measuring 0.14mm diameter (though given the limited precision of my calipers, they could equally well be 0.15mm which is a standard strand size used in some smaller metric conductors). 0.25mm²? The series resistor is 150kΩ (as usual, ±5%, carbon film, and 0.25W judging by size), which may be just OK for a high-brightness neon of the size used (4.8mm diameter by 11.6mm long).

Tip-over switch.jpg
(The rear-right foot doubles as the button.)
Tip-over switch.jpg (49.94 KiB) Viewed 38554 times
Thermostat.jpg (57.18 KiB) Viewed 38554 times
The tip-over switch, selector switch, and thermostat are all standard fare, and appear substantial enough for the current draw (ratings of respectively 16/16/10 amps @ 250VAC). (The selector switch is rather boring to look at, so I haven't included a photo of it.)

The resistance wires measure (using a caliper with 0.01mm precision) at just over 0.5mm diameter so about 24AWG – fair enough for the rating (4.17A nominal through each). Cold resistance measures 60–61Ω each (a little higher than the nominal calculated 57.6Ω from the given rating). The wire is attracted quite strongly to a magnet, so almost certainly FeCrAl again.

Thermal cut-out.jpg
Thermal cut-out.jpg (110.57 KiB) Viewed 38554 times
The bi-metal switch is set for 80°C, the thermal fuse (which I quickly blew, by jamming the fan for a test) for 110°C.

In this sample at least, it's the front-most element that's connected on the half-power setting. There is some tarnishing visible (as a blue tint) on the portion that got red-hot. At less-than-red-hot temperatures (say, in a heater with a stronger fan), these wires can stay shiny for many years (apart from burned remains of dust and other airborne particles that may accumulate in use).

Pinched wire.jpg
Pinched wire.jpg (38.6 KiB) Viewed 38554 times
Warped bracket (L).JPG
Warped bracket (L).JPG (110.21 KiB) Viewed 37064 times
Warped bracket (R).JPG
Warped bracket (R).JPG (130.44 KiB) Viewed 37064 times
Another unpleasant thing is the lack of care taken when assembling this heater: One wire was pinched, and both element assembly mounting brackets were bent out of shape (placing the elements closer to the grille than they should have been, and also putting the brackets up against the grille which might have created a second breach of the double-insulation standards).

Weird crimp.jpg
Weird crimp.jpg (50.77 KiB) Viewed 38554 times
They also did something bizarre with the connector to the neon+fan, inserting the conductors into the crimp from the wrong end and then folding the wires back before crimping the insulation. :wtf:

As you can also see, they used latching blade receptacles (which is reasonable enough). Unfortunately, a few of them are loose, which further cripples my (already weak) confidence in this heater.

So the grades are:
Build quality: F. Given the low standard, I've felt compelled to issue a recall of this model.
Airflow (fan setting): D. It can be felt close up, but it's unlikely to put up a fight with even a budget stand-alone (axial) fan, let alone a Vornado.
Fan noise: D. It isn't horrible in this regard, but is not quiet for the limited air movement, nor at a level "Silent PC" enthusiasts would be satisfied with.
Heat distribution: D. It may fare better in this regard than many air conditioners (including those installed in my house) as mounted high-up to prioritise cooling over heating (not by accident, I should add – especially here in Australia where the cooling part is all but mandatory in many areas), but it would still have to be rather close to where you most need the warmth, in order to reasonably utilise it.
Manual (although I couldn't be bothered photographing the whole thing for this post): D+. It does at least admit that "A correctly specified, undamaged extension cord may be used with this heater provided it is used in a safe manner", although it still makes the usual overblown fuss about coiling up the cord (which I've also tested out with the aid of another heater – rated for the full 2.4kW in Aus/NZ – and, methodically bound into 5~6 layers, a spare extension cord with the normal 1.0mm² cores still stabilises at just lukewarm to the touch even with a pillow beneath it) and tells us to keep "all objects" (their words) a full metre away from the unit (as I've again tested, also overkill for anything other than a radiant heater). One section even feels the need to remind us that "The Fan Heater generates heat during use" – yeah, thanks. :lol:
Warranty: D. (2 years for "normal domestic" use, reduces to 3 months elsewhere)


In New Zealand, two other brands have had to recall some heaters of this same model, and Goldair don't sell it anymore; while my unit doesn't have the date code shown there, it's shoddy enough in its own right that I really don't trust it any more.

I can't recommend these "flat" fan heaters anyway, even if you find one that's decently built. (Is it any wonder I haven't seen a stand-alone fan in this form?)
If the room is large enough for heat distribution to matter, I recommend an upright model with a normal axial fan (e.g. the Kambrook KFH6x0 examined above); alternatively, heaters using a tangential ("cross-flow") fan may work well, if available (I know of at least one vintage model).
If you want to heat just a small room and don't need great heat distribution, then I'd get a simple fanless convector heater like the De'Longhi HCM2030 (near-silent and more reliable). (Or, perhaps, get a convector with an independently switched "boost" fan.)

I ultimately gutted this heater for the parts (which themselves are more-or-less OK; it's the horrendous casing and slapdash assembly that ruin it). I did my own glow-wire ignition test on the pillar broken off the upper half of the casing, but it seems to be self-extinguishing (see here).
Last edited by LongRunner on April 10th, 2021, 6:15 am, edited 10 times in total.
Reason: Final amendments, added airflow temperature
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia

Projecta 200A booster cable

Postby LongRunner » March 19th, 2018, 3:21 am

Projecta SB200SP automotive booster cable

I've really scraped the bottom of the barrel this time. :blush: I got this at a Bunnings Warehouse, although they're available through other retailers as well. There is a fair range of cables provided by Projecta (going up to some fairly high-end stuff, it appears); but this is just the second-cheapest variant, rated for 200A, and ostensibly using 9mm² copper-clad aluminium wires: Not a good start, but to be expected at the price. There are other problems, though. Firstly, the length is claimed as 2.5m, but that's including the length of the clamps; if you subtract those, it's closer to 2.2m.

End connection.JPG
End connection.JPG (33.83 KiB) Viewed 36335 times
It appears that the wires have been folded back within the end crimp connections, although I don't know if this is a serious problem.

Electronic Box.JPG
(Okay, maybe calling this thing "electronic" is pushing it a bit.)
Electronic Box.JPG (34.17 KiB) Viewed 36335 times
My main point of concern is with this part. As you can see, it has a 3mm red LED, which illuminates (quite brightly) when connected to a battery with the correct polarity; I think that's a neat enough idea, but let's check inside the unit:

Components.JPG (49.58 KiB) Viewed 36335 times
So we have the LED with a series resistor (1.5k ±1%, metal film, size suggests 0.25W but it better be a mini 0.5W to properly handle 24V use), and a MOV set for 39V (which is correct for the given voltage range). I would want to see a diode to prevent the LED from receiving reverse voltage, though (LEDs based on GaAs or GaP and related materials will usually survive reverse voltage provided the current is limited, but I still don't think it's the best practice).

(There was a bit of hot glue between the MOV and case, but it was covering the value so I had to remove it.)

Components removed.JPG
Components removed.JPG (22.43 KiB) Viewed 36335 times
The way the components connect to the cable is very crude – two pins just more-or-less "hook" into the wires. Worse still, the case provides no meaningful strain relief for the cable, so the cable (which is quite thick, remember) can slide back-and-forth and stress the components.

Soldered in.JPG
Maybe not perfect, but it should be better than no solder.
Soldered in.JPG (50.06 KiB) Viewed 36335 times
So when putting it back together, I soldered the "hook" connections. (Not shown in the photo, I also added a small piece of PVC tape to hopefully provide a little extra protection against an internal short-circuit.)

In conclusion:
Fail-200px.png (11.15 KiB) Viewed 34951 times
At least it worked (my sister's car needed the boost), so hasn't gone totally to waste. I'd hope the higher-rated cable offerings are better than this, but wouldn't know short of getting them to examine (not likely to happen in a hurry)…
Last edited by LongRunner on April 30th, 2018, 7:09 am, edited 1 time in total.
Reason: Added Fail award
Authoritarianism is for wimps.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, 2 * WD3003FZEX-00Z4SA0, Pioneer BDR-209DBKS, Seasonic G-360, Chenbro PC31031, Windows 7 Pro (though I do want to build a second system with GNU/Linux).
Posts: 880
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia


Return to Off-Topic

Who is online

Users browsing this forum: No registered users and 11 guests