I appreciate the shielding Apple added, especially with two resonators in the HF band. The shielding is, BTW, probably to prevent the device from causing interference, not to prevent other devices from interfering with it. (It's not a radio, after all.)
Note that if you do choose to buy the cheaper one without any shielding, it's your responsibility (in the US) to prevent it from interfering with licensed users of the RF spectrum. (In this case, 25m shortwave broadcasters on 12MHz and government-run time-transfer services, like WWV and WWVH, on 25MHz.)
If your device is particularly annoying to some shortwave listener, expect to get a letter from the FCC telling you to shut the thing off.
(I looked up a few of the letters, and I really appreciate the politeness from the government: http://www.arrl.org/files/file/Part%2015%20Letters/04-06-24-... "There are
several simple and inexpensive steps that can be taken to eliminate interference from battery chargers, and we would be glad to send them to you if needed.")
The complainant even provided the fix (Toroid cores, often seen as those lumps on USB cables) for free, installed and everything. In response the emitting neighbor went and...
"You apparently discarded the toroid devices, tossing them in Mr. (deleted) yard. The devices were simple and non-intrusive, and we can envision no reason whatsoever for you not to continue using them, other than a desire to be uncooperative."
Dabbling in RTL-SDR for a bit quickly brought me up against the interference monster, and since the USB sticks are typically powered by bus if no direct power modifications are done one of those sources of interference was hooked right into the receiver. Toroid cores either clipped on or looped were indeed the answer for quieting the cable. This doesn't help much if the interfering device is emitting really badly off the board itself though so manufactures shielding their devices as a standard is rather welcome.
According to Bunnie's investigation into microSD cards[1], that low serial number probably indicates that the components were produced on a "ghost shift", when a rogue worker comes in at night and runs the plant off the books
I'm equally impressed by the Apple adapter. They make expensive stuff, but they legitimately don't cut corners. It's nice to see there's still one consumer electronics company that still does that.
What is even more horrible is how people think they can buy a PC for cheaper than the price of its parts. I am thinking commoditization of the entire device is flawed while commoditization of parts is not.
I guess it's nice to see that they put some thought into it but it still seems, well... over-engineered. If you are trying to be the monolith that your customers buy EVERYTHING from then you should be a little more realistic... nobody wants to shell out an extra $25 for a fancier chip with a little shielding.
It bugs me that Apple customers seem to get ripped off by their brand for simple things & then they thank them for it. I'm glad a 3rd party market for Apple stuff finally emerged. I remember the dark ages when you couldn't even tell your friend "check ebay I'm sure theres one of those..."
Part of the reason I am still in the PC world vs Mac (I build my own powerhouse desktops for research data-mining & gaming, buy cheap laptops for practical use) is that the multitude of companies making parts means they find the proper price/quality balance much more often.
I'm glad this author gives credit to the off-brand. I'm very impressed with the quality of cheap Chinese electronics, and if it breaks you have 8 or so more tries to get it right before you reach the same amount spent. & plus, let's face it the main problem with this stuff is that you lose it, not that it breaks so I have a tendency to sway way toward the cheaper end with non-essential cabling & components for that reason.
Inside my research PCs is a different story... no expense spared on CPU/GPU/drives/mobo/etc. :D
Very few UL power adapters are as good as the ones Apple makes. I ran a bunch of my USB charger "cubes" at medium load and looked at their output on my o-scope one time. Most of the UL listed adapters had way more noise than the Apple one, even at just 500mA. There was one other manufacturer's charger that was as good as the Apple charger, but I can't remember off the top of my head who it was... I think it might have been HP or Asus or something.
The lack of shielding is rarely any problem. Two oscillators is not better than one, since PLLs can generate the required clocks (480MHz = 25 / 5 x 96, 100MHz = 25 x 4, 125MHz = 25 x 5); SMSC has some USB/Ethernet solutions that use the same 25MHz frequency.
> The Windows drivers are the exact same digitally signed ones that Microsoft distributes through Windows Update
AFAIK there is a standard device class for "USB ethernet controllers", so any chipset that conforms to it will work fine with the standard drivers.
As for the unmarked chip - many IC companies are not averse to creating custom designs for a specific customer (and marking it however you want) if you're willing to buy enough. I don't think it's ASIX since they don't have the single 25MHz clock source; more willing to bet on Microchip or SMSC.
You're spot on about the PLL stage, a few people on Reddit pointed the same thing out when I first posted it. Makes sense, although I still think it's interesting they added it internally to an otherwise (apparently) cloned design.
You're also right that USB CDC does provide the option for a generic USB Ethernet device, however this silicon is ASIX-specific (not just the USB IDs.) ASIX's Windows drivers include their own system driver binaries, and the ASIX Linux driver has a lot of ASIX-specific stuff in it.
I think it's kind of possible ASIX made this themselves as some kind of no-name branded unadvertised market segmentation effort. I can't understand what their rationale would be exactly but hardware companies do unusual things sometimes...
If you go only on the definition of a clone being "compatible interface", then there are tons of other examples of that in the electronics industry - it's more commonly known for simpler parts like voltage regulators (how many companies make a '7805?), opamps, transistors, etc. but also occurs with more complex ones too.
IC companies make unadvertised products all the time, for anyone who is willing to buy enough... look at Apple's Lightning cable and TI's BQ2025, for instance.
> I don't think it's ASIX since they don't have the single 25MHz clock source; more willing to bet on Microchip or SMSC.
Note that it identifies itself to the system as ASIX, though. Of course, it could be that other companies actually have a license to make ASIX things...
USB descriptors are programmed in the EEPROM, so they could've made it identify itself as anything they wanted. If you can dump the EEPROM, that can provide more clues - the configuration format does differ between manufacturers.
I recently purchased a Gigabit ethernet adapter for my rMBP. If you're fine with something a bit larger and are willing to go above "dirt cheap" I'd strongly recommend the Inateck HBU3VL3-4. It's made by a German company so I'm not sure how available it is over in the USA.
It has an ASIX AX88179 which is a USB 3.0 to Gigabit chip and it my tests I was easily able to get around 800-900Mbps. It also has 3x USB 3.0 ports, so not only do you not lose your USB 3.0 port you gain 2 more. I've not dismantled it but it feels solidly built.
Here in the UK I paid the same price as the Apple USB to Ethernet Adapter, which is only 100BASE-T and offers no hub.
Fantastic, I'm in the market for a USB gigabit ethernet adapter as the thunderbolt one from Apple (and all other thunderbolt adapters it would seem) disable the sleep on windows. I found this on Amazon in the US under the brand "Unitek."
We have used a similar dongle at our company for a 2nd ethernet adapter on our Foxconn-built boxes and found out that they usually don't have unique MAC addresses. This made our lives very difficult because Cobranet audio network uses level-2 addressing. We eventually found a supplier with real MAC addresses.
You know it is usually pretty trivial to assign whatever MAC you want to them.
(I have an unbranded Android phone that assigns a random MAC address every time I turn WiFi on/off. Apparently this happens since they didn't bother to initialise the EEPROM and just left it at all zeroes. But this means being harder to track, so I consider it more a feature than a bug...)
To clarify, my interest in whether the IMEI is zeroed and/or programmable is about how the various mobile networks use IMEI, particularly: - China Unicom uses them to detect continued use of subsidized handsets on their network (and will send the customer a text if they're not using it (eg if they've sold their subsidized phone) - I believe T-mobile USA use IMEI to limit access to their 'free data for life' promotion to specific tablet models - Several UK mobile networks have different data tariffs for phones vs. tablets vs. dongles vs. mi-fi devices
I'm wondering whether there is any/widespread use of these android devices to bypass restrictions and, eg use a cheaper tablet data plan on a mobile phone with tethering to multiple devices.
Were you able to get reliable performance sending CobraNet over USB Ethernet adapters? I once wrote a simple audio recording program for the Linux console for diagnosing CobraNet issues, but haven't tried it over a USB adapter.
An old analog hacker rule of thumb is that if your product is going to emit RF, you'll find spurious RF on input and output terminals (including ground terminals), which you can check with a scope. Such a pre-test is cheaper than getting a field test done and failing.
The scope test might be do-able on a comparison basis between the two designs. It's possible that Apple over-engineered their shielding.
It's not necessarily a cloned design. The cheap part may simply implement standard protocols, and use the name of the Apple part to ensure driver compatibility without too much testing. Still, that seems rather under-handed if it's what it seems.
OP isn't alone - I find these parts dissections fascinating as well. Posts like these show you how and why hardware is becoming a comodity. It's going to be a very interesting century.
>I’d love to learn more about these secretive industries and the engineers who work in them.
Isn't it pretty well known that 1) they aren't secretive and 2) they don't actually design their own hardware. They start off by producing the legitimate item, in the factory that's contracted by the company who designed the product. If the product is successful or easily fits in their existing knockoff fabs, they plans are basically copied over
They aren't secretive, but they are hungry, and they are foolish.
I'd venture the shanzhai chinese geeks fiddling relentlessly with hardware in Guangzhou are much closer to Steve Jobs' ideal than the average Apple Store clients.
On one side you have rich kids who can afford a $30 dongle, knowing perfectly (after reading this article) that they are giving away $25 to a marketing department and that they are bowing to their own voodooish cargo cult. These guys are simply afraid of the remote possibility of having to spend 5 mn of their precious time checking a connection, and maybe too, something really catastrophic! lose ONE of their 10.000 instagram pizza photos because of a subpar component in their setup.
On the other side you have these ugly hackers who are photocopying hardware components for 1 year old companies named Yunxingdianziguangzhoujituanyouxiangongsi, and their logo is ugly, and these guys are Chinese! I know them: they are hungry, not only hungry for money, but also hungry for knowledge, recognition, they are curious about the world, about the future, and they have no fear for bad taste: a pink iPhone clone with teddy bear shape? check!
It's a poor analogy for many reasons, not the least of which is to SJ:
- Taste mattered a great deal - he definitely did have a "fear for bad taste"
- Things being perfect and working perfectly out of the box without "5 mn of their precious time checking a connection" mattered a great deal
- User experience mattered more than price
- SJ wasn't interested in commodities
Also I don't know why you would refer to the guys as "ugly" and why you exclaim that "and these guys are Chinese!" There's something subtly racial about your view in the whole post. Would it be any less of an exclamation if they were Peruvian?
Wow, you got it wrong: it must be my mistake, I was a bit too assuming on the context.
- For me being "foolish", in the mouth of Steve Jobs, means not caring at all about what other people will say or think. I see a link with the peculiar "bad taste" of shanzhai: they do not care what "good taste" people think of pink Teddy bear shaped phones.
- I am living in China since 10 years, my wife is Chinese, my kids are half-Chinese, all my colleagues and many of my friends are Chinese. My comment was not "subtly racial", it was (trying to be) ironically emphasising the fact that in some people's disdain of shanzhai cheap gadgets, there is also the fact that those who build them are Chinese.
Just like 30 years ago saying a car was Japanese meant something cheap, copied, second-grade, right now the same applies to Chinese manufacturing.
And, in fact, often justifiably so! There are three kind of products quality made in China:
1) The very cheap that is sold in developping countries (have you ever seen those electric plugs with a single hair of copper inside for conductivity? cost nothing, burns after 10mn use).
2) The high quality sold to developed countries with Western brands. E.g. the Apple usb dongle in question here.
3) The medium quality sold internally and in developed countries as cheap alternatives to 2) above.
By the way, this 3) class is the most interesting because as the Chinese middle class raise its income level, it will have to become better and better, until a Chinese brand will do better than what is imitated currently (just as Japan did with cars, hifi, etc.).
I hear often that Chinese people cannot produce quality because of some of their "cultural atavism", and I think this is very wrong. In the past, they were manufacturing the best of breed in many domains, and in all their region of influence having "北京" (Beijing) stamped on goods was enough to triple its price.
While that does happen, not all knockoffs are just counterfeits made to the approximate specs of the original. There are a ton of knockoffs that have radically different internals that required significant engineering effort to design.
Well, maybe. The original has 47 power pins, more than the counterfeit has any pins. It's likely that the original designers included these to improve noise margins in various parts of the chip. It's also likely that the 32-pin version has rather worse problems with power and ground noise.
It _is_ possible, of course, that the original has excessive margins and the counterfeit one is a tighter design. We'd have to see the eye diagrams on both sides to be sure.
But all of that speculation is not comparable to actually seeing the clone work in real-world situations.
I would've liked to see the OP do a benchmark on transfer rates between the original and the clone. (Unless he aleady did one and didn't find any difference.)
"When I ran a TCP throughput test with iperf, they both performed well. The Apple adapter measured throughput of 94.3Mbps. The cheap adapter measured 87.4Mbps. By comparison, the builtin ethernet on my laptop measured 94.8Mbps (after being set from gigabit to 100Mbps.)"
So, there is a difference, but it's relatively small (and possibly acceptable for the things people use ultra-cheap ethernet adaptors for).
I would've liked to see a UDP based packet loss report, and also at a couple of different packet sizes.
The throughput drop could be due to packet loss, maybe from more noisy circuitry, or it could be due to a bottleneck on the USB side of things. Smaller packets are good at finding those because they generate more events per unit time.
I saw this with a bunch of crappy WiFi dongles. The radio can run at the 54Mbps or whatever but the USB interface tops out at 20.
It might work well when attached to another device with high (i.e. industry-normal) noise margins. I'd be quite interested to see the performance of two adapters connected together, or connected with a cable wrapped 10 times around a pencil.
Sorry, didn't see that (just dived right into the teardown. :) So it's a ~7% difference, indicating that it's not a straight clone but different silicon with a compatible interface.
This part pretty clearly isn't a knockoff, though; it's a version made cheap almost to the limits of practicality (or maybe beyond, if you care about producing radio interference) through design changes.
We use a ton of USB adapters where I work and are using them all day long with several different systems. A few months ago we got a batch of cheap adapters and one noticeable difference was the cheap ones get really hot compared to the more expensive name brand adapters. Appear to work OK, though.
If so, why don't they make an exact clone of the Apple one at $6 cost and sell it at $15, or even $10? The identical thing but at much better price. Win-win.
Because nobody will buy them because there is no way you can convey the fact that being an exact clone gives more "quality" than being a cheap knockoff. It's already a hard sell a $300 android phone from a $100 android phone, though you would say there is infinitely more differences between the two, compared to two usb-ethernet adapters.
Some people will buy Apple just because it's Apple and they want "original" stuff; others will buy Apple because they read these articles and know that the parts are high quality; others will buy the cheapest one they can find among amazon, ebay, monoprice and korean websites (probably laughing at those who buy Apple parts).
I don't think there's an easy market for "non-Apple and midprice".
I'm in this business. It's really tough to market based on quality without having a big name, but it's always easy to market based on price. So a few companies offer quality parts at high prices while hundreds of smaller companies fight it out for the low price.
That, unfortunately, probably wouldn't work well. It'd be hard to convey that it's better enough to justify the price difference to the buyer. Also, if you're doing that, why not just price at $10, _claim_ to be better quality, but clone the clone instead of the Apple one? The average consumer has no way to check, and good luck getting a refund on this sort of thing.
If you look at the vendors on AliBaba/AliExpress (where the prices are usually too good to be true anyway), the AX88772ALF chip alone goes for between $4 and $14 USD with a median of $8.
The taobao link in the original post leads to a listing which does not specify the chipset or support for Mac. This suggests it's a cheaper chipset. The 4-5 best-selling cheap USB-to-Ethernet adapters on taobao use the RD9700 chipset, and can be had for 22RMB including shipping.
The cheapest Asix 88772-based one I can find from a reliable seller is 25RMB including shipping.
The RD9700 is "USB 2.0" and what they don't tell you is that it's actallly "USB 2.0 (FS)", i.e. the 12Mbps full speed mode.
It also comes in a 32-pin package, and comes from a shady-looking company (http://www.corechip-sz.com/ ) that has parts numbered 6872 and 6873, so the "RD6877" on the PCB of OP's clone suggests it might be a RD6877.
This is interesting - and more importantly, is there any part of the ethernet standard they don't support? I only ask, as at my university there were a number of networks that students simply couldn't connect to if they used any ethernet adaptors. Oddly enough, thunderbolt adaptors, and ethernet on the motherboard worked fine, just USB adaptors didn't work.
There have been times in the past when I've found almost every USB ethernet adapter available on the market to be of a single chipset. If one very common chip at that time had a poor interaction with your school's switches, boom.
Most of the time, the cheap no-name electronics I've encountered have been pretty feature-complete, or even feature-laden, and any bugs were usually minor annoyances rather than fatal flaws.
But the quality control is a crapshoot, so a good chunk of them tend to be DOA. Of the ones that work fine at the start, many will fail in the following weeks, months, and years at much higher rates than you'd expect.
The problems I have run into with the cheap versions of hardware like this is that they tend to fall apart. For example, my last mini-DVI to HDMI adapter performed just fine... until the glue holding the female end in broke loose. I glued it back, and then the plastic started to crack.
The guts of it (the electrical bits) looked fine, worked properly, etc. It was the mechanical design that gave me problems. Replaced it with a (ridiculously expensive at $29.99) higher end version and that has lasted me quite some time.
Well, note that the cheap one is somewhat slower in the iPerf test (though, for the sorts of things people use ultra-cheap ethernet adaptors for, this would rarely matter). Also, while I'm not sure if it's the case for this one, a lot of ultra-cheap network hardware has the same MAC address for _every device in existence_, which can cause entertaining issues.
Note that if you do choose to buy the cheaper one without any shielding, it's your responsibility (in the US) to prevent it from interfering with licensed users of the RF spectrum. (In this case, 25m shortwave broadcasters on 12MHz and government-run time-transfer services, like WWV and WWVH, on 25MHz.)
If your device is particularly annoying to some shortwave listener, expect to get a letter from the FCC telling you to shut the thing off.
(I looked up a few of the letters, and I really appreciate the politeness from the government: http://www.arrl.org/files/file/Part%2015%20Letters/04-06-24-... "There are several simple and inexpensive steps that can be taken to eliminate interference from battery chargers, and we would be glad to send them to you if needed.")