An Engineer In Shenzhen

Initially, I was very excited about adding wireless capabilities to my product, and I was even more excited to find a plethora of available solutions in China. I could get any kind of remote, receiver, transceiver, encoder or decoder, FSK, OOK, ASK, in any frequency band 315MHz, 433MHz, 900MHz, 2.4GHz, at any power level. I found someone who had a remote with six buttons, which was exactly the kind of remote I needed. They had a matching receiver module, which had the decoder for handling the transmitted signal. I was even able to hook the receiver up to my product without any changes to the microcontroller firmware; it integrated directly into the same circuit as the rest of the buttons. And I contacted the manufacturer of the remote and was told I would even be able to customize the printing on the buttons and change the color of the remote. It was a perfect solution.

Well, it was a perfect solution for a prototype. The realities sank in and caused me a lot of delays. Here are some important notes on building wireless products if you are going to sell them in the United States, Canada, or Europe.

In the United States, the governing body for wireless products is the FCC. In Canada it’s the IC, and in Europe, it’s ETSI. Each has slightly different rules, but in general passing FCC means you should be ok anywhere else, and each region wants imported products to follow those rules.

Circuit

• Antenna design is a black magic that few people understand, and I’m one of the ignorant many. To get a consistent and optimized range and performance out of the transmitter/receiver pair is very difficult, and requires finely tuned and precisely valued components. This is not likely in extremely cheap products, and the range and bandwidth suffers. You will need to consult with an expert in wireless technologies in order to optimize your solution for your product.
• You may be transmitting at a certain frequency, but depending on the circuit, you are probably also transmitting at harmonic frequencies or other spurious frequencies. This happens more the cheaper the parts are. Your intentional frequency power levels may be below limits, but transmissions outside of the band must be at significantly lower power levels, and it’s easy to violate those.
• Certification is done using the exact parameters of your final product. You don’t want to submit to a certification house anything but the final version of your board, because any change to it can significantly affect the output of the antenna, leading to violations of power levels or harmonics. Any change during production requires recertification as well, so you want to make sure you are in complete control and confidence of your factory. If they change a part, it can put you out of compliance.

Regulations

• The FCC has a lot to say about intentional transmitters. First, most frequencies are off limits, except for a few called ISM or unlicensed bands. If you will be doing international sales, the most consistently available frequencies are 900MHz and 2.4GHz. These frequencies are unlicensed, meaning you don’t need a license to operate devices at those frequencies, but they are NOT unregulated. Think of it like a city playground. Everybody is allowed to use it, but if you start throwing sand at people, you’re going to get in trouble.
• The FCC wants you to not throw sand in other people’s faces. They’ll let hobbyists do their own thing within certain limits, but anyone developing a product for sale has a lot of rules to follow. This means filling out some paperwork if you are developing a product that says you will be transmitting at certain frequencies and powers from certain locations. It also means no beta test units going out to early customers. You can’t sell, you can’t even market, and you can barely develop a product if you don’t have FCC certification for transmitters.
• In order to make the FCC happy, you have to take your transmitter to an approved testing house which has a special anechoic chamber used to measure transmission levels at all frequencies. The product is put in this chamber, and powered on and put in a test mode to continuously transmit, and then rotated around as the meter gathers data about what emissions are coming out of it. This typically takes a few hours, after which the testing house will compile a report which is then filed.

An anechoic chamber for testing RF emissions. (Image courtesy of Adamantios. CC)

China Specific

• Very few Chinese products have any kind of certification. It costs too much for the factory to pursue it, and then they’d have to amortize the cost into the cost of the product. Wasted if the product is being developed for China or any country that doesn’t care about FCC.
• If you want to pay for the certification, you risk that your factory will at some point change a part and put you out of spec and make your certification invalid.
• If they do have certifications, they are sometimes fake. It’s helpful to look up the cert in the FCCs databases. The neat thing about the FCC database is the submitted documents sometimes have the exact part list, schematic, and board layout of the thing being tested. It’s easy to tell if the FCC number being used is not for the product being claimed.

The Good News

It’s not all FCC misery, though. The good news is that there are companies that develop modules that have modular FCC approval. This means you can drop these modules into your design and they have already passed FCC testing and you don’t need to go through that process. Naturally these modules are more expensive, but in lower volumes and faster time to market, it’s not a bad solution. There are lots of modules for any purpose, so it’s not too hard to find one that will do what you need, without having to design your own and fight through the black magic and bureaucracy of the FCC and wireless antenna design.

China is hazardous to your health. A good friend of mine with the startup Nomiku recently took ill while working on production in Shenzhen and had this to say:

Ugh, just found out the meds I ate have lead in them from the doctor I visited today. China got me good, once again.

There are many kinds of ways China will kill your body. Let’s examine a few.

Traveler Issues

These are the problems that aren’t specific to China but rather are likely to befall anyone who travels to a foreign country. There’s diarrhea, flu, jet lag, and adjustment to climate (such as going from a colder region to the tropics). It’s hard to blame China specifically for these issues, but they do need to be mentioned.

Bad Food

Some food is fantastic. China definitely knows how to do things right if you happen upon the right restaurants. Unfortunately, sanitation standards aren’t very high, and it’s common to get food poisoning from street vendors. But it gets worse than that. The fakes about not just in electronics but in food as well. The recent scandals on gutter oil and fake pig ears have rocked China and disgusted the world, and those are the ones that got caught. There’s really no way of knowing your food is safe, so you just hope it’s been cooked long enough and your gut has the fortitude to take it.

Getting Run Over

This merits a whole heading. With motorcycles driving on the sidewalks, running lights constantly and driving in the wrong direction, and in general the whole of China only having a couple decades of driving experience, I sometimes wonder if I should be wearing a helmet while walking. The only safe place is the subway, because the trains have the least likelihood of being where the pedestrians are.

Environment

Manufacturing takes a toll on the environment, and in China there’s a lot of it. On many days it’s not unusual to only be able to see a block or two in the big cities. That’s a lot of particulate matter in your lungs, going to town on your respiratory system. And the awesome part is that there’s nowhere you can go to escape it. There’s no ‘breath of fresh air’ to be had. Running is harder, stairs are harder, and your lifespan is probably shorter every day you are there. On the plus side, it doesn’t feel like such a big city when you can only see a bubble two blocks wide.

SRSLY?

I couldn’t think of a better name for this category. Besides watching out for the things that seem almost understandable, there’s a whole category of dangers that make no sense at all and make you wonder if China has some great conspiracy and everyone is involved. A friend had a massage, and it turned out that having ears cleaned is a common step in this service. Instead of using soft tipped things, they use a pointy stick, and they go deep. So deep that she was bleeding from her ears and a trip to the hospital revealed that her ear was scratched, but her eardrum was not damaged. Eastern medicine is another category of strange practices that are probably more placebo than solution because they didn’t work for a skeptic like me. The motivation behind this post is my friend being prescribed pills with lead. There may be some legitimate reason, in the same way that chemotherapy is legitimized, but I’m not quite ready to give China the benefit of the doubt here.

Hacks abound, too. The kind of hack that has high voltage lines taped to walls and heavy objects suspended precariously. It may work for now, but everything you see or use is a disaster waiting to happen.

Conclusion

One of the purposes of this blog is to inform engineers traveling to China about the things they are likely to expect. To summarize, expect to get sick, and expect to get hurt by something completely unexpected. China can mess you up, and will be unapologetic doing so.

China is fraught with fake parts, but that’s not always a completely bad thing that will ruin you. Sometimes it’s something that will save you lots of money because it works well enough for your needs and is readily available.

Two of the components for my product (the DigiTally, a portable electronic scoreboard) are DC/DC voltage converters (one at fixed 3.3V for the microcontroller, and one at 9V for the LEDs). The parts are pretty simple and do the job well. They are made by National Semiconductor and readily available from Mouser at $2.70 for 1 and $1.17 for 2500. This is for the LM2575 3.3V SMT part. In Shenzhen, I asked for an LM2575 SMT component, and they rushed off to find my part, returning shortly with it and charging me only about $0.40 for each.

The component was pin compatible, with the same footprint as the National part. It worked exactly the same. I didn’t need to modify my design, and I had managed an enormous savings off the cost of my product. It was awesome.

But there was some cost. There are no datasheets for this part. I can’t even figure out what company made it. I looked around for datasheets for other variations of this part and found that there are a few, and that they sometimes have slight changes in the recommended values of nearby attached components. This is slightly disturbing because it means I may not be using the right values for components, but I have no way of knowing. It is essentially a small black box that does what it is supposed to as far as I can tell, but I don’t have the information to back it up, meaning there is some risk in my design. I don’t even know if these are legitimate parts or if they are rejected defects or expired. It’s a gamble.

The other challenge is finding these parts in the first place. I got lucky (maybe) to find this component, but what about my other components? There is no way of knowing if there are other parts that can be legitimate cheaper alternatives to the name-brand components I’m using now. There is no Mouser or Digi-Key for Chinese manufacturers.

I’ve found that cheap resistors abound. I can get SMD resistors of any value for essentially free ($0.002 each). But when I ask for name brand parts, the price jumps to comparable to, or sometimes even higher than, Mouser.

There are two lessons in this. First, that China can be a source of extremely cheap parts and significant cost savings on components. But second, that because it is so hard to verify the provenance of these parts and their specifications, there is some risk in using them, and it may be better for critical components to use name-brand parts with clear datasheets and support.

Sparkfun has an interesting article on something they call, the “pit of despair.” It’s the point of volume production where you’re more than you can make easily at home, less than you can go directly to a contract manufacturer, and right at the point where you’re not making enough money to quit your day job but you’ve committed yourself to a lot of work. It’s an awful place to be.

The problem with small hardware startups is that it’s all too easy for them to get into this space. They start off making prototype units, selling them at a loss or tiny profit. Then as they scale up with glimmers of mass production in their eyes they hit this pit along the way, where they can’t afford to get a factory and their volumes aren’t high enough for one anyway. They’re stuck, and quality suffers, and there’s no time to do development to improve the product. The only way out of the pit is with a large infusion of cash or popularity (which brings with it cash).

At a soldering party at the Sector67 hackerspace. The Pit of Despair enjoys company.

From the article:

A surprising amount of products fall into the third bucket. I call this bucket the pit of despair. These products get far more attention than the creator(s) expected. The product was well-designed but may have been designed for kitchen production, where they planned to build tens of units. When demand grows beyond thousands, the creators are often forced to make tough decisions: the income may not be enough to release them from their day job, and the amount of time required to build the product begins to gobble up evenings and entire weekends.

Read the full story at Sparkfun: Pit of Despair

Occasionally I get contacted by other people in or near Shenzhen who have also written about the craziness that is the Shenzhen electronics market. Here’s one such post. Mitch Davis has an interesting perspective about forming lasting relationships with people and how important it is to get a foot in the door and build a long term partnership with suppliers that will save more money in the long run than shopping around for the best price at that moment.

Sellers generally only want to do business if you’ve done business with them before, so it’s a chicken-and-egg situation. The reason is the Chinese concept of 关系 (guānxì, loosely pronounced as “gwunshee”, and even more loosely translated as “relationship”). …

One way to overcome this is to get an introduction from someone already doing business with them. The sociological theory is that neither side is unlikely to do anything stupid (like cheat) to the other, because it would reflect poorly on the introducing friend. And to put someone in a situation which makes them look bad is something you just don’t do in Chinese culture.

Read the full article for an interesting story of bootstrapping those relationships.

Source: Hackvana

I met Ian of DangerousPrototypes and Tully and Liao in Shenzhen, and just discovered they have a video tour of the electronics market that is much better than the coverage I gave of the place. This video gives a good sense of what the market is like.

And here’s the link to his article.