The same day we went to the sanitary napkin factory, we also went to two other factories. I broke them up into two parts because there were too many factories to write about and it would have turned into a huge post.
After lunch we headed to the Okano PCB Assmembly house. Okano is a joint venture between Okano in Taiwan and AQS so it was easy to set that tour up. Unfortunately, Okano didn't want pictures being taken inside the factory because one of their large customers was Nintendo and they didn't want the PCB pictures and assembly process for them to get leaked on to the internet. I was only able to take pictures of the initial setup to go into the factory. By now, we're all pretty used to the gear needed to go into an SMT assembly factory. All PCB assembly houses are paranoid about ESD since they result in soft failures that are difficult to diagnose. For Okano, we had to get dressed up in the standard ESD frocks, hair nets, and shoe condoms. This was the first picture I got of all of us geared up to go into an assembly house though since Huawei wouldn't allow cameras.
The other SMT assembly line that everyone saw was at Huawei and it was a world class SMT line. That was actually not so good because they couldn't go in and check things out close up since a line stoppage would have resulted in a large loss for the manufacturer. At Okano, it was much better because everyone was able to go in and get their hands dirty checking out the SMT process up close. SMT lines have common sets of components. The standard pattern is to have the solder paste applicator, automated visual inspection of solder paste thickness/coverage, pick and place, AOI (automated optical inspection), reflow, and final AOI. Okano also had an intermediate step which was a manual component insertion station for components that couldn't be inserted via pick and place. This would definintely not have been tolerated on the Huawei line because of the volumes they do, however it's a reality of design, especially for a PCB assembly house that does not control the design.
The main area of interest for me was the kinds of pick and place machines that Okano was using. They were using nice but slightly old Panasonic automated pick and place machines. From my estimate, they had about 8 separate lines, mostly using the same machines, although some lines had more complex equipment. I really liked watching their chipshooter, which is a specialized pick and place that is just used to "shoot" passives on to the boards. These usually require less accuracy so a chipshooter would just load up on a bunch of passives and do rapid fire machine gun type chipshooting without having to return to the base to pick up additional passive components.
I was talking to the engineer that was showing us the lines and noticed that they had idle machines in the corner that weren't in use. I asked him if the machines were still operational and he said they were, but they weren't being used at the moment. He then said that if I was interested in them, I could possibly purchase them from Okano. Bunnie suggested I rent some space in Shenzhen, purchase equipment and make sure it works. If it needs work, then its easy to get repair people to fix any problems, as well as get extra feeders. Once it's operational, then I can just ship it out to myself in Tokyo. Unbeknownst to Bunnie, this triggered a weird fantasy like dream. I suddenly had bizarre ideas of starting a hacker house in Shenzhen where the first floor was a hackerspace, 2nd floor had an SMT line, 3rd floor had CNC and laser cutters, and the fourth floor were living quarters. All right next to the Hua Qiang Pei electronics markets. Argh! I need to get these ideas out of my head. I can't be starting hackerspaces all over the world...
After that tour was over, we were then ferried off to the final destination of the day which was a speaker factory. The speaker factory was located a ways out and when we arrived, the owner of the speaker factory showed up with a huge stack of cash in his hands. I'm not sure if he was trying to impress us or just got paid from one of his customers, but it looked like he had a brick of about $20k USD in his hands. The only time I saw cash carried around like that was when I lived in a neighborhood that had drug dealers. Mainly, China is primarily a cash-based economy. I was talking to Bunnie about this because he also carries a fat stack of cash around with him in China. Apparently, most people carry it around because cash is the universal problem solver there. If any issues come up, apparently a bit (or more) of cash can solve most problems. It's an interesting concept and I'll have to explore that more in the future.
The speaker factory was a medium sized factory that had around three or four floors. We only toured the first two floors which was where the main assembly took place. I had always thought speakers were quite simple mechanisms but they're actually fairly complex. There are multiple small components that need to be layered on top of each other to produce the final product. The main components of a speaker are the chassis, magnet, voice coil, suspension, and cone. The assembly flow seemed to be to start out with the chassis which is the speaker's frame. The spider is then attached, and the voice coil assmebly is glued on. From there, the paper cone is attached, the voice coil windings are attached, the ferrite is magnetized, exposed wires are covered with resin, and the terminals are attached. During this tour, it went kind of fast and since I was taking a lot of pictures, I kept on falling behind everyone. I tried to ask how much it cost to get some custom speakers made, but there wasn't much chance because of that. I'm always interested in how much custom devices cost because it's just an option that you never really think is available.
This factory mostly turned out small speakers that seemed on the low end of the cost scale, as opposed to some of the speakers I see in Akihabara which are hardcore audiophile quality. I'm not exactly sure how I would want to customize a speaker unless there was some specific frequency response I wanted to go for. Anyways, it was interesting to see how they were assembled and how all the small pieces fit together.