| << return to blog entries
2010-05-30 07:44:03 (7590 views) 1. May have eliminated the need for the ferrite bead by implementing very simple EMI suppression circuitry directly within the motor overmold; this helps keep the arousal inducer small. For a definitive answer we'll have to wait until our final day at the EMC lab where will we obtain all the compliance test results (and destroy a few units in the process thanks to the european LVD requirements). This is the last thing that must happen because we cannot modify the product's hardware after this or we'll have to retest everything (few hours of testing cost several thousand dollars).
2. One day at the factory revealed that we're still 1 step away from having a simple, reliable manufacturing process for this part, however we know now that in the worst case we can do part of the work by ourselves (the part which requires precision and concentration). In the best case we've already found a material that they can use to streamline the process - but we need to verify this by doing further testing, and for this we need to produce more of the sample metallic pieces we talked about. From now on we will no longer discuss the specifics of how we're overmolding this part as it's gotten complex enough to constitute a true "trade secret"; besides things are changing daily so there's no point in taking anything we write about it too seriously.
The important thing is that we've been exposed to a very large number of problems with this part (and with molding rubber in general) and now understand exactly how to fix each one, but we're still trying to optimize the process so that it doesn't require much attention at all. There are still a few "regular" molding problems that keep occurring (too little or too much rubber, poor adhesion to the cable, cable slightly burned by the mold) but these are problems that were solved before in at least one attempt and so it's just a matter of putting all the tricks together (no small feat, to be honest) and Mr Tang's workers getting some practice at producing this part.
This know-how will be vital as we extend our product line in the future and integrate more rubber components.
3. Right now as you know from previous posts we are producing the very final DL2000 prototype, something which became necessary due to RF range surprises and latest achievements, and the need to recharge the remote's batteries (can't ship fully charged li-poly packs due to regulations and more). The receiver's top has some structures which may or may not increase the "beep" volume, which gets from annoyingly strong to very faint when the waterproof enclosure is closed. We'll see where we get with that one.
4. We've found another battery provider. So now instead of having one unreliable guy who doesn't know the first thing about li-poly safety and won't return emails, we have one company for the li-poly pack, the guy speaks English decently and they have all regulatory papers for importing their batteries into other countries, AND another company for the receiver's non-rechargeable li-ion cell. Long story short we've eliminated one unreliable component from the system.
5. This coming week we'll probably have an appointment for the spray painting, this should be fun and we'll get to test out many different styles and colors
6. We've TENTATIVELY solved all the problems related to part stock disappearing on us! Already 4 components had to be replaced, and the microcontroller (which can't be replaced) was finally found on a site, after even the manufacturer declared they're out of stock for months to come. We have enough microcontrollers to make 50 units. The rest of the parts should be something we can replace even though they keep disappearing one by one (never imagined sourcing would be such a hassle). For instance we just replaced one diode with another type, but this new type eats up just a little more current so the remote battery may last a tiny bit less (not a huge problem now that it can be recharged easily). We keep on fighting! Let's hope we have no more trouble sourcing components. On the positive side the RF components which were previously estimated to have a 3 month lead time, have already been produced so this is a huge good thing that happened on the sourcing front. We'll have enough RF components for 50 units very soon. After that we'll have funds to do bigger production runs and we're hoping we won't have to monitor parts stock weekly as we're doing now.
Comments
Post new comment
 |
By equinox316 at 2010-05-30 09:54:24 Reply 
By dreamloverlabs at 2010-05-30 10:11:21 Reply