Do we get an extra reel by posting a review? LOL! j/k Ryan!
I did not pay (much) attention to the actual specs on the strips. They're not coated! Nor do they have a sleeve (this was mentioned in the original post, I just read right over it.) So they are what I would call 'bare' strips. This is important if you plan on using these outdoors or without any kind of enclosure. Other than that, they're 48 pixels per meter (which worked out perfect for my R&D process). Detaching them at the solder joints is a tad tricky. You can cut where indicated, but if you want to cut an actual solder joint, you're not left with a whole lot to solder onto afterwards, so I always detach the joints themselves. This one is a bit tricky because the solder pads are tiny circles and they stack them on top of each other at the ends. So the best way I could do that was to insert a thin set of fine tipped straight tweezers in between the layers right around the solder joint and as I heated it up, gently push the tweezers in and separate the layers. This leaves both the top and bottom pads intact and ready to have wires soldered back on them. Some strips have larger pads and they're soldered side-by-side making it easier to simply cut the joint. This one has the edges on top of each other. But once separated, it's easy to get wires back on. Another thing to keep in mind, especially for those who have dealt with WS28x strips before, VCC and GND are next to each other on these. The WS2801 strips I have are different in that VCC and GND are on the outside with CLK and DATA in the middle. I actually prefer that as there's a much smaller chance of accidentally creating a solder bridge between VCC and GND. On the plus side, it has two GND pads which makes for a stronger, more secure connection. I wish they did the same for the VCC pad but alas. That's it, strips work beautifully.
So my reason for getting one was two fold. On the one hand, I wanted to make sure I wasn't doing something wrong versus the WS2801 strips I have being just too slow for what I am wanting to do. On the other hand, I just wanted another IC to play with. I've always only worked with the WS28xx series ICs and strips so why not. I already had a sneaking suspicion that it was the WS2801 strip (note, 2801, not 2812, although their speed is the same) but I needed to make sure. And rather than waiting 2 weeks for my vendors in Shenzhen to send a test reel to me, Ryan, you were a savior here.
So, I'm working on a custom, full color POV system, like a poi baton. Most, if not all of the stuff I see have static patterns and are limited in color, mainly because everyone stores it in memory of which there isn't a whole lot to begin with. So my system stores images on a uSD card. The program then reads each individual column of the image and pushes the data out to the string, constantly. At 48 pixels tall, I just need three bytes, r, g, and b. So I read in a total of 144 bytes (48 pixels) each time and push it to the string, then read the next 144 bytes, etc., etc. The only delay is with the uSD itself. On average, it takes 108 usecs to read 144 bytes (out of cache) and about 1250usecs every 4th read to refill the cache. Plenty of time to goof around with.
Turns out, I was right, the WS2801 is simply too damn slow. Here's what the WS2801 strip would do when I tried to push data to it at full blast:
And this is what it's supposed to look like. This is running on the LPD8806 IC:
Now, I could've just as easily slowed the process down on the WS2801. It would take almost a full 2.5ms to get it to work right. Add the 1.2ms from the uSD and I was looking at close to 4ms for each update on the string. That was way too long.
I've now switched to an LPD8806 design and thanks to the strip that Ryan sent, I was able to both confirm my suspicion as well as continue development. Here are some other images taken using the LPD8806 strip: