Maybe you can help me with a little dilemma i am having! Hope this is the right place to post.
For a POV project i am thinking about, I am restricted to WS2801 or WS2811/2 because the form-factor (package?) of the LED node only is available with a choice of those!. Im looking to choose the best IC for the job, given the circumstances.
In my past research I have seen quoted that 2801’s refresh is 1mHz and in someplace i have seen it listed as high as 1.25mHz -whats the real measurement?
Also 2811/12 refreshing at either 800Hz or 400Hz? i dont even need someone to verify this, we ALL know how slow these drivers are.
My instinct says higher refresh rate will make it more optimal for POV purposes, but, would not there also be a upper threshold for refreshing ‘too fast’?
What am i missing here? Just making sure by going with the 2801 im not shooting myself in the foot.
That 1mhz rate with the ws2801s is the data rate, not the refresh rate.
I don’t know what the pwm refresh rate on the 2801s is, the ws2812 refresh rate is only 400hz and is fucking terrible for POV.
What package/form factor are you using? The apa102’s 20khz refresh rate (note the k in there) is fantastic for POV.
Ah - the WS2801 has a 2.5kHz refresh rate
AH. yes that makes a great deal more sense. spec i just pulled from WS for the 2801 says “PWM free-run capability(refresh rate(2.5KHz)”.
as for package - best case im looking for 12v nodes at something between 45-60mm diameter - 6 LEDs if they are one sided. 12v to go some distance with it all before the start of the piece.
would be dope if i could find some ‘MiSphere’ style white balls with an lpd8806 driving 3 per side - but all im finding are 2801or 2811, with option for custom spacing. My MOQ is like 150pcs for this iteration of the piece, so getting a custom pcb design is not really an option.
can you imagine 3 apa102’s per side of a sphere?. THAT would be super fly.
im confused about clock rate vs data rate, and how they play into the best case requirements for PWM. faster is better?
I know that better PWM refresh means ultimately a more solid and cohesive afterimage (less dots)
so, the 2801 has a faster PWM refresh than the 2811, and a clock line. . so unless i can find a better product with a better IC, seems like 2801 is it.
final thought does the presence of a clock line mean no speckling?
Data rate: How quickly you can send data to the leds. With the WS2811’s, you can send data at 800kbps, with the WS2801 it is closer to 1.25Mbps, with the APA102 it’s around 24Mbps.
Refresh rate: How many times per second does the led chipset run through its pwm cycle. The higher this number, the better for POV the chipset is. The WS2811 does this 400 times per second while the WS2801 does it 2500 times per second and the APA102 does it 20000 times per second. This affects the time “width” of your PWM pulses. Let’s say you have the red led set at 128 (50%). At 400Hz, the red led will be on for 1.25ms then off for 1.25ms At 2500Hz it will be on for 0.2ms then off for 0.2ms. At 20000 it will be on for 0.025ms and off for 0.025ms. If your led is moving, the time the led is on and the time it is off for becomes the “width” of the striping that you might see. The thinner and more frequent those stripes (e.g. the shorter the time the led is on then off for) the less likely you will be to see the differences (hence, the POV).
The presence of a clock line or not doesn’t mean much as far as speckling. That may have more to do with the fact that if you try to send data to the WS2811 more than 400 times per second it behaves badly.
duh, sorry, currently not giving 100% focus to what im trying to say.
what i was trying to say was: higher PWM should mean less speckling or striping or whatever you want to call it… clock line should increase stability.
Yes, higher PWM rate will mean less striping. Clock line gives you more flexibility in timing (not as rigid as the WS2811’s) and the potential for much higher data transfer rates as well.