Barnabybear
New elf
Hi Al, I’ve linked the data sheet below. I don’t have any of the pixels to test but from the sheet:
If just Red was supposed to be on – no output.
If Red and another colour was supposed to be on – the Red would not light, the other colour would light at times that Red was not on i.e. if Red was 25 and Blue was 50, when the Red and Blue where due to be on nether would light but when only the Blue was due to be on it would light. The result being a Dim Blue (25) as opposed to Red (25) and Blue (50). There is also the stagger output to take into account later on page 8. So even if colour was supposed to be Red (25) and Blue (25) there would still be a time where Blue still lit when the Red finished its time period (which it started before Blue).
“When the external control data is transmitted, SDI uses the 1-24bits received data as the display data, meanwhile,SDI2 will discard the 1-24bits data and use 25-48bits data as the display data.”
It looks like it skips the whole pixel.
“GS8208 gets into RGB test mode right after power on. If the device does not receive the external display data for a long time, the chip will go into the built-in display patterns mode.”
“Under the internal display mode, the first chip’s SDI/SDI2 needs to be connected to GND.”
“Under the internal display mode, the first chip’s SDI/SDI2 needs to be connected to GND.”
From a controller or WS281x pixel running the OLD 50uS reset timing, in most cases I suspect they would operate correctly but this is not guaranteed. Whilst the controller may only be catering for a ‘reset’ LOW of 50uS, because by default the WS281x data line is LOW, if there is no data to send, a ‘reset’ LOW of greater than 50uS it produced. However in the case of high frame rates (40FPS) and long pixel strings 500+ dependant on when the data arrives, there is no guarantee that there will be additional time added to the 50uS to reach the 300uS required to trigger an update. In general whilst appearing to work fine most of the time in testing, whilst out in the show a delay in E1.31 packets arriving could cause random freezing.
The GS8208 data signal does not conform to the WS281x specification.
Part of signal as percentage of time-----WS281x-----GS8208
T0H------20%-----25%
T0L------80%-----75%
T1H------48%-----75%
T1L------52%-----25%
Zeros aren’t far out and may be within tolerance but the ones are a mile off. I’m not sure that Ws281x pixels following GS8208 would work or be a good idea without a lot of testing.
- If you lose 1 led out of the RGB does the pixel go dark with no colours shown or does it just show the other 2 colours.
If just Red was supposed to be on – no output.
If Red and another colour was supposed to be on – the Red would not light, the other colour would light at times that Red was not on i.e. if Red was 25 and Blue was 50, when the Red and Blue where due to be on nether would light but when only the Blue was due to be on it would light. The result being a Dim Blue (25) as opposed to Red (25) and Blue (50). There is also the stagger output to take into account later on page 8. So even if colour was supposed to be Red (25) and Blue (25) there would still be a time where Blue still lit when the Red finished its time period (which it started before Blue).
- With the data backup does the next pixel process it's own data or does it process the data from the previous "dead" pixel. In short if you have a matrix or similar do the lines stay in sync with 1 pixel missing or does the missing pixel cause everything downstream to shift 1 pixel.
“When the external control data is transmitted, SDI uses the 1-24bits received data as the display data, meanwhile,SDI2 will discard the 1-24bits data and use 25-48bits data as the display data.”
It looks like it skips the whole pixel.
- On power up how long is it before the internal test function starts up and does it require the data signal to be open circuit, 0V or 5V to start the test signal.
“GS8208 gets into RGB test mode right after power on. If the device does not receive the external display data for a long time, the chip will go into the built-in display patterns mode.”
“Under the internal display mode, the first chip’s SDI/SDI2 needs to be connected to GND.”
- Related to 3. Will the pixels resort to a test signal if they don't get a valid WS281x type signal for a given amount of time. This can be important if pixel controllers don't send data out at all times.
- Does each chip produce the test signal or does the first pixel that doesn't see data and it gets sent out along the data lines?
“Under the internal display mode, the first chip’s SDI/SDI2 needs to be connected to GND.”
- If the first pixel produces the test signal is it sent out in a WS281x type signal?
- The GS8208 runs off the same data signal as WS281x. Can the GS8208 be daisy chained in with WS281x pixels with the WS ones before and/or after? Worldsemi screwed around with their WS2811 timing and I was wondering if that has any effect on GS8208 pixels.
From a controller or WS281x pixel running the OLD 50uS reset timing, in most cases I suspect they would operate correctly but this is not guaranteed. Whilst the controller may only be catering for a ‘reset’ LOW of 50uS, because by default the WS281x data line is LOW, if there is no data to send, a ‘reset’ LOW of greater than 50uS it produced. However in the case of high frame rates (40FPS) and long pixel strings 500+ dependant on when the data arrives, there is no guarantee that there will be additional time added to the 50uS to reach the 300uS required to trigger an update. In general whilst appearing to work fine most of the time in testing, whilst out in the show a delay in E1.31 packets arriving could cause random freezing.
The GS8208 data signal does not conform to the WS281x specification.
Part of signal as percentage of time-----WS281x-----GS8208
T0H------20%-----25%
T0L------80%-----75%
T1H------48%-----75%
T1L------52%-----25%
Zeros aren’t far out and may be within tolerance but the ones are a mile off. I’m not sure that Ws281x pixels following GS8208 would work or be a good idea without a lot of testing.
- If you have the RGB brightness all set to 5 (out of 255) for instance is the current 18mA or closer to 1mA
Last edited:
