RobG wrote:Pixels are not line drivers, so using null pixels helps, but it's not perfect. They have to be part of your sequence, which is also a drawback.
It really bothers me to have you keep repeating this as it spreads dis-information... I've used MANY null pixels before and have NEVER had to adjust my sequences or addressing or anything to compensate for them. If you are using any of the more "full featured" controllers, you may just have to adjust the settings in your controller. Almost all of the "high density" (4 ports or more) controllers out there have simple settings for null pixels at the start of the strings. If that's the only place it's needed (most common use case), pretty much any of the controllers will do. For the more complicated scenario of null pixels in the middle of a string, there are controllers that can handle that as well(the j1sys controllers and the new FalconV2 can), but admittedly not all do (SanDevices for example don't). A null pixel is very cheap ( http://www.aliexpress.com/store/product/DC12V-WS2811-null-pixels-to-regenerate-the-data-signal-with-13-5mm-3-core-black-waterproof/701799_32279633328.html ) and with most controllers and use cases, very simple to use.
That said, I do admit this buffer board is even simpler. 🙂
Wait, so you are saying that
have NEVER had to adjust my sequences or addressing or anything to compensate for them
but then you also say
If you are using any of the more "full featured" controllers, you may just have to adjust the settings in your controller. Almost all of the "high density" (4 ports or more) controllers out there have simple settings for null pixels at the start of the strings. If that's the only place it's needed (most common use case), pretty much any of the controllers will do. For the more complicated scenario of null pixels in the middle of a string, there are controllers that can handle that as well(the j1sys controllers and the new FalconV2 can), but admittedly not all do (SanDevices for example don't).
Maybe I'm speaking out of turn, but I think that he is distinguishing between making a one-time adjustment in the controller configuration versus compensating in the sequencing.
Well, whatever it is, the truth is you have to keep track of null pixels. If you move/add/remove null pixels, you have to make changes, in the controller or in the sequencing. Maybe I was too vague when I wrote "...part of your sequence," I should have written "...part of your setup."
Also, Ray's null pixels are 12V WS2811, meaning you need a different one for 5V, then another one for different type of IC, and so on. My board is (almost) universal.
Whatever you want to use, it's your choice. I have a setup in which null pixels will not work, so this is what I will use.
If you're putting nulls ANYWHERE at this point you have to compensate for them, which is a HUGE pain in the rear end, especially when lights are up on the roof and you're making sequence edits. This board is CHEAP and there's no work involved in figuring out it's location in order to compensate for it....Rays' nulls are nothing more than a live pixel that you still have to account for. Using nulls at the beginning of a string as it comes out of the controller is a SIMPLE edit, anyone with knowledge of what a null pixel is and access to the setup page can do that. The problem comes in out in the yard when you've done a 300 pixel run and now you need to do a 30 foot jump to the next set without having to run a new homerun from the controller.
RogerH wrote:Here's an easy way to explain it. I have a mini tree using 98 pixels. After pixel 98 I use the signal output and power , 3 wires, conne t it to a max485 chip. from that chip I run a cat5 cable using 2 twisted pairs pretty much as far as I need to. At the end of that cable I use another max485 chip to translate the signal running over the twisted pair of signals back to a single signal. What I end up with is signal and power to go to my next mini tree. Mind you, I will not power anything except the two chips with the cat5 cable. The signal and ground goes to the next mini tree and power injection is used at that point. Roger
I am thinking the 2nd board is not required. Use the first to drive the differential signal. Terminate the twisted pair into a termination resistor as is proper for RS485.
Due to the high impedance of pixel inputs, pick off the clean + data and feed the next pixel string. Leave the data compliment tied to the termination resistor and nothing else.. Eg, fo not ground it.
This should be good for several hundred feet if the far pixels run on their own local power supply. The pixel minus would need connected back to the controller ground as a signal ground reference is still required for the pixel input data.
In summary, from the line driver send Ground, and Data + and data -. Terminate the twisted pair at the far end with a resistor of 120 ohm for Cat5. Connect the ground and data + to the next pixel string and inject power.
A twisted pair receiver should not be required except for runs over a couple hundred feet where long run attenuation is excessive.
When can we order this board?
Technician wrote:I am thinking the 2nd board is not required. Use the first to drive the differential signal. Terminate the twisted pair into a termination resistor as is proper for RS485.
Due to the high impedance of pixel inputs, pick off the clean + data and feed the next pixel string. Leave the data compliment tied to the termination resistor and nothing else.. Eg, fo not ground it.This should be good for several hundred feet if the far pixels run on their own local power supply. The pixel minus would need connected back to the controller ground as a signal ground reference is still required for the pixel input data.
In summary, from the line driver send Ground, and Data + and data -. Terminate the twisted pair at the far end with a resistor of 120 ohm for Cat5. Connect the ground and data + to the next pixel string and inject power.
A twisted pair receiver should not be required except for runs over a couple hundred feet where long run attenuation is excessive.
I am thinking that this is something that I'd only do in a pinch or just to see how it works. The reason that I say this is that I don't consider the signal at the end of a 50 or 100' cable to be a good signal reference. In many situations this would be perfectly fine, but there are others where there may be either a lot of noise or DC offsets present (the ground wire has a non-zero resistance). This depends on the overall set of ground connections in the display and on how the ground wires are run (what is next to them).
It is nice to see some diversity in the offerings for this sort of thing. Null pixels a pretty easy to add, but this approach is effectively transparent and can drive longer distances than a standard null pixel.
I am already in the uAmp group buy so I probably won't go for any of these boards this year, but would be curious how this board and the uAmp stack up against each other for next year.
Are some of those components on the board some sort of line termination resistors? If so, are they series (source) or parallel (destination) terminators? And what line impedance are they configured for (assuming that very few people would be assembling these themselves)? I assume that you are not planning on using Schottky diodes as terminators.
