I know. I talk too much. After further evaluation it can be done. I would add a converter to the nulls but don't output power from the nulls to the 300 pixels since the only reason the nulls are needed is to drive the signal. Then add a converter to each of 6 50 pixel groups and I would get the full 5 volts to the pixels. Am I right?
Sent from my Nexus 7 using Tapatalk
what voltage are you going to be driving down the line?
Just an FYI, there has been many changes in CMOS since the 4000 series. Not all has the 3.3V threshold. Many custom chips have input signal conditioning (hysteresis) and other threshold levels. Some pixels allow for some input voltage drift associated with voltage drop on the input ground wire related to pixel current draw.
Current draw on the data lines when above 0 and below 5 on the pixels I have is very very low. Current is measurable when the string is drawing current so the power leads have some voltage drop.
For example, assume a 1 volt drop. The Plus line would drop 1/2 and the minus line would drop 1/2. What this means to the data line at the pixels is as follows.
Measuring from ground at the power supply, 12 volts will be down 1/2 volt to 11.5 volts at the first pixel. The Minus line will be UP 1/2 volt at the first pixel to 0.5 volts. The DATA line with a 0-5V data stream will be seen at the first pixel as -0.5 to 4.5 volts. This is part of the problem often seen with long lines from a controller and power supply as the power lead losses are greater with longer leads.
The Null pixel if inserted midway in the above example does a local voltage shift so it would see 1/2 the above voltage drops and re-create the signal with the new ground reference. It would be as follows.
Pixel positive would be 11.75. Pixel negative would be +.25. Data would be -.25 to +4.75 going in and shifted related to the pixel ground. Data out would be referenced to the pixel ground so related to the power supply ground the data to the pixel strip would be low of +.25 and recreated against internal 5 V ref for the 12V pixel, high would be +5.25.
Remember when you have voltage drop in DC power, the ground is not the same throughout the system because there is voltage drop in all current conductors.
Dropping a volt from the power supply to the first pixel will drop half in each power conductor.
Since pixel data is a data signal, are you measuring it with an oscilloscope? Average DVM readings will vary greatly depending on what the pixel controller is sending. Even if you shut off vixen, the pixel controller may be sending a repeat of the last valid data received.
You lost me on the technical jargon. I have attached a diagram of what I'm intending to do.
I did a test and increased the voltage to 5.4 volts and all the problems disappeared. I don't think the pixels were intended to handle that much so I dropped it down to 5.2 volts. The flickering came back but not as bad. The colors were also slightly off.
I hope switching over to 12 volts and buck converters that I can sustain 5 volts at the pixels. Notice that each ball has a separate feed to the pixels. I also have it connected to the backside of the pixels so there's voltage coming from both sides (injection).
Yes this will work. Two null pixels may be overkill. I put a null pixel every 20 feet and it all works fine. Make sure you use a good heavy gauge wire to run the power.
