Pixie8 170 Nodes Per Port No Pi Video

[dibblejr]

Controller Info

Light O Rama. Pixie8 V4

FW. - 1.06

12v

500w PSU

HU set in Advanced Config 170 Nodes Per Port , RGB color order

Pixel Info

LOR Square Nodes 

12v

100 pixels per string

Light O Rama HS Network (RS485HS Red Adapter)

Although I have tested and tested up to 170 pixels per port from Beta to yesterday I never took the opportunity to record a video.

During a new build for a member here yesterday I took the opportunity to put more of those old theories, beliefs and statements to rest.

The video shows exactly what the title of the thread is. It’s pretty simple to see and understand.

Pixie series controllers are able to control at least 170 pixels without power injection (pi).

The new documentation indicates they can control 200 pixels per port and my thought about that is it must be with the pixielink device. 
 

As of this writing I have not attempted to test my pixielink device for number of pixels.

JR

 

 

 

Edited February 25, 2021 by dibblejr

[k6ccc]

JR, A reminder that the LOR pixels (at least in the past) have all been low power nodes.  They draw about 30mA per pixel instead of the FAR MORE COMMON 55 - 60mA per pixel (at full white).  I have little doubt that 170 12 volt LOW POWER nodes will work without power injection.  Do that with standard pixels and you are drawing about 10 amps at full white.  Neither the pixel wiring nor the fuse on the pixel controller will take that for very long.

Note that even with low power pixels, you are drawing about 5 amps with 170 pixels which exceeds the rating of the Pixie series controllers.  It will also eventually blow the 4A fuse if you leave it at full white - although with only a 20% overload it will take a long time.

Also note that I personally have blown 5 amp fuses when 100 pixel strings on my pixel tree went to full white in my show (and the one time I had video of it, it was only for a couple seconds before the fuse blew).

 

[dibblejr]

1 hour ago, k6ccc said:

JR, A reminder that the LOR pixels (at least in the past) have all been low power nodes.  They draw about 30mA per pixel instead of the FAR MORE COMMON 55 - 60mA per pixel (at full white).  I have little doubt that 170 12 volt LOW POWER nodes will work without power injection.  Do that with standard pixels and you are drawing about 10 amps at full white.  Neither the pixel wiring nor the fuse on the pixel controller will take that for very long.

Note that even with low power pixels, you are drawing about 5 amps with 170 pixels which exceeds the rating of the Pixie series controllers.  It will also eventually blow the 4A fuse if you leave it at full white - although with only a 20% overload it will take a long time.

Also note that I personally have blown 5 amp fuses when 100 pixel strings on my pixel tree went to full white in my show (and the one time I had video of it, it was only for a couple seconds before the fuse blew).

 

These are the only LOR pixels I have. They are replacements for the LOR Sing Quartet if I ever needed them. Since the guy I built this controller for wanted the LOR pigtails those were the only ones I had to connect.

Note- all of my other pixels. HC, Rextin. Paul. WW have all responded the same. Perhaps since I do not test all white very long that is a possibility, however this video was to show they can run 170 nodes without PI which would have nothing to do with the fuses.  But the end user would need to test rather than go off the old knowledge of over 100 pixels requires pi.

The argument has always been 100+ nodes REQUIRES pi. Which I have proven in the past with 450 nodes without pi and that test had full white for 30+ seconds.

Also after testing for shows my pixels with S5 are set at 30% intensity. This year going to test 20% intensity if I can see it in the preview. 

JR

[k6ccc]

1 hour ago, dibblejr said:

The argument has always been 100+ nodes REQUIRES pi.

I don't think anyone has said that 100+ pixels REQUIRES power injections.  The general recommendation has been use PI over 100 pixels at 12 volts or 50 pixels at 5 volts. and the STRONG recommendation is to test with the pixels that you will be using. 

1 hour ago, dibblejr said:

Which I have proven in the past with 450 nodes without pi and that test had full white for 30+ seconds.

Do that test and put an amp meter on it and tell me that it meets the current spec for a Pixie (or any other) controller.  Even with low power pixels, that will massively exceed current spec for any pixel controller I've ever seen specs for.  Just because it worked, does not make it a good engineering design.  Next test is measure the voltage at the end of the string with 450 pixels at full white.

I can tell you that I have multiple strings of 12 volt ribbons that are part of my landscape lighting, that could not run 100 pixels at full white without substantial color shift at the end of the strings without PI.  I also had to run my pixel tree at 30% this past year because I could not run any higher than that without the flickering getting too bad.  This is 100 pixel strings of 5 volt pixels and PI was planned, but my COVID got in the way.

My point has always been that you need to test with the pixels that you will be using.  One other tidbit that can come back to haunt you for people who have their controller set to reduce the light level.  If you have a pixel fail in the mode where it sends all white commands down the string, that will be 100% white - not the reduced level.  So all of a sudden your 300 count pixel string that works fine with the controller level set to 25%, has a problem when pixel 37 fail a turn the remaining 263 pixels to full white (and stays that way until you notice it and power it off).  That's also one of the reasons that I don't tend to run long strings.

 

[hasslerk]

2 hours ago, dibblejr said:

This year going to test 20% intensity if I can see it in the preview. 

JR, I created a spreadsheet for this intensity.  If you are interested I can convert it for use in preview and send to you.  It follows the same proportional curve as other LOR pixel curves.  I was also thinking of moving to this curve this year anyways.

[dibblejr]

1 hour ago, k6ccc said:

I don't think anyone has said that 100+ pixels REQUIRES power injections.  The general recommendation has been use PI over 100 pixels at 12 volts or 50 pixels at 5 volts. and the STRONG recommendation is to test with the pixels that you will be using. 

Do that test and put an amp meter on it and tell me that it meets the current spec for a Pixie (or any other) controller.  Even with low power pixels, that will massively exceed current spec for any pixel controller I've ever seen specs for.  Just because it worked, does not make it a good engineering design.  Next test is measure the voltage at the end of the string with 450 pixels at full white.

I can tell you that I have multiple strings of 12 volt ribbons that are part of my landscape lighting, that could not run 100 pixels at full white without substantial color shift at the end of the strings without PI.  I also had to run my pixel tree at 30% this past year because I could not run any higher than that without the flickering getting too bad.  This is 100 pixel strings of 5 volt pixels and PI was planned, but my COVID got in the way.

My point has always been that you need to test with the pixels that you will be using.  One other tidbit that can come back to haunt you for people who have their controller set to reduce the light level.  If you have a pixel fail in the mode where it sends all white commands down the string, that will be 100% white - not the reduced level.  So all of a sudden your 300 count pixel string that works fine with the controller level set to 25%, has a problem when pixel 37 fail a turn the remaining 263 pixels to full white (and stays that way until you notice it and power it off).  That's also one of the reasons that I don't tend to run long strings.

 

Jim

I ran my entire show with many props well exceeding 100 and some 600 nodes without pi.

My custom pumpkin faces 465- 525 nodes each

Monster trees 400+ nodes

Mini trees 150 each

Just to name a few

They are compact but my roof outline has a max run in two sections 156 nodes each

And there are a few people that stated “pi required over 100”

I will find and link the thread which wasn’t very long ago.

I have had pixels fail up until this last season I has only had 3 pixels fail. I never experienced the all white though.

This last year was bad, I think Covid forced factories to not QC their products and I ended up with around 3000 failures most of which are returnable.

I don’t think 5v pixels can go very far but I have never tested other than the guy I repaired his Pixie16 360 tree last year. I’ve never had any 5v.

But yes most of us have always stated “ user should test to verify” that’s always the case. Sometimes online research leads people to the old information when pixels were in their infancy in the USA.

It’s been beat to death but I will link the latest thread.

Jr

Just a quick example

 

Edited February 25, 2021 by dibblejr