Pixel colors are off down the length of the string.

[fasteddy]

If they are like the candy canes i build i have them connected together, so the bottom of the candy cane is where th power and data comes in and the top is where the extra cable is added to then connect to the second candycane.

So what im suggesting is using that 18guage cable as your supply to the pixels but also then connecting the 18 guage cable from the first to the last pixel. This way you eliminate some of the voltage drop created in the cane

---- pix1-----pix2----pix3-----pix4------pix5------ (candy cane)
|-----------------------------------------------------| (18 guage bypassing pixels and connected in series)

So basically you run the 18 guage in parralel with each cane so you eliminate the voltage drop through each candy cane as your not relying on the small cable that is connected between each pixel to then supply power to the second candy cane

 

[bob_moody]

Eddy,

Thank you. I understand the explanation but question why even send power to the last pixel?

Why not just parallel power and data up the cane and only return data to keep the sequence intact and parallel off the power coming in to send to the next cane?

Not trying to be difficult here, just trying not to destroy the 8 canes I have built and trying to understand the extra wiring

Bob

 

[fasteddy]

You dont want to parrallel the data lines only the power

Basically what is happening is that you are loosing voltage through each candy cane because voltage is lost due to the guage cable between each pixel and this is also determined by load going through those cables. So if you connect your 18 guage canle from pixel 1 to the last pixel in the candy cane then there will be enough current capacity to then power the second candy cane because that voltage is not having to go through each pixel.

Im under the belief that you have 4 candy canes connected together, one to the other so by doing this you are providing a much better voltage level for the second candy cane because it doesnt have to go through 22 guage cable and pixels to supply the second candy cane and also the small 22guage cable between the pixels doesnt have to carry a high current, instead it will flow down the 18guage.

You were worried about having too much injection, so this suggestion will allow less injection points without having to provide seperate power injection for each candy cane (or each second candy cane)

It all has to do with lowering the resistance and increasing the gauge cable so the current can be supplied to the other candy canes that are connected. This will reduce the voltage drop seen at the second candy cane. repeat this with each candy cane until you need injection.

 

[bob_moody]

Eddy,
I think we are on the same page.

Here is what i did. I have coming into the candy canes 4 pin waterproof connectors with 18ga wire.
There is one going in and one coming out.
I tied the +12 together on each of the connectors then tied in the 12V to the pixel

Pixel16
^
|
Pixel2
|
Pixel1
__|__
/ \
+12In +12Out

I am still send data up the pixels from the input and bringing data back to the output from pixel16

This has improved the results somewhat....
With all 16 pixels lit white the new results are:

V-Drop v-Drop
Cane Before After
===============================
1 11.5 11.8
1,2 10.1 10.8
1,2,3 9.0 9.9
1,2,3,4 8.3 9.0

With all 4 canes on (white) I have picked up .7 volts, reducing the total loss in 4 canes to 3 volts. Which if I recall you mentioned was about the threshold.

I still need to go back and revisit the current draw and see if it has changed.

Is this better, worse or insignificant?

Thanks again for putting up with me on this...

Bob

 

[fasteddy]

Eddy,
I think we are on the same page.

Here is what i did. I have coming into the candy canes 4 pin waterproof connectors with 18ga wire.
There is one going in and one coming out.
I tied the +12 together on each of the connectors then tied in the 12V to the pixel

Pixel16
^
|
Pixel2
|
Pixel1
__|__
/ \
+12In +12Out

I am still send data up the pixels from the input and bringing data back to the output from pixel16

This has improved the results somewhat....
With all 16 pixels lit white the new results are:

V-Drop v-Drop
Cane Before After
===============================
1 11.5 11.8
1,2 10.1 10.8
1,2,3 9.0 9.9
1,2,3,4 8.3 9.0

With all 4 canes on (white) I have picked up .7 volts, reducing the total loss in 4 canes to 3 volts. Which if I recall you mentioned was about the threshold.

I still need to go back and revisit the current draw and see if it has changed.

Is this better, worse or insignificant?

Thanks again for putting up with me on this...

Bob

Yes we are on the same page

You will find that your current draw has increased due to the higher voltage allowing more of the 2811 chips to drive at the full 18.5mA

So now your injection at candy cane 5 should yeild better results

 

[bob_moody]

Eddy,

Here is a view of the four canes. I'm still seeing the part of the 3rd and 4th cane going pink when on full white. However, when in motion and color changing you really cant tell. I'm just wondering if it can get any better without individually feeding each cane?

Bob

 

[fasteddy]

you must consider that these candy canes with 16 pixels are power hungry. So the 4 candy canes are equivalent to more than 1 strip, so in reality you would have to reinject power between the third and fourth candy cane (16 modules x 3 = 48 pixels) The reason these modules come in strings of 20 modules is because anymore than 20 modules starts to require injection.

So the voltage drop is only really an issue when at full current draw when all white because this is when the maximum current is required.

You could try reducing the output to 85% when using white and you wouldnt see a great deal of light output difference because the lower current will reduce the voltage drop being seen.

 

[dariansdad]

A suggestion may be to use the 18 guage to supply and then also use the same 18 guage to then bypass the pixels to the end of the candy cane. So basically you will be powering each candy cane from both ends because your just adding another cable that is in parrallel to the candy cane pixels. This will enable you to use less injection with the current design

Not to hijack this thread but in the attached drawing, Is this what you're saying about power injection at the end?

 

[Bill Ellick]

A suggestion may be to use the 18 guage to supply and then also use the same 18 guage to then bypass the pixels to the end of the candy cane. So basically you will be powering each candy cane from both ends because your just adding another cable that is in parrallel to the candy cane pixels. This will enable you to use less injection with the current design

Not to hijack this thread but in the attached drawing, Is this what you're saying about power injection at the end?

That would be correct for injecting power from each end for that "run" of pixels. If you had a longer run say of 150 pixels, then you could inject power at the midway point as well if needed or at 1/3 and again at 2/3 of the way along the string. Roughly every 50 pixels or so is a good rule of thumb for power injection for 5 volt pixels but it really depends on the current requirements of the specific pixel you are using as far as current required to operate them.

 

[bob_moody]

Well I am in the home stretch of completing the 8 candy canes and I would like to stop for a moment to tell everyone here how much I appreciate the help in working out the power issues.

There were so many "Ah-ha" moments that it all kind ran together into one long "ahhhhhhhhhhhhhhhhhhhh!"

I am a visual learner and seeing the results of current consumption, voltage drop, wire size effects was a real educational journey. I never put together, for example, the number of pixels I was using in relation to say a string (strip) of pixels which, had I made that relationship, might have saved myself a lof misery.

My biggest disappointment in the whole process was how far along I was in completing the project before I realized the issues. Of course I built and tested, but it wasnt until I got enough of them together that I started seeing the power issues manifest themselves. I really didnt want to tear up my props while trying to resolve the issue so it took some creative thinking on how to fix it.

I am now to the point that I have 8 very bright candy canes with 3 injection points (cane 1,4,7... which is 48 pixels per injection--AHHHH.. the Pixel Strip analogy!!!) and a little less than a 2V drop measure at the end of cane #8.

So. Once again, THANK YOU... to everyone here at ACL for thier assistance.

Bob

 

[Bill Ellick]

Sounds great Bob.
Glad to hear that it is working out for you. I know what you mean as I am one who just does not grasp things from reading very well (as well as I have mild dyslexia) and really need more of a hands on to get things to work.
Trouble is that I am getting old enough that when I have that "ahhhhhhhhhhhhhh" moment, I am afraid that I may pee my pants! LOL
But you are well on your way with things now so build away and make those pesky little pixels shine brightly!