Coming to grips with voltage drop.

[BundyRoy]

I have worked out my layout and would like to just double check I have a handle on the issue of voltage drop.

I plan on having a matrix of 10 strips of 3m long 12V ink1003 30 leds/m (this). It will be 9m one way from my controller so say 10m one way from the psu.


Wasn't sure what current to use. So the specs say 7.2W/m. So based on P=VI then 7.2=12*I. I=.6A. First question:

Is that 0.6A/m or 0.6A total or completely wrong?


Now say I wanted to check voltage drop with 14/0.20 cable. From the bible it has a resistance of 0.043 ohms/m. Second question:

Is the resistance calculated for the cable length (20m in this case) or cable length plus light strip length ?

I will assume it is just the 20m of cable in order to continue my working. The R=20x0.043=0.86 Ohms

So V=IR. V=0.6x0.86=0.52V


So assuming I have assumed correctly (not that likely) I have a voltage drop of 0.52V. Question 3:

What voltage drop is acceptable?

I seem to remember something about a tolerance of 5%. This equates to 12x.05=0.6V. By my numbers I am getting close to this and should look at getting more cross sectional area of wire.


Question 4:
If I have a cable that has enough cores so that I could double up the wires on the +ve side but only run a single wire on the -ve does that help? Or should I stop mucking around and get bigger wire.


I haven't ordered any wire yet. Hence why I'm checking calculations and options now.

Other relevant info that may be required:

30m x 30LED/m = 900 Pixels = 2700 channels.
Using a Pixlite 16 controller. Handles 340 pixels/output.
I have 3m x 30 pixel/m= 90 pixel/strip.
Need to use minimum of 4 outputs. (2 lots 270 pixel, 2 lots of 180 pixel.)


I assume the power requirements are okay for one PSU. 30m x 7.2W/m = 216W. This equals 216/350W = 62% load.

Sorry for the long post and multiple questions but it is hard (at least for me) to look at each issue in isolation.

Thanks

Roy

 

[AAH]

So many questions.
The first answer is it is 0.6A per metre.
The 2nd answer is it is only for the cable leading to the strip so in your case it is 20m.
With ink1003 I would expect a voltage drop of up to 3V would probably still give you perfect results but that is just a guess as I've never seen specs on the regulator that is used in conjunction with the ink1003. That regulator is totally seperate from the pixel so it would need to be 2 specs sheets or a combined 1 that is supplied.

As you are looking at putting up a 30m matrix with 0.6A per metre you're talking about some serious current (18A) and as such some serious voltage drop. I personally would be tempted to mount a power supply at the matrix. The power supply doesn't have to be the same 1 that is running your controller (only the 0V lines need to be commoned).

As an alternative to running a mains power supply out at the matrix you could use several of http://www.ebay.com.au/itm/24V-to-12V-10A-120W-DC-to-DC-Power-Converter-Regulator-Adapter-Module-TN2F/291041320986?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222003%26algo%3DSIC.FIT%26ao%3D1%26asc%3D20140106155344%26meid%3D6799900158177615794%26pid%3D100005%26prg%3D20140106155344%26rk%3D1%26rkt%3D6%26sd%3D271272594944&rt=nc or similar to provide localised regulation at the matrix and feed that converter with 24V which reduces the current through the cables and gives you a true 12V at the matrix.

 

[BundyRoy]

Thanks Alan. At least I got some of the basics sort of right. I think I can put a power source next to the matrix on the roof so that hopefully will solve the majority of my issues.

My biggest load on an output is 3 strips (9m). Based on my previous calculations this is 9 x 0.6 = 5.4A. The controller can only handle 4A per output. Does powering the matrix at the matrix instead of through the controller get rid of the load on the controller or does it still see a current of 5.4A. I'm assuming it takes the load off the controller as the controller effectively only has the data and common wires connected.

One last question. You said the ink1003 could likely handle a voltage drop of say 3V. I assume this is because it has an inbuilt regulator to reduce the 12V and operate on 5V LED's. If I'm using 12V 2811 strip elsewhere what is an acceptable voltage drop for it.

 

[AAH]

Hooking up a power supply adjacent to the matrix gets rid of pretty well all your problems. You don't connect the +ve output from the pixlite to the matrix pixels in this case. The -ve must be connected. You therefore have no load on the pixlite other than the data line.

As far as the voltage drop on the 12V 2811 you can probably get away with up to 3V drop but that is the voltage drop both in the cable and also along the strip so you have to be highly aware of your cable sizes and injection points if you're doing a long cable run. I ran some 12V pixel modules on my eaves last year and I used some 4mm2 cable to get from the power supply to an injection point a fair way along the run. I had to inject at several other points too and even with all the points I injected at I could notice some colour variation when the pixels were all on white.

 

[BundyRoy]

If I'm putting the psu at the matrix should I be putting some in line fuses on the positive wire from the psu as a safety precaution or is this a waste of time.

 

[AAH]

You should always have a fuse between the power supply and the lights. The onboard fuse on controllers is reasonably suitable but in your case you should be using a few fuses for the power injection points. The fuses are generally there to protect the wire from any faults in the lights or wiring.

 

[BundyRoy]

Thanks for all your help Alan.

I think I've got a grip on it now. We'll see how we go.

 

[BundyRoy]

Doing my calcs for the voltage drop. It appears to me that with strip lights that have power usage of 7 W/m and higher then if you have a strip length of 5m you are looking at currents in the order of 3 Amps. If my understanding is correct then this 3A load is getting close to or exceeding the rating for outputs on controllers. The high current also means that unless very big cable is used then the voltage drop is going to be significant unless the controller and power supply is very close to the strip (or at least the power supply).

Is it common to have to have the power supply adjacent to the strip and downstream of the controller. I'm just concerned that everyone talks about using security cable (0.5mm2) and based on my calcs this should be resulting in large voltage drop in a short distance.

 

[AAH]

I'm too lazy to look at links at the moment or even to look to see if there were any. The security cable is ONLY suitable for getting the data from the controller to the pixels IMHO and all the power has to be supplied via heavy duty figure 8 or similar cable. I don't really know what other people have done but I do know a lot of people get it wrong on 1st, 2nd, 3rd attempts as they aren't putting enough power into their pixels.
The controller outputs kind of don't come into it as far as the 3A goes. DC controllers are rated in amps per output but pixel controllers supply 1 or 2 data lines which are very low current and then the 2 supply lines which can work out at up to 30A over the 680 pixels (4 universes) that some controllers can control per output. No controller that I know of has fusing and power distribution that will allow control of that sort of current direct from the pixel controller.

 

[BundyRoy]

Excellent, Thanks Alan, that's what I wanted to hear. It backs up what my calculations are telling me. I just wanted to get it right as I have to order it in. Can't just duck back and swap it at the store if I get it wrong.

2 Questions.

If I'm powering the lights downstream of the controller do I need to take the ground/- ve/common wire (whatever it's called) back to the controller or can it just go to the psu same as the +ve. I'm guessing it does need to go back to the controller to complete the data circuit (as well as go to the psu).

The data wire. Does the data signal travel any better/further down a thicker wire or is using thicker wire just wasting money.

 

[AAH]

There must be a common ground (-ve) connection between any/all power supplies, the pixels and the controller.
The configuration possibly makes a slight difference but I don't think that it will affect things in most cases. As long as the pixels get nice fat power wires the wiring back to the controller isn't really all that important.

Low capacitance wire is probably the best choice rather than thin or fat wire but like the order in which the -ve wires are ran it is probably only a marginal thing.

 

[davrus]

There must be a common ground (-ve) connection between any/all power supplies, the pixels and the controller.
The configuration possibly makes a slight difference but I don't think that it will affect things in most cases. As long as the pixels get nice fat power wires the wiring back to the controller isn't really all that important.

Sorry for my confusion... but somewhere someone told me that the cross connection of the common supplies that happens where you inject power, is the best place to achieve the common ground between power supplies. Thus it is unnecessary to join -ve to -ve at the power supplies.


But ..... what happens when you are only getting data from the controller, and all power comes from a different power supply than the power supply that powers the controller. Wouldn't the right place to get the return path for the data signal be to cross connect the -ve at the power supplies ?