Sorry if this is not strictly Docsis related but I'm confused and need a little help. I'll use the Arris 152220 as an example. It's a 1550nm DFB return laser. The specs say the return levels should be 12dbmv/6Mhz. Part of my confusion is that the cmts will adjust return power from the modems based on channel width. I'm also not entirely sure whether or not the levels should be different based on channel loading. Reading the spec literally I assume 12dbmv per 6Mhz is by my understanding already taking loading into account so the level should be 12 whether I'm using 1 channel or 4. To add another twist there is the difference in peak levels between signal generator and spectrum analyzer or sweep gear vs modem levels. What do I do for different channel widths? Should 3.2Mhz channels have more padding than 6.4Mhz channels to bring the levels up?
I'll try an example and maybe someone can help me work through it. If I make a simple setup for the headend transceiver with a 4 way splitter. Just for simplicity lets assume 2 6.4Mhz channels and one 3.2Mhz channel. The fourth port will feed my sweep gear. If I setup my sweep gear to give me 12dbmv at the transmitter input and use inline pads between the headend transceiver and the 4 way splitter to give me 0dbmv at the sweep receiver, is that correct for the 6Mhz channels? Should I put a 3db pad on the port feeding the 3.2Mhz channels to raise their levels or leave them alone because 2 3.2Mhz channels at 9dbmv would put me back at 12dbmv/6Mhz? How would that affect a tech in the field checking levels with a standard cable modem meter?
As you can tell, I'm lost. Help please.
you need to bypass the cmts to set your return transmit levels, using a return sweep. And for optimizing your laser link, you'll need to calculate how much RF you're going to have hitting the laser vs what that laser wants to see. It is set up independently of your return, and yes you need to adjust it based on how many and how wide your return carriers are, plus you need to account for noise (because it adds to the rf power as well)
Here is a pdf you will find helpful:
http://www.sctedakotaterritory.com/uploads/1/6/6/8/16687504/cisco_return...
Thanks, that doc is actually the reason I'm asking these questions. I can't figure out how to correlate the two NPR curves taken from that document. One is in power per hertz the other is total composite power. I don't understand how they convert one to the other.
Don't worry about the charts too much, they're just a guide - your chart would look different depending on your specific equipment and system specifications. Look at the specifications of your equipment and calculate your total possible input power (including noise), and adjust your input power (return padding in the node) to match your equipments optimum input power. Make sure the light is hitting the receiver at the optimum level. Then set your return (modem) transmit level with padding in the headend.
So, say you're transmitting back at 38db through your -20down forward test port on the node with a single CW, You do your math on loss through the node and gain, and hit your return laser at x db(optimum level), once you know that, calculate from there what it would be at that level if it were your full load (you'll be adding db of rf power), then change your return pad (which should be on the output of the gain stage of the nodes return, usually) to bring your final number back down to your original x db. (make sure your padding accounts for any additional noise that may come back on the system) Sometimes an extra wrench is thrown in by the return laser input pad, sometimes the pad is on the true input, sometimes it is after a gain stage, and it makes a big difference where it is, because it will change your x db. Once that is set, you can set your return(modem) transmit level by padding the headend output.
Its important that once the node return padding is set that your techs don't try to change those, because they'll be changing your laser input.
If you have a qam injector you can measure quality as you balance the link (takes two people, one at the node, one at the headend - and make sure you restart the measurement when you make a change because most meters show a measurement over time and not a current instance) Looking at the return with a spectrum analyzer in the headend can be helpful also, because you can see some problems in your return path caused by the laser or receiver if your adjustment is too far off.
At least, this is my understanding, and how I've done it in the past, I know there are several other people here and on the line-man.com forums that have a lot more experience than I do with different kinds of equipment. It can be a confusing topic, and I know I have to re-read how to do it everytime because I don't do it often enough, because once it is set, you should never have to change it unless you change your loading.
Thanks for trying to help. I'm gonna get this through my thick skull somehow. I'm pretty sure I'm making it more difficult than it is but there is definitely a thing or two I'm missing. The only spec (for rf input) on the return tx is 12dbmv/6Mhz. I just can't work out if/how I need to calculate for different channel widths and total load. Apparently I'm dumber than I thought.