Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Noobe, need some help in power drop along long wires

  1. Mar 17, 2006 #1
    Hi,

    I am new to the forums here, I have a radar detector that I want to hardwire and remotely view the lcd in my rear view mirror. I'm using cat5e cable to make the connections, the length is about 8'. I finally got the lcd in the rear view mirror and soldered all the connections. It works, but, as the text scrolls it becomes weak and dimms to the point that it is not legible. I am not an expert at electrical stuff but I think the voltage is dropping and the text wont show.

    My question,
    is there an electrical component that I can wire into the cable or on the lcd to make sure the voltage or power does not drop to the point that the text is not legible?

    Any help would be great.

    thanks
     

    Attached Files:

  2. jcsd
  3. Mar 17, 2006 #2

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    So all you did was cut the LCD out of the radar detector, then extend the wires that originally connected the two pieces? From the looks of it, you are soldering your cable directly to the outputs of an integrated circuit.

    This may not work, because a long wire requires a driver, a component that essentially amplifies a board-level signal and is designed to push a signal over a long length of wire (with its associated capacitance and resistance). Cat 5 cables have a characeristic impedance of about 100 ohms, while a copper trace on a board may easily have an impedance of less than an ohm.

    If you know what kind of signals are on the signal wires (TTL perhaps?) you can buy off-the-shelf drivers that will solve the problem.

    Without knowing more about the nature of the signals that are being transmitted across your wire, this is about all the advice I can give you.

    - Warren
     
  4. Mar 17, 2006 #3
    This is SOOOOOO wrong. Cat 5 may very well have a surge impedance of 100 ohms. But how the heck can you say 'a copper trace on a board may easily have an impedance of less than an ohm'? No circuit board trace, strip-line or otherwise, is likely to have a surge impedance of less than an ohm. Now the resistance of the copper itself darn well better be less than an ohm, but these two things are not the same.
    -
    Blazin, I don't know where exactly you split things apart, but if the frequencies feeding the LCD display are low enough you shouldn't have to worry about a line driver. The most that will happen is that the corners of the square waves will be slightly rounded off. However, if the frequencies are high enough and the length of the Cat 5 is long enough then yes you will have problems with loss. However, this is not because of the resistance of the wire in the Cat 5.
     
  5. Mar 17, 2006 #4

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    AverageSupernova,

    It's also irrelevant, as is your comment about high-frequency attentuation. It's only 8 feet of cable. The high-frequency attenuation of Cat 5 is essentially zero below 1 MHz, and it takes tens or hundreds of feet of Cat 5 to create any appreciable loss below 10 MHz. (By the way, I love the way you aggrandize the harmless mistakes of others and then proceed to make bigger mistakes of your own.)

    I also note that you didn't actually give any suggestions on how he should go about fixing the problem.

    The reason why I brought up characteristic impedance is beecause I suspect that his "design" is suffering from reflections. If he used an approrpiate line driver, with an appropriate output impedance, he wouldn't have reflections and would likely solve the problem.

    You could try to solder a 100 ohm resistor between the signals and ground on your radar detector's board. It wouldn't be quite as robust a solution as a line driver, but you'd at least get the impedances a little better matched.

    Ideally, you'd have an oscilloscope available to you. One glance at the scope would tell you whether this a reflection problem or a high-frequency loss problem.

    - Warren
     
    Last edited: Mar 17, 2006
  6. Mar 17, 2006 #5

    berkeman

    User Avatar

    Staff: Mentor

    My guess is that the problem is the capacitance of the 8' of Cat-5 cable. Twisted pair like that is about 13pF per foot, and the capacitance of the LCD that th LDC driver generally drives will be on that order, I think. So the capacitive load presented to the LCD driver is way higher than it was optimized for, and hence you are getting a smaller drive signal at the display itself.

    As chroot suggests, if you can figure out which are the drive signals and how they work, you may be able to insert some buffer drivers before the 8' cable to boost the output current capability of the LCD drive IC. I would suggest trying to get the IC part number off the LCD driver chip, and looking up its datasheet for more info. Also, I would back-terminate the driver lines at the buffer IC, before the signals are driven into the twisted pair wires. Also, you will want to choose the twisted pair line assignments so that each signal has a paired ground with it in each twisted pair. I understand that the LCD drive signals are pretty low speed, but it's good hygene to think about being careful how you run signals down wires to a remote destination.
     
  7. Mar 17, 2006 #6

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    berkeman,

    My first guess was capacitance, actually. I went and looked up the specs of Cat 5e, and saw 5.6 nF per 100 meters, and decided it was very small and irrelevant for an 8' length. If I had just typed it into my calculator and realized that's 17 pF per foot, my face would have looked like this: :eek: :rofl:

    Most ICs are designed to specified driving 15 pF loads, which is typical of interconnects between ICs; a 135 pF load is going to be very tough on the chip, and may even damage it (by heat dissipation) over time, depending upon how it was designed. It'll have to move roughly 10 times as much charge per volt as it was designed to move. Given that it's current limited, signal transisitions will be ~10 times slower than they should be.

    My advice remains the same though: get a line driver.

    - Warren
     
  8. Mar 17, 2006 #7
    Well thanks for all the suggestions so far but most of what has been discussed makes little to no sense to me.

    But my understanding was that cat5e is a straight through cable if the standards are not mixed eg 568a and 568a. A crossover would defin not work in my application.

    Now as far as lcd's go, I did read up a little about them and apparenty they are basically point on a 4x4 or #x# graph where the electrical signals lights the dot on the matrix and thus it is illuminated. No what I do not understand is how they work in the physical sense, ie is + on x axis and - on the y axis and based on these the matrix it lit. I could not find this to be true on mine since the leads are I III II II. Thus I would assume the first 4 matrix boxes are one whole section and the second part is on another.

    In any case, I would appreciate the education if anyone has the time to explain.

    On the install I did install the wiring and the radar detector and lo and behold it works. the wire was shortened to about 4' since I was able to tuck it in the headliner along witht he sound lead for the voice alerts speaker in the rear view mirror. The text is not legible when it is scrolling and bright text, but when the display is only showing the direction and the mode it works fine.

    I attached a pictures.

    So to recap I need somehting called a line driver, I shall research this more. One last q, this line driver, would this go on every wire? Or would this go back to the issue of how lcd works or how this particular lcd works as far as + and - of the wires like I asked above.

    thanks again
     

    Attached Files:

  9. Mar 18, 2006 #8

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Several of your wires carry power; they don't ever change, so capacitance is not an issue, so you won't need any line drivers for them.

    The signal wires, however, would each need their own line driver. You can buy chips with several line drivers in a single package.

    - Warren
     
  10. Mar 18, 2006 #9
    I see, what is the technical term for a line driver? I looked it up in my jameco catalog and could not find anything.

    Also, this line driver is essentially a chip, component... that measures the voltage across the wire and regulates it to keep it at a constant rate?

    Thanks
     
  11. Mar 18, 2006 #10
    Ok guys, let's take a step back. We still really don't know the type of signal going between the 2. I'm guessing that the driver is on board to the display. I have worked with displays of this type where there is just serial data sent to the the display. If the characters light up correctly then I'd say all the data is getting to the display correctly. Is this display LCD or vacuum florescent? It looks like vacuum florescent to me. Before we send this guy off looking for parts he doesn't know what are let's evaluate this a little better. Incidentally chroot, the reason why I didn't give much of a reply in my first post concerning the actual problem is that I felt there was not enough information to base an accurate and helpful response. No answer is better than an incorrect one.
     
  12. Mar 18, 2006 #11

    chroot

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    The technical term is "line driver."

    They're integrated circuits (chip), and are essentially amplifiers. They're designed to provide large currents to drive large capacitive loads (like long cables), with the appropriate impedance to match that of your cable and eliminate reflections.

    There are thousands of line drivers and buffers available from distributors like Digi-Key and Arrow.

    - Warren
     
  13. Mar 20, 2006 #12

    berkeman

    User Avatar

    Staff: Mentor

    Good point by supernova. The picture appears to show an active green dot matrix display, not a passive liquid crystal display. My comments about capacitance only apply to directly driving an LCD, where each pixel (at the intersection of each row and column) presents a few pF of capacitance to the driver.

    If there is a simple serial link between the radar unit and the display driver electronics, then I think the most likely cause of the problem is bad signal integrity due to reflections in the 8' length of cable. Even if the overall data rate is not real high, ringing on any clock signals can cause double-clocking of the data, or other problems in getting the right data across. That's why I would still back-terminate the signals that are being driven from the radar unit to the display unit. You need to figure out what lines do what, and be sure to make good clean twisted pair connections from the radar unit to the display unit. One signal line and one ground per twisted pair, and back-terminate each driven line from the radar unit with about 50 Ohms or so (the Zo of the twisted pair minus the output impedance of the driver). This should clean up the signals enough so that you have good signal integrity and clean clocking of the data at the display.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Noobe, need some help in power drop along long wires
Loading...