| Noise in audio/video and R/C  systems of Radio Controlled vehicles whether they be robots or elaborate model boats can be caused by a number of  things including poor power  ground schemes and poor power distribution, modulation of the batteries by controllers, signal ground loops, bad audio/video shielding, improper signal levels between audio/video components and Radio Frequency Interference.  This growing page delves into diagnosis and solutions.
         SERVOS: The usual application in an R/C plane
        plugs the 8" long servo pigtail right into the receiver and it works fine.  
        Larger planes and bigger projects use more powerful servos and add extension cords.  
        Often servo operation becomes unreliable and there are a number of products, from RF
        chokes to elaborate optically isolated amplifiers, to cure the problem.  VANTEC discovered a characteristic of one servo amplifier
        Integrated Circuit  that can defeat the best of these solutions and led to a lot of
        frustration.  Although our work was centered around the Futaba S3302 what we
        discovered is common to many IC's and is a possible source of trouble in any three
        wire servo that uses an IC.  Here's the problem: all the power for the servo is
        supplied through the "extension" cord. The best wire has resistance. Four feet
        of "smaller" wire like the #24 AWG used for servo pigtails, has considerable
        resistance. When the servo is commanded to move, the motor starting currents can cause a
        voltage drop across the power wiring.  Bear in mind both the "red" +4.8VDC
        power wire and the ground wire experience this voltage drop and that the ground wire in a
        3 wire servo is part of  the command signal.   Meanwhile the signal wire source
        is referenced to the "real" ground, the ground bus of the receiver.  The
        result is a negative pulse into the servo IC when the servo motor starts.  Monolithic
        IC's don't tolerate negative inputs well because of their intrinsic substrate diode; they
        momentarily cease to be servo chips during the starting pulse. VANTEC has solved this problem by putting a "negative-input-tolerate" buffer
        inside the servo case.  At the end of the 8" servo pigtail works too. 
        Needless to say, our buffer was NOT an IC! Consider: 
        1. Build your own extensions with #18 or bigger power wire, #24 is fine for the signal
        wire.  If you use RF chokes make sure the 2 power lead units are rated for 5 times
        the servo's rated running current.  Minimize the extension length and use
        quality connectors. 
        2. If you use a buffer amplifier avoid IC types.  Install the buffer right at the
        servo's pigtails.  VANTEC has found a common emitter NPN transistor
        to be a universal input circuit.   Use resistors to limit the current to the base
        current and establish a controlled impedance voltage divider.  A small 470 pf cap B
        to E filters noise and RFI. Complete the buffer with another inverting stage. Optical
        isolation is not necessary and can be accidently defeated if attention is not paid to
        ground loops.  Likewise converting servos to "4 wire" with a duplicate
        ground. 
        3. A very large capacitor to bypass the power wire right at the servo's 8"
        pigtails can supply some of the motor's starting current pulse thus lowering the voltage
        drop across the wiring.   5000ufd to 25,000 ufd. Unfortunately the capacitor
        solution is relatively heavy and bulky. Note that  "Farad" memory backup
        capacitors may not be suitably rated to supply high currents. 
        SPEED CONTROLS:  Brushes arc creating Radio Frequency
        Interference. If your motor is not already so equipped install a .001ufd - .01ufd ceramic
        disc capacitor directly across each motor's brushes and between each brush and their motor
        case.  These are yellowish round parts sometimes labeled "102" -"103
        with two leads.  If motor operation still interferes with your R/C set (jittering
        servo's) RF chokes may also be required at the motor in series with each lead.  Use
        chokes rated to the DC current of the motor. 
        NOISE IN PROMOTIONAL ROBOT AUDIO/VIDEO SYSTEMS: Promotional robots
        often incorporate a number of sub-systems together which interfere with each other,
        particularly when they are operated off the same 12 volt Gel Cell battery. Radio control
        systems with servos and PWM-type motor controllers introduce loads on the battery that
        modulate the battery voltage, often at an audio rate. Commercially available audio/video
        components are sensitive to this modulation on the battery and amplify it.  As the
        battery runs down its internal impedance rises aggravating the situation. Although
        wireless microphone receivers can add noise via RF interference and noise, generally
        audio/video noise comes from the battery modulation and poor ground schemes. Some
        solutions: 
        1) pay special attention to the power distribution wiring in the whole robot including
        sneak ground paths.  Place all the audio/video components in a compact area and
        dedicate an isolated chassis mounted with insulating washers for the audio/video
        components. Tie the audio/video component power grounds together at a single isolated
        ground summing node. DON'T incorporate any other components into this ground, just
        audio/video components. Then connect this audio/video ground to the main summing ground.
        Remember heavy motor currents impose a "hum" on even the best grounds.  The
        signal grounds and shields for piping the audio/video from one component to another should
        be employed but watch for sneak ground paths (control elements?). DON'T use the robot
        metal frame to conduct power or signal although a single point ground for the robot frame
        is wise. Wire speakers with individual pairs; DON'T use the frame.  Match signal
        levels; DON'T put a line level output into a microphone input 
        2) employ an LC power filter to feed the audio/video components.  Putting a
        "Farad" filter capacitor on the isolated audio/video chassis and feeding it
        +12VDC through a suitable choke can smooth out power line noise. 
        +12Volt battery-----wire----choke---|---to
        audio video components 
        fused and switched       in series  | 
                                      
        1 Farad 
                                      
        capacitor 
                                           
        | 
                                           
        | 
        Negative-------wire-----------------|---audio video ground 
        battery post 
         
        Make the terminals on the capacitor the the summing nodes audio video components: the
        positive terminal of the capacitor is the audio/video power terminal and  the
        negative terminal is the isolated audio/video ground summing node.  Automotive/Marine
        audio/video components have a fuse incorporated into their power lead; leave these in the
        circuit while tying all the audio/video power wires together at the audio/video Power
        summing node. DON`T incorporate any other components into this power summing node.  
        The David Levy Company at (562) 404 9998 has a suitable 15 amp choke CK-3510 for $3.90 and
        a One Farad 16Volt capacitor PL1F for $77.00. 
        3) Use a separate battery for the audio/video/video components, crude but effective.
          If not a permanent solution it's a good diagnostic tool.    |