Misc.(Edit)

Radio Wave Production?

• Electric current → magnetic field ∴ oscillating current → oscillating magnetic field
• Not "EM wave" (as Maxwell supposed)
• Electric oscillation can only occur in a conductor
• Faraday Effect: Rotation of the plane of polarization of light by application of a magnetic field
• Electric fields are created when:
• A charged particle (negative or positive) creates an electric field around that particle
• An oscillating magnetic field creates an oscillating electric field within a conductor.
• Magnetic fields are created when:
• A charged particle (positive/negative) has magnetic properties and acts as a small magnet.
• Charged particles are aligned in a material, a magnetic field will be created around that material.
• Flowing electrons cause the electrons to align (but there is no direct connection between current and creation of a magnetic field)
• An electron's electric field is at right angles to its magnetic field---so the magnetic wave emanates perpendicular from the current direction

Ground

• Understanding Ground Loops
• GFI tester
• Don't just live with (60Hz always?) hum
• Start by using single power strip instead of two different plugs
• Actual details and explanation at the Wikipedia: Ground Loop page.

Testing An Electronic Circuit - Making A Circuit Board

• Static checks before live checks.
• Check +V
• Check all ground locations
• That's all (it's a short video. Notable bonus is the way the text twiddles in at the start. Obskur4tionz Styl3.)

Tube Radio Repair Essential Tools

• Isolation transformer
• Multimeter (obviously)
• Variac
• When starting up for first time, smoke test: slowly go from 50V to 75V to 100V to 117V, pausing 3-4 minutes in between each step

No video for this one, just a tip:

• You can find AAAA batteries inside 9V batteries

SMD Soldering

• Several techniques, all you need to know.

Oscilloscopes(Edit)

Choosing an Oscilloscope

• 10MHz Bandwidth, 100MHz+ = excessive
• 100Ms/sec = good
• 2-channel = good, 4 = overkill
• Digital: Analyze voltage, freqs for you, Save waveforms for later
• Stay away from USB scopes
• Recommends PDS5022S
• Shouldn't pay more than $400, used

Sine wave vs. Square wave

Transistors(Edit)

Very Cursory Presentation

• Touches on how the gate is able to make a bridge between the collector and emitter

Testing Transistors by a Very Fast Japanese Guy

• Model # determines NPN or PNP
• Determine pins by:
• Base = leg that reads 1Ω to both sides
• Collector = leg that reads no continuity from Base
• Emitter = leg that reads continuity from Base

Haltingly Presented Analysis Prep and ...next, Common-Emitter Analysis

• i_C = ϐi_B
• V_BE = V_ɣ
• Cutoff: Open, V_BE < V_ɣ iC=0 because i_B=0
• Active: Just i_C = ϐi_B
• Saturation: Short, V_CE=0, i_C = i_short
• (These videos have good info, but the presentation is excruciating. Too slow at times, too fast at others. Getting to the information is like playing Frogger, there has to be a better set.)

DIY:

• N-MOS
• (Turns on when gate voltage is applied)
• P-type wafer, 100 crystal orientation, into furnace w/steam until oxide layer is 5k-6k Angstroms thick, color chart will tell you when you're there (~6hrs)
• Active area etch silicon dioxide with HF acid ("Rust Stain remover") and vinyl mask
• Spin Phosphosilicate on wafer (e.g. using CPU fan as turntable), then kiln @ 1000 deg C, 5 min, which puts a high concentration of phosphorous onto the silicon layer
• Cool down, back in acid, etching away film, done when water beads up in active areas
• Intermediate oxide layer + drive in: 15 min in steam in furnace, stop steam, 25 min drive-in phosphorous, lowering resistance and forming PN junctions
• Etch gate oxide area (.8-1k Angstroms). Mask, open gate region. Furnace 1k degrees (w/o steam), grow thin layer (use color chart again). "Most difficult part"
• Make contacts: another mask, cut through Source/Drain, then use e.g., conductive epoxy
• Add substrate ground to get rid of parasitic PN junction

Amplifiers(Edit)

iPod Amp from Salvaged TV Parts

• TDA2616Q + 2A regulated 6V DC PS + Speakers + Tupperware

WA2KWL (Jonathan Gordon), who is the man:(Edit)

User Page

Digital RFI

• AM Radio detects RFI
• Square wave is fundamental + odd harmonics (the faster the rise/falls, the more the harmonics)
• 166 MHz (6'), 1 GHz (1'), will find 1/4-wave, 1/2-wave and full-wave wires/traces to use as antennas
• Spread-spectrum clocking (enabled in BIOS) reduces RFI output.
• TRIAC dimmers/speed controllers clip AC sine wave to change duty cycle. The chops create sharp edges (aka, harmonics) == Divide by triangle wave?

Tips for troubleshooting a complicated analog circuit on a solderless breadboard

• Run wires as short and parallel to each other as possible.
• Color-code wires
• Keep component leads far enough apart that they don't short
• Try to avoid going over the top of ICs
• Be aware of adjacent rows' stray capacitance
• Diagnose in stages (!) Either one at a time in sequence, or BinaryChop Debugging

D-type Flip-Flop (computer PSU)

• Shows how the momentary-on switch resets a flip-flop to start a PSU
• Also mentions the Data and Clk inputs on the flip-flop that could work as a power-off timer.

Test for Electrical Circuit Continuity

• Doesn't like mini phono plugs (because they can short during insertion), does like barrel plugs (which don't)
• Set MM to continuity, place one lead on plug and the other on the opposite wire termination, then hold them and move the wires around at different angles to test for intermittent connection
• To find ground, connect one lead on the shielding of a serial port terminal, then probe using other lead. Anywhere on the circuit you get a beep is a ground connection.
• Also, continuity testing fuses (out of or in-circuit)

AC Adapters

• AC-to-AC Unregulated: Just a transformer; measures higher without load than with.
• AC-to-DC Unregulated: Transformer, bridge rectifier; measures higher without load than with.
• AC-to-DC Regulated: Transformer, rectifier, filter caps, voltage regulator.
• AC-to-DC Switcher: Bridge rectifier, HV filter cap., stepdown transformer, pulse width modulator, switching transistor, rectifier diode; Probably lighter than the transformer version.

LED Semiconductor Physics Made Easy

• Valence shell <-> Band Gap <-> Conduction band
• Energy required to move from valence to conduction band defines insulation/conduction. Home run hit vs. Cannon vs. Thump.
• Silicon doping:
• Silicon has 4 valence electrons, creates strongly bound electrons
• Boron, aluminum, gallium = 3 valence electrons; dope with these and you have one electron "missing", hence P-type
• Nitrogen, Phosphorous, Arsenic = 5 valence electrons; opposite: N-type
• Electrons going from N-type to P-type gives off energy: different wavelengths for different band gaps
• Doping with other substrates gives different band gaps, e.g., Gallium Arsenic, Aluminum Gallium Arsenide (=Red: 620-780nm), Aluminum Gallium Phosphorous (=Green: 490-570nm)
• Wikipedia elaborates ( http://en.wikipedia.org/wiki/Band_gap )
• Gap is measured in Volts.
• The reverse: "the band gap determines what portion of the solar spectrum a photovoltaic cell absorbs."

Ethernet 10Base-T Manchester Encoded Signaling

• TX+ (non-inverting) TX- (inverting)
• Non-return to Zero Encoding
• TX+ rising = "1" bit, TX+ falling = "0" bit
• 100ns period (10MHz) possible, but this only happens on consecutive bits (1 to 1, 0 to 0). Whenever there is a change (1 to 0, 0 to 1) the signal is 5MHz wide (!).
• Spectral Analysis verifies that most of the signal is near the 5MHz range---which reduces bandwidth.

AC Copper Wiring vs MotherBoard Copper Trace

• House wiring has no transmission line properties---just connects switch to load
• U.S. AC = 60Hz cycle, 16.67ms (0.0167s) period, 5 Mm length (New York to California)
• Because the wavelength is so much greater than the wire length, the wire is not a transmission line (only presents a small amount of resistance)

• When you get to the 7th+ harmonics, PCB traces are going to act like antennas, like a series of inductors and capacitors
• Xc (capacitive reactance) = 1/(2 pi Freq C)
• XL (inductive reactance) = 2 pi Freq L
• As frequency increases, Xc decreases (thus more voltage is shunted to ground), and at the same time XL increases (attenuating the voltage more).
• Causes timing problems, motherboard = unstable.
• Also, on traces, electricity's speed of propagation slows down to 50% of C.

Power Factor

• True Power (e.g., watt meter)
• Apparent Power (e.g., Multimeter "amp" measurement)
• Power Factor = True Power / Apparent Power

• Dealing with harmonics from switching: Filter capacitor integrates square waves into sine waves (see 7:48 to end of video)
• Dealing with Inductive Amps: A resonant capacitor-to-ground keeps inductive negative amps from flowing back to generator

Computer PSU

• ATX 20+4 (the 4 break away for use on 20-pin)
• One of the pins has 2 orange wires: one is the supply and the other is a sensing wire to make sure it's exactly 3.3V
• Multiples of same voltage because connectors have a significant voltage drop
• Causes of burnt-out PSUs:
• Long duration voltage surges (solution: Line-interactive UPS)
• Motherboard capacitor fails
• Lightning (solution: Surge protector)
• Faulty ATX Connector Pins
• Excessive moisture
• Excessive current draw
• Leads to heating
• Caused by too many peripherals
• Best way to test is to feel for excess heat with your finger
• One solution is to use different connectors (e.g., use both PCI Express connectors rather than the "Y" one)
• As processor speeds increased, circuits went from 5V to 3.3V and then further: the motherboard gained its own switched-mode power supply fed by the 4-pin connector.
• 4-pin Molex HDD connector
• Yellow = 12V (powers motor)
• Red = 5V (electronics)
• Testing
• Ground Green (power on) to Gnd to turn PSU on
• Have to cut to test Amps (= in series)
• Risk, Control of PSU Repair
• SPST Switch
• Gnd goes directly → Chassy
• Hot (Blue) and Neutral (Brown) → RFI Filter → PCB → Bridge Rectifier → High-voltage Filter Cap (the big one) → High-voltage Switching Transistor (50-100kHz) and Primary Windings of Transformer (acts as an Isolation Transformer) → Secondary Windings (e.g., one 12V and one 5V)
• Division between high-voltage ("hot common") side of PCB and low-voltage (earth ground).
• Safety: Isolation transformer between PSU and wall when tinkering
• Variac = Variable AC Voltage Controller (like a volume controller). Not isolated unless isolation transformer is added in.

Cable TV Signal

• 100mV
• QAM = Quatrature Amplitide Modulation: Phase-shifting, Amplitude variations
• Each channel is 6MHz wide, with a spike for each carrier

TV Repair:(Edit)

Dave's TV in Grants Pass, OR

• Video #1
• Checking for bad caps by looking for bulges, but it's better to use a capacitor checker (reputable ones: Cap 88 Equivalent Series Resistance (ESR) Meterby Electronic design; Capacitor Wizard; Blue ESR Meter)
• Replacing with different capacities is iffy, but you can go up in voltage (and this can sometimes be a great idea)
• Remember capacitance decreases in series and adds in parallel
• Make non-polarized capacitors by putting two polarized in opposite directions (?? - verify)
• Use a flashlight to see through the back of a PCB
• TV repair case histories (!) Something like this for radios?
• Video #2
• Hairdryer to heat up components to test for intermittent overheating effects
• Freeze spray for the opposite
• Telephone earplug-inductive coil amplifier device - can listen for magnetic oscillations (e.g., 60Hz hum, switched-mode power supply)
• Small neon bulb - can be used to detect high voltage (e.g., flyback transformer)
• Video #3
• Don't buy excessively cheap components
• Video #4
• Bad fuse is probably because of too much current draw within the circuit itself (as opposed to a fluke surge)
• Sub in a circuit breaker while the fuse is still popping
• Variac to slowly ramp up power
• Light bulb in series on AC line as debugging aid (demonstrates short circuit)
• Inductive Ammeter (AC only)
• Looping wire around clamp multiplies the amp reading (= more sensitive reading)
• Video #5
• Heat temporarily improves a capacitor's quality rating, freezing lowers it
• Resistors increase under heat
• If taking a hairdryer to a circuit is enough to make it work, then it's probably a bad capacitor
• Video #6
• Switched-Mode Power Supply Troubleshooting
• Choosing correct Ground Reference Points
• Different ground symbols (e.g.: hot-ground/3-lines, isolated-ground/hollow triangle)
• Separations in PCB can be a good clue
• Measure on rectified side, not on high-frequency AC side
• Video #7
• Finding bad solder joints
• Requires attention to detail, magnification (e.g. loupes, old camera lenses)
• A clue is darkened areas where excess heat may be
• When in doubt, just re-solder
• Sand bad leads/pads, or possibly make your own access and use a jumper wire
• Always discharge big caps before working with circuit
• May be cracks if board has been jolted/dropped - can be very difficult to spot
• Video #8
• Shut-downs are a common circuit response to any problem
• Talks about LCD backlights a bit (e.g., recommendation to salvage them off of old monitors)
• Good debugging sequence starting at 3:11 - No power at terminals, found transformers, turned out to be xformer controlling xistor
• Check transformers for shorts with a coil tester AKA coil ringer (DIY?) (see video #10 tip)
• Video #9
• Varying brightness on TV = varying beam current = varying power draw
• Hot-running IC might be a problem (try freeze spray)
• Compare power draw to back of TV's expected draw (varies with picture, though). High draw is probably a component problem.
• Watt-meter + Variac = handy. Circuit problems can cause jumps in power disproportionate to voltage change (e.g. PSU problem)
• Slowly turning up variac is a good check of voltage regulator - by definition it should not change much with increased voltage
• Wear ground strap (or manually neutralize)
• Video #10
• Checking transformer with a VOM
• Test resistance across coil windings for a sketch
• A coil tester is better
• Ensure no shorts between primary and secondary windings
• Switched-mode Power Supplies are common problem
• Feedback circuits are the complicated part of an SMPS
• Knowing as much as you can (see link in Video #6) about SMPS and inverters
• Very servicable
• Think in terms of block diagrams (starts at 5:38)
• E.g.: Power Supply board, microcontroller unit (MCU), T-con (controller board), driver ICs (drive screen), screen, backlight & its inverter board, feedback circuits (tell PSU/inverter board to shut down if problem), audio amp, Tuner,
• Know approximate voltages at various block
• Learn when to bail out or escalate
• "Make yourself like a flow chart" (!)
• Playing with LCD, Polarized Filters
• Also: Watertown TV Repair

Then there's John from Preher-Tech, who moves faster and zips past more details (though he does drop more advertisements):

• Intro to major circuits PSU, Inverter, Main Board, LCD Controller Board, Misc
• Mentions CCFL (cold-cathode flourescent) for backlight
• Feedback circuits everywhere to shut down in event of trouble
• LCD Controller Board ("T-Con"?)
• Part 2, more misc + LCD itself, Backlight
• ...the follow-up videos are long overdue.

Inductors(Edit)

Electroluminescent Backlight (Or, how to get 100V AC @ 150 Hz from 5V DC)

• Video starts at 1:14 and ends at 2:00, or just skip straight to their web site
• Inductors resist change, so if their output is suddenly cut off (as in when the first MOSFET is switched off, controlled by the AVR microcontroller's internal timer), a voltage spike occurs while the inductor "tries to keep the current flowing".
• When this happens, the monster diode passes the kick-back voltage into the capacitor (I don't fully understand that storage part. Capacitors hold a charge, but how exactly does that relate to a high voltage?). The reverse-bias of the capacitor works as a peak detector.
• As the other MOSFET oscillates (at a much lower rate, also controlled by the AVR), the negative terminal of the electroluminescent device is pulled low and current flows through it.
• A higher frequency for the first MOSFET gives less time for the inductor to build its magnetic field, so the resulting voltage spike is smaller.
• Driven by this C code

Ellsworth:(Edit)

• Debugging Bus Contention - Electronics Repair Techniques

Just interesting (as in, not necessarily useful):(Edit)

• Commodore 64 Laptop Demo
• Best of the SpaceLab

Totally Irrelevant:(Edit)

• Search by Image
• Villager TSBs