It's that time again. Only two years have passed this time, but I've got another hundred schematics to post. This page contains my computerized electronic drawings from 2008 to 2010.
I would like to think these express my developing electronics skills. Many of them relate to my investigations at the bleeding edge of bandwidth, for instance the pulse generators, high bandwidth amplifiers, logic stages and such. Some of these schematics appear on other pages; I'm not sorting them out, but most are indicated. Most of them should be of better quality than last series, i.e. complete and probably functional. There are still mistakes, so again I offer these with no guarantee. I will offer some description on each this time. (Of course, I offer no guarantee that my description is at all correct either...) So without further adieu, the list.
Statistics: 101 images, 311KiB, average size 3157 bytes. All converted to .PNG. Organized completely alphabetically. This file (index.html), 31KiB.
A textbook TL494 circuit, this produces regulated 24V from a 12V input (automotive, perhaps).
3W capacity, dual +12/-6V outputs. Suitable for floating gate drivers, for instance.
I was asked to design an RF induction heater to heat powdered metals. This was one possible design I came up with. The potential transformer senses the output voltage so that, when one transistor turns off, the output voltage commutates, turning on the opposing transistor for a fixed delay (set by the simple one-shot timer in the top-right corner). By varying on time, the frequency is controlled, thus controlling output power. Control circuit not shown.
I was asked to design an RF induction heater to heat powdered metals. This was another possible design I came up with. Because high speed optocouplers are expensive, this model opts instead for local latch-and-timer logic. The circuit is a little more involved, because it's doubled, and the controller must manage high and low side timers independently (times have to be kept equal). Control circuit not shown.
Adjustable stabilized current source for a 555 timer circuit. This replaces the charge resistor, generating a linear ramp on CT and providing electronically variable frequency.
This circuit was designed to use a quad of droolingly high current darlingtons I obtained from an old school reel-to-reel data tape drive. Unfortunately, it suffered from bias instability because the VBE multiplier never has the right tempco to keep it stable at the low bias currents required to keep dissipation down. (It would probably be stable at several amps output, but that would also dissipate 100W+ easily, and I don't have heatsinks to handle that.)
An elaborate SPDT diode switch. RON is diode internal resistance, the currents must be greater than peak signal current driven through the network, and the control logic levels must be able to sink/source the current.
An elaborate SPDT JFET switch. Although analog switches are usually drawn with a G-S resistor, current sources were used here to improve switching speed. (A side-effect of the small positive gate current is enhancement of the channel due to injected charge carriers, so RDS(ON) should be lower than rated.) To balance this offset, a current of Ig is subtracted from the output. A twin-T design was chosen to maximize isolation at high frequencies (over 20MHz).
Bipolar transistors, particularly those with symmetrical junctions, also make excellent analog switches. Since RCE(ON) is typically much better than a JFET's, only two switches are necessary. Two disadvantages: higher currents are required and symmetrical BJTs are very hard to find.
Typical power-input filter circuit for an automotive environment. The 1.5KE18 protects from spikes, the FR102 prevents reverse connection, and the CLC filter keeps RFI on seperate sides (keeping ignition noise out of the circuit, and filtering the circuit if it's a switching circuit).
Prototypical blocking oscillator circuit. Note: when the transistor turns off, the inductor produces a flyback pulse of undefined voltage. Some load must be applied to the output to protect the transistor.
A simple battery-operated boost supply.
A modification of a posted guitar clipping circuit.
A tutorial on parasitic elements (inductances primarily) in a switching circuit. Shown are the lead inductances between the gate driver supply and ground leads, MOSFET output transistor, motor (making this a buck supply), freewheel diode and supply inductance. Although the consideration of these elements deserves not only an article but many books, the general idea described by the accompanying waveforms shows (left column, top) 1. logic-level drive, 2. gate drive supply bounce (due to shoot-through and gate charging current), 3. gate over- and under-shoot (including miller steps), 4. drain voltage (right column, top), including overshoot due to stray inductance between MOSFET and diode, 5. motor current (inductive, not resistive) and 6. supply voltage, showing shoot-through (due to diode reverse recovery) and overshoot.
While entertaining myself brainstorming LED lighting driver circuits, I created this circuit, which drives a series of LEDs from a higher voltage source using a buck switching regulator. Without the feedback winding, this circuit is simply a linear current source, so it will operate the LEDs safely over a range of supply voltages. With the feedback winding, it becomes a hysteretic oscillator, setting duty cycle as necessary to keep the current waveform within range. As a result, efficiency rises from a resistive 40% up to 70%, a noticable gain. Unfortunately, the BJTs shown are quite inefficient. A MOSFET version is significantly improved. The ultimate downside to this circuit is the voltage dropped across the 1k resistor, which prevents this from being useful at high voltage, such as in a line-operated (160VDC) LED light bulb.
An equivalent circuit for a cadmium sulfide photocell lab. Ri models the leakage current, while RCdS represents the active component. RL is the load resistor. The voltage V is measured over time when the photocell is subjected to a bright flash, thereby measuring the carrier decay in the semiconductor.
A representative schematic for a current-mode induction heater, inspired by the textbook CFL circuit. This circuit was tested briefly at low voltage, but didn't produce useful results; the transistors tended to pull towards a high frequency rather than locking to the tank resonance.
A textbook feedback-corrected MOSFET current sink element.
Left: double diode referenced current sink; right: "ring-of-two" current sink.
A simple differential amplifier circuit, which can be used directly as a simple discrete operational amplifier. Improvements include a more powerful output stage (perhaps a complementary emitter follower), current mirror (instead of Rbe) load for the differential stage, darlington inputs (for reduced input bias current), and so on.
Flowchart for the robot control program in an electronics class.
Flowchart for the robot control program in an electronics class.
Schematic for robot (Olimex AVR board plus additions).
Interface connections for robot (early setup).
Interface connections for robot (final setup).
RTL (Register Transfer Level) diagram of a VHDL project for an electronics class. This circuit consists of two LFSR (Linear Feedback Shift Registers), setup with the same taps and initial seed values. Logic allows a clock pulse to be skipped at the press of a button, thus desynchronizing the registers. (Switch debouncers and clock dividers not shown.) A data input can be XORed with one register, then XORed back; when the registers are in sync, there is perfect correlation and the output comes through unchanged. When out of sync by even one clock cycle in either direction, correlation is approximately -40dB (as measured by spectrum analysis).
Logic level input, speaker driver circuit, for audio demonstration of the pseudorandom generator above. A square wave tone can be sent into the above circuit; when synchronized, it comes through clearly. When out of sync, only white noise is heard.
Simple parallel port EPROM reader.
Self-excited, regulated, current limited, flyback type DC-DC converter. Runs unusually fast, 500kHz at full output. D1 is a small diode, smaller than 1N914; BAT86 may be an acceptable substitute. All 1N914s in the circuit are running at or beyond datasheet limits and should be replaced with MBR160 or similar.
Related to the above circuit, this is a single transistor, self excited forward converter. All 1N914s in the circuit are running at or beyond datasheet limits and should be replaced with MBR160 or similar. The main disadvantages to this circuit are low efficiency (much of which is burned in the timing resistor R2) and lack of current protection (overloading will short out Q1).
A typical RTL (Resistor-Transistor Logic) flip-flop, with edge triggered input A and level triggered input B.
A typical CMOS logic flip-flop, with edge-triggered input A and level triggered input B. A 2N4403 output provides sufficient drive to operate an LED and buzzer.
An example of half-wave rectified sine wave duty cycle. Unlike square pulses, the cusps have a duty cycle which depends on the voltage level you use to define it.
An illustration of choke-input-filtered half wave rectification. If current is discontinuous (as shown), the two hashed areas must be equal. If current is continuous (t4 = t1), then the areas must still be equal, with the result that V2, the average voltage, changes accordingly.
A high current regulated electrolysis power supply, with high efficiency, high power factor and digital control. This page illustrates the power input and PFC section, capable of producing 410VDC at up to 1.2A from any input over the 90-250VAC range. Testing has shown the FAN7527 to be quite unstable, especially with negative resistance loads (such as a switching power supply).
A high current regulated electrolysis power supply, with high efficiency, high power factor and digital control. This page illustrates the forward converter and synchronous rectifier sections. In order to obtain high efficiency at 100A output, a synchronous rectifier must be used, yielding an impressive 0.2V drop maximum at full load.
A high current regulated electrolysis power supply, with high efficiency, high power factor and digital control. This page illustrates the PWM control circuit, which drives the forward converter and synchronous rectifier sections. Testing has shown the FDS8333C's to draw incredibly high shoot-through current during the 40 nanoseconds they take to transition; during this time, about 20A peak is gulped from the supply rail (mostly the 10uF tantalum and 0.1uF bypass at the TL598), drawing anomalously high current and making things heat up. This was later changed to a 1 ohm series resistor supplying the gate drivers, limiting shoot-through current to a quiet 5A.
A high current regulated electrolysis power supply, with high efficiency, high power factor and digital control. This page illustrates the microcontroller and PWM DAC filter sections.
A high current regulated electrolysis power supply, with high efficiency, high power factor and digital control. This page illustrates the remaining microcontroller peripherial circuits: RS232 interface for remote control, a 20x4 LCD display, external PWM drive and 1-of-16 selector for driving external process devices.
A dramatic improvement on the tired old 2N3055 (gag!) flyback driver circuit. Adjustable duty cycle allows adjustable output, while a proper drive circuit and SwitchMode transistor power the flyback transformer efficiently and powerfully. For best results, use a horizontal output transistor, the original FBT primary winding and a 160VDC offline supply (warning: use an isolation transformer).
Gate drive transformer H bridge driver. Since open-collector BJTs do not handle reactive current, a choke-input filter is required to force the dead-time voltage to zero. Downside: fall time is limited by inductor current, which already has to handle magnetizing current. As a result, switching can be dangerously slow.
Gate drive transformer H bridge driver. Since open-collector BJTs do not handle reactive current, a complementary emitter follower supplies drive. Downside: requires bipolar power supplies, and the additional switching stage adds delay and skew. Dead-time drive is limited by the 1k resistor to ground, which is poor.
Gate drive transformer H bridge driver. Since open-collector BJTs do not handle reactive current, complementary emitter followers drive the transformer. For open-collector outputs, current sources were added to improve saturation voltage. Downside: although deadtime drive is excellent, the two series emitter followers have a total 2.8V dead band where the voltage is uncontrolled. As a result, this circuit is unsuitable for low-threshold transistors.
Inspired by CC_Buck.png above, this circuit drives a series chain of LEDs from a high voltage supply, wasting no power on a series dropping resistor. Instead, current is controlled by a buck topology. Unfortunately, while the switching transistor is off, current cannot be monitored, because all nodes in that loop are near +V. To overcome this, a PUT (Programmable Unijunction Transistor) oscillator restarts the single-transistor driver periodically, providing frequency modulation control -- not an ideal solution to the problem.
A typical joule thief LED driver circuit.
A simple common-emitter amplifier circuit to modulate the intensity of a high power LED.
Logic level input motor driver.
This circuit drives a motor in some direction until its current increases above a threshold (about 1.1V / Rs in this case), causing it to turn off.
Representative diagram of an LLC output network, typical of quasi-resonant switching supplies and early induction heater designs.
Series resonant LC network with lossy C.
NPN equivalent to a PNP pushbutton-triggered circuit.
A latch suitable for overload protection, this circuit block comes from my newer induction heater circuit.
Op-amp input clamping circuit.
Op-amp input clamping circuit. The 555 circuit apparently has to run from -9V.
Op-amp input clamping circuit. The 555 circuit apparently has to run from -9V.
A big fat electronic lamp dimmer. This classic circuit (or one similar to it) has been used for decades in high power circuits such as motor drives, where SCRs were the only device available to switch the high power demanded in such applications.
Someone asked for a light-to-PWM circuit. Feeling quite lazy, I drew this in about ten minutes. It will indeed produce PWM, but the frequency varies considerably, which may not be desirable.
This circuit sequences one output to turn on after a short adjustable delay. One output is shown; more LM393 or 339 sections can be added to expand the number of outputs.
Speculative current source element for a power MOSFET amplifier. The fundamental downsides to discrete MOSFETs are transconductance and threshold voltage, which have a substantial range over temperature and manufacturing. To compensate this, a differential amplifier was added, making a simple current mirror as in Current_Sink.png. The high side was singled out for special consideration because it must have floating supplies and its fundamental voltage response is a follower (Vout = VIh - RcIh), whereas the low side will simply be a current sink.
The complete circuit, showing two power mirrors (as above), the mirrors to drive them, and the bootstrap supplies as required. A differential pair provides input gain, while the tail current is adjustable to set total current, thus making this an operational transconductance amplifier (OTA). Unresolved are potential oscillation/compensation issues, and how to set class A bias current independently of peak output current.
A typical precision full wave rectifier circuit, using two op-amps and four diodes.
This circuit was built to observe photoflash response; the compensation was developed based on the square wave response to a superbright red LED (which, amazingly, was just barely visible to the black-cased photodiode).
This simple circuit contains a hysteresis oscillator and pulse width modulation comparator.
Read description here.
Circuit used to program my Z80 scroller.
Representative high- and low-pass RCRC filter circuits.
A single transistor self-excited blocking oscillator, this flyback supply is capable of a reasonable 30W output in a surprisingly small size, at a maximum frequency of 200kHz.
A possible circuit for driving a DC permanent magnet motor with regenerative braking. Note: for regeneration to work, the input PWM must never go completely to zero, otherwise the motor will simply be shorted out. The controller should have a strong integral pole in its response, so that PWM is changed slowly; sudden changes will result in the motor kicking the rider off the scooter (or whatever the project is).
A possible (partial) formulation for the solution to the "resistance of two points in an infinite grid of resistors" nerdsniping problem. Can you guess the approach? (I think I have the coefficients correct!)
A high bandwidth, high impedance, temperature compensated buffer stage suitable for an oscilloscope.
Serial to parallel (e.g., boundary scan) shift register.
This revolutionary (and impressively useless) circuit is the completion of an analogy. Consider: voltage sources are available in two flavors, shunt (e.g., TL431) and series-pass (e.g., LM7805). But current sources are only available in one style, series-pass. These simple circuits complete the analogy, providing a shunt current source. In both cases, a resistor provides a current greater than or equal to the desired output current over the rated range; a current sense resistor, voltage reference and voltage amplifier (VBE and a BJT in the left example; a TL431 and differential pair in the right example) adjust a shunt current to keep the output current constant.
A partial redraw of a circuit for Gregg "T3h" Geek, showing shunt feedback on a circuit.
A high pass filter with "realistic" component values. I think the response is Butterworth?
Soil moisture sensor. Note: the electrodes should be made from thin platinum wires to prevent corrosion. No filtering is applied, so the valve may easily chatter under noisy conditions (how noisy is the resistance of soil, anyway?).
A one-shot resettable flip-flop with pulse output.
Another joule thief circuit.
A remarkably simple "discrete" 12V regulator; functionally similar to a 7812, though the dropout voltage will be lower at low current. (The 2SD1273 specified is a superbeta power transistor, so the dropout at high current won't be too bad.)
One possible realization of a TL494 driving a gate drive transformer. This uses the choke-input filter approach seen in H_Bridge_Transformer_Driver.png.
A fully featured TL494 based forward converter, designed for powering a tube oscilloscope (hence the 6.3V 10A filament winding and 250V output). A blocking oscillator provides startup current, isolated CRT filament power and high voltage. Unfortunately, the blocking oscillator is open loop, so it will dissipate excessive power at high line voltage.
Tube oscilloscope vertical attenuator, preamp and deflection stages. Should provide approximately 5MHz bandwidth, not bad for using no distributed amplifiers, and only about 1/4 the bandwidth of a comparable solid state circuit.
A modification of my old preamp circuit, incorporating a guitar tone stack circuit.
A modification of my old preamp circuit, incorporating a guitar tone stack circuit and "dirty/clean" switch.
A possible realization of tubes in half bridge. In particular, note the screen resistors Rs, required to prevent the screens from going Chernobyl when driving inductive loads (where plate voltage will be negative for a portion of the cycle).
A small tube amplifier power supply, suitable for automotive operation. Capacity about 10W.
One possible realization of a type 2 phase detector. Note the deadband when the comparators are producing synchronized pulses.
This preamplifier is intended for high bandwidth and could be used as a gain-of-10 block in an oscilloscope.
See project description here.
Z80 "Tetris" display. This page shows the microprocessor and support hardware, including a periodic interrupt and I/O decoding.
Z80 "Tetris" display. This page shows the memory and input latch.
Z80 "Tetris" display. This page shows the LED output latches. The two downsides to this approach: the Z80 will be busy almost continuously writing data to the LEDs, and with a low duty cycle (up to 32 x 64 bits = 1/256 duty cycle), the LEDs will be very dim indeed, even if driven with high current.
Web page maintained by Tim Williams. All rights reserved.
