Curve Tracers — Vol 3: How They Work — Curve Tracers

The semiconductor tracer, the pulsed-HV tube tracer, the octopus — and what every control does

Figure 1 — The three-block architecture again, now as the thing to trace with your finger while reading the controls below. Source: hand-authored SVG.

Vol 1 gave you the three blocks — step generator, sweep supply, X-Y display. This volume opens each of the three main kinds of tracer to show how those blocks are actually built, then walks the standard front-panel controls so you can sit down at any tracer and know what the knobs do.

3.1 The semiconductor curve tracer: stepped base current, swept collector

This is the Tektronix 575/576/577 and Heathkit IT-3121 architecture, and it is the one to understand first. Two things happen at once.

The collector sweep. The collector supply produces a rectified-AC ramp, not a regulated DC voltage. In the classic designs the mains-frequency sine wave is rectified into a series of half-sine humps; each hump sweeps the collector-emitter voltage VCE from zero up to the peak set by the front-panel control and back to zero. Because the sweep is derived from the AC line, it repeats at line frequency (or twice it), which is exactly the refresh rate you want for a flicker-free display. On the Heathkit IT-3121, for instance, the collector supply is deliberately pulsating DC from a transformer and diodes — a simple three-transistor Darlington buffer varies the peak from 0 up to 40 V or 200 V depending on range — rather than an expensive regulated high-voltage supply. Sweeping from a rectified line also means the device only sees full voltage momentarily each cycle, which helps keep dissipation down.

The base staircase. Synchronized to the sweep, the step generator advances the base drive by one increment between sweeps. A counter (one TTL counter in the IT-3121) clocks a small DAC, and the DAC output feeds a precision current source that turns each voltage step into an equal base-current step — the IT-3121 offers steps from about 0.002 mA up to 10 mA per step across its ranges. Set 5 steps and the generator dwells at 0, then 1, 2, 3, 4, 5 increments of base current, one per collector sweep. The result is Vol 1’s fan: the beam paints a curve at IB = 10 µA, the staircase clicks up, the beam paints the IB = 20 µA curve above it, and so on. Reverse the polarity switches and the same machine tests PNP devices with the curves in the opposite quadrant.

A subtle but important point: for a bipolar transistor the natural control variable is base current, so the step generator is a current source. For a FET (JFET or MOSFET) the gate draws essentially no DC current, so the control variable is gate voltage, and the step generator is switched to produce voltage steps instead. A tracer that can do both — like the VBA Curve Tracer, which offers stepped current or stepped voltage — can characterize BJTs, JFETs, MOSFETs, and Darlingtons from the same front panel.

3.2 The tube curve tracer: why it goes pulsed

A vacuum tube wants very different conditions: hundreds of volts on the plate, a screen supply for pentodes, a negative grid bias as the control, and heater power. The classic Tek 570 did this the direct way — real HV supplies and a swept plate voltage — which is why it is a large, heavy instrument.

The modern DIY tube tracers solve the same problem with a pulsed high-voltage technique, and this is the single cleverest idea in the category. Ronald Dekker’s uTracer designs charge a large reservoir capacitor to the desired plate (and screen) voltage from a compact boost converter, then dump that voltage onto the tube as a short pulse — on the order of a millisecond — while sampling the plate and screen current during the pulse. Because the tube conducts only for that brief moment, the average power the little instrument must handle is tiny even though the instantaneous plate voltage and current are those of a working power tube. That is how a uTracer fits a several-hundred-volt tube characterizer onto a palm-sized board: it never has to dissipate the full plate power continuously, only source it for a pulse. The measured points are digitized and the curves are drawn in host software on a PC rather than on a built-in CRT.

The pulse approach also protects the tube — a shorted or gassy valve conducts for a millisecond, not indefinitely — and it lets the same hardware sweep a family of curves by stepping the grid bias and repeating the plate pulse at each grid setting. The uTracer6 and uTracer NXT and eTracer dives in this category go into the reservoir sizing, the current sensing, and the safety interlocks; the takeaway here is pulsed HV = big voltages, tiny average power, host-side plotting.

Because these instruments genuinely apply several hundred volts (and store real energy in that reservoir capacitor), they demand the same HV bench discipline as any tube gear — see this project’s shared bench-safety notes.

3.3 The simple case: the “octopus” component / signature tester

Figure 2 — Two-terminal "signatures": an open reads as a horizontal line, a short as vertical, a resistor as a sloped line, a capacitor as an ellipse, a diode as an L-shaped knee. Source: hand-authored SVG.

At the opposite end of the complexity range sits the octopus — the poor man’s curve tracer, so called because a shop-built one tends to sprout wires. It has no step generator and no swept high voltage at all. It simply applies a small AC excitation — often about 1 V (or a few volts) at line frequency, current-limited by a series resistor — across a two-terminal device, puts the voltage across the device on the scope’s X input and the current (sampled across the series resistor) on the Y input, and lets the scope draw the resulting Lissajous “signature” in X-Y mode.

You read an octopus by shape, not by numbers: a resistor traces a straight sloped line whose angle encodes its value; a capacitor (or inductor) traces an ellipse, because current leads (or lags) voltage; a good junction traces an L-shaped knee; an open reads as a flat horizontal line and a short as a vertical one. Its killer feature is in-circuit testing: because it only wiggles a volt or so, you can probe a component (or a whole node) without desoldering it and with the board unpowered, and compare a suspect board point-for-point against a known-good reference — the technique called analog signature analysis. It will not give you β or gm, but for finding shorts, opens, leakage, and blown junctions on a dead board it is fast and forgiving. Both the Heathkit-style semiconductor tracer and the octopus are worth having; they answer different questions.

3.4 What the standard controls do

Sit down at almost any semiconductor curve tracer and you will find the same clusters of controls. Knowing them is most of the battle.

The step / base controls set up the family: steps per family (how many curves, typically 0–10); step amplitude (µA or mA per step for a BJT, volts per step for a FET); step polarity / offset (which quadrant, and — on tracers that support it — a positive offset so you can bias a MOSFET above its threshold). More steps and finer amplitude give a denser fan; fewer, coarser steps make a specific parameter easier to read.

The collector / sweep controls set the horizontal extent and the power the device sees: peak sweep voltage (the maximum VCE/VDS the ramp reaches), the voltage range (e.g. the IT-3121’s 40 V vs 200 V ranges), and polarity (NPN vs PNP, N- vs P-channel).

The series / load resistor — often labeled series resistance, load, or max peak power — sets a resistor in series with the collector sweep. This does two jobs: it limits the peak current (and therefore the dissipation) the device can be driven to, protecting it from destruction, and it sets the load line the device is tested against. This is the control that keeps you from vaporizing the DUT, and the discipline is always to start with a high series resistance / low peak voltage and open the throttle only as far as you need.

The display / scaling controls set what the picture means: vertical (current) sensitivity in amps or milliamps per division, horizontal (voltage) sensitivity in volts per division, and on instruments that plot to your own scope, the corresponding V/div you dial into the oscilloscope’s X and Y channels. On the vintage Tek machines the alphanumeric readout printed these scale factors on-screen; on a Heathkit-plus-scope or a DIY-plus-scope rig, you are responsible for knowing them, because a curve is meaningless until you know how many milliamps a vertical division represents.

With the architecture and the controls in hand, Vol 4 turns to the actual skill: reading the family of curves the instrument draws — pulling β, gm, the Early effect, breakdown, and saturation off the screen, and using the tracer to match devices.