Lab 1.2 — Scope DC & Ripple Measurement
← Course 2 syllabus · Module 1 · Prev: « Lab 1.1 · Next: Lab 1.3 »
Goal
Learn to measure both the DC level of a supply rail and the tiny ripple riding on top of it — two numbers on very different scales — with the oscilloscope. The key skill is coupling: DC coupling shows the absolute rail voltage, while AC coupling strips the DC so you can crank the sensitivity to volts-per-millidivision and actually see millivolt ripple on a several-volt rail. You will characterize the WANPTEK bench supply and the cheap MB102 breadboard module, and quantify how clean each one is. For DSP/firmware work a noisy rail is a noise floor you can never filter out downstream, so knowing your supply’s ripple is the first entry in every mixed-signal error budget.
Recommended reading
- O-S&S Ch. 1 — DC (constant) vs. AC (zero-mean) components of a signal, and the idea of decomposing a signal into an average plus a fluctuation. This is exactly what DC vs. AC coupling does in hardware. → Oppenheim, Signals and Systems
- Your Siglent SDS1104X-E manual: vertical coupling (DC/AC/GND), automatic measurements (Mean, Vpp, Vrms), cursors, and the bandwidth-limit (BWL, typ. 20 MHz) function.
- PEI Ch. 7 — Hands-on Electronics: using an oscilloscope (coupling, triggering, probing) at reference depth.
- Prerequisite: Lab 1.1 — the probe must be compensated and set to 10× before any of these numbers are trustworthy.
Equipment & parts
- Siglent SDS1104X-E + compensated 10× probe on CH1 (from Lab 1.1).
- WANPTEK 30 V / 10 A bench supply.
- MB102 breadboard power module (the small module that plugs onto a breadboard and outputs 3.3 V / 5 V from a barrel jack or USB input) + its input source.
- Breadboard + jumpers to bring each rail to a probe point; optionally a small resistive load (e.g. 100 Ω or 1 kΩ) to draw a little current.
Safety & don’t-break-it
- Ground the probe to the rail’s ground, not to its positive. The probe ground clip is earth-referenced. Clip it to the supply/module −/GND, and the tip to +. Reversing them shorts the rail to earth through the clip.
- Stay within the probe’s input range. With a 10× probe on a 3.3 V or 5 V rail you have huge headroom; no risk here, but keep the ratio at 10× so amplitudes read correctly.
- AC coupling does not make it safe to probe high voltage — it only removes DC on the display. Irrelevant at 3.3–5 V, but internalize it now.
- Set the WANPTEK current limit (per Lab 0.1) before connecting any load resistor. Start around 100 mA.
- Don’t confuse “no signal” with GND coupling. If you leave a channel in GND coupling you will see a flat line regardless of the input — a common self-inflicted “my supply is dead” moment.
Background
Any real DC rail is \(v(t) = V_\text{DC} + v_\text{ripple}(t)\): a constant plus a small zero-mean fluctuation (switching ripple, mains hum, load transients). The scope’s coupling selects what reaches the vertical amplifier:
- DC coupling passes everything, so the trace sits at \(V_\text{DC}\) and you read the absolute rail voltage. But to fit a 5 V rail on screen you need a coarse V/div, and a few mV of ripple is invisible.
- AC coupling inserts a series capacitor (a high-pass with a cutoff around a few Hz), removing \(V_\text{DC}\) so the trace re-centers on zero. Now you can select a very sensitive V/div (e.g. 5 mV/div) and the ripple fills the screen.
The measurements you want:
\[V_\text{DC} = \text{Mean}(v), \qquad V_\text{pp,ripple} = \max v_\text{ripple} - \min v_\text{ripple}, \qquad V_\text{rms,ripple} = \sqrt{\overline{v_\text{ripple}^2}}.\]
For a supply, ripple is often quoted as mVpp or mVrms. The bandwidth limit (20 MHz) matters because at full 100 MHz bandwidth the scope also shows broadband noise and pickup that isn’t really supply ripple; enabling BWL gives a fairer, repeatable ripple number.
Procedure
Part A — Measure the DC level (DC coupling).
- Bring the rail to a breadboard point. Probe tip to +, ground clip to GND.
- CH1 menu: coupling DC, probe 10×. Set V/div so the trace sits a few divisions up (e.g. 1 V/div for a 3.3 V rail, 2 V/div for 5 V).
- Open Measure and add Mean on CH1. Read \(V_\text{DC}\). Cross-check against the Fluke DMM DC-volts reading — they should agree closely.
Part B — See the ripple (AC coupling).
- Switch CH1 coupling to AC. The trace drops to center on 0 V (the DC is now removed).
- Increase sensitivity: reduce V/div to 10 mV/div, then 5 mV/div, until the ripple is a few divisions tall. Adjust the timebase to see the ripple’s structure (try 2 ms/div to catch mains-related 50/60/100/120 Hz components, and 2 µs/div for switching ripple).
- Turn on Trigger → Edge → CH1, and adjust the level to stabilize the ripple waveform.
Part C — Quantify it.
- In Measure, add Vpp and Vrms on CH1 (these now describe the ripple only, since DC is removed). Record both.
- Enable the bandwidth limit (20 MHz) in the CH1 menu and re-read Vpp/Vrms. Note how much the number drops — that difference was broadband noise, not true ripple.
- Use Cursors (voltage cursors) to bracket the ripple peak-to-peak manually and confirm the automatic Vpp.
Part D — Compare sources, and load it.
- Repeat Parts A–C for the MB102 module and for the WANPTEK rail. Expect the MB102 (a small switching/LDO module) to be noticeably noisier than the bench supply.
- Add a 100 Ω or 1 kΩ load (set the WANPTEK current limit first) and see whether ripple grows under load.
Deliverable & expected results
For each source, one DC-coupled capture (showing \(V_\text{DC}\)) and one AC-coupled capture (showing ripple), with the measurement badges visible. Predicted DC is the set/nominal value; ripple is source-dependent and is what you are characterizing (leave predicted ripple as an order-of-magnitude expectation).
| Quantity | Predicted | Measured |
|---|---|---|
| WANPTEK rail, Mean (DC) | set value (e.g. 5.00 V) | … |
| WANPTEK ripple, Vpp (BWL on) | low, ~single-digit mVpp (bench supply) | … |
| MB102 3.3 V rail, Mean (DC) | ≈ 3.3 V | … |
| MB102 ripple, Vpp (BWL on) | higher, tens of mVpp (cheap module) | … |
| Ripple Vpp with BWL off vs on | off ≥ on | … |
Analysis & reconciliation
Confirm the DC-coupled Mean matches the Fluke DMM to within both instruments’ accuracy — if not, suspect the probe ratio (Lab 1.1) or a GND-coupled channel. For ripple, explain the BWL-on vs BWL-off gap as broadband noise/pickup removed by the 20 MHz filter, and identify the dominant ripple frequency from the timebase (mains-harmonic hum vs. high-frequency switching). Compare the two sources quantitatively: the bench supply should be a cleaner rail than the MB102, and that difference is exactly the kind of number that becomes an ADC noise-floor limit in Module 5. Note whether adding a load changed the ripple and reason about why (regulation, output impedance).
Going further
- Add a decoupling capacitor (e.g. 100 nF ceramic + 10 µF electrolytic) across the noisy MB102 rail and re-measure the ripple — a direct preview of why every chip in later labs gets a bypass cap.
- Use the scope’s FFT math on the AC-coupled ripple to see the ripple spectrum (mains harmonics vs. switching frequency); a warm-up for the FFT labs in Module 6.
- Log the ripple Vrms vs. load current and plot it (host Python) to characterize the supply’s regulation.