Keysight vs Rigol Oscilloscope: Lab-Grade Accuracy vs Value, which Is Right for Singapore Engineers?

The Oscilloscope Market Has Changed, but Not in Every Way That Matters

Ten years ago, a 200MHz oscilloscope with a modern user interface and deep memory cost SGD 6,000 or more from any established brand. Today, Rigol sells the DHO1204 (a four-channel, 200MHz instrument with 25 Mpts memory), for approximately SGD 1,200. That is a genuine and remarkable disruption. The question is not whether Rigol has closed the price gap (it clearly has), but whether it has closed the performance gap in the dimensions that matter for your specific application.

What has not changed is fundamental measurement physics. An oscilloscope measures voltage as a function of time. Its ability to do that accurately is determined by its noise floor (how much intrinsic noise the instrument adds to the measurement), its bandwidth (the frequency at which its amplitude response degrades by 3 dB), its waveform update rate (how many complete waveform captures it processes per second), and the quality of its vertical amplifier and analog-to-digital conversion chain. These are hardware properties, and hardware has a cost floor that market disruption cannot simply eliminate.

Spec sheets are also frequently misleading when read in isolation. A "1GHz bandwidth" specification means something different on an instrument where that bandwidth is achieved through genuine wideband analog frontend design versus one where it is software-unlocked from a nominally lower-bandwidth configuration. Understanding what the specification actually reflects in the hardware is the first step to an honest comparison.

Keysight InfiniiVision Series. The Engineering Benchmark

Keysight's InfiniiVision X-Series spans from the 1000X (entry lab grade) through to the 6000X (high-performance R&D and compliance), all with genuine hardware bandwidth, not software unlocked. The DSOX and MSOX product lines cover bandwidths from 50MHz on the 1000X up to 6GHz on the 6000X series.

MegaZoom IV ASIC

The defining technical advantage of the Keysight InfiniiVision series is the MegaZoom IV dedicated waveform processing ASIC. This is a custom silicon chip designed specifically for oscilloscope waveform acquisition and display processing. On the DSOX3054T, it enables a waveform update rate of over 1,000,000 waveforms per second. A figure that is not achievable with a general-purpose processor and display pipeline. In practice, this means that the DSOX3054T can capture infrequent, randomly occurring glitches that a scope with a 30,000 waveform/second update rate would statistically miss in the intervals between captures. For engineers debugging intermittent faults (power supply glitches, protocol violations, EMC-related transients), this is not a minor specification. It is the difference between seeing the problem and not seeing it.

Noise floor and vertical performance

Keysight's noise floor on the InfiniiVision series is typically around 0.6–0.8 mV RMS at the 1 mV/div vertical range. The most sensitive range where noise performance matters most. This reflects the quality of the input amplifier design and the ADC. Lower noise means you can measure smaller signals without the instrument's own noise masking the signal content, and it means amplitude measurements are more accurate at low signal levels.

Memory depth

Standard memory depth on InfiniiVision X-Series ranges from 1 Mpts (1000X) to 4 Mpts (3000T) per channel, with the 4000X offering up to 1 Gpts in the optional deep memory configuration. Memory depth determines how long a capture can be at a given sample rate. Deeper memory means you can capture a longer time interval without reducing sample rate, which matters when you are looking for rare events in a long protocol stream.

Active probes and probe ecosystem

Keysight's probe ecosystem is a genuine differentiator. The N2140A active probe (400MHz, 10:1, 1 MΩ input impedance, <0.4 pF tip capacitance) dramatically reduces probe loading compared to a passive 10:1 probe, which typically presents 10–15 pF of capacitance at the measurement point. At frequencies above 200–300 MHz, this loading effect causes measurable signal attenuation and waveform distortion with a passive probe. For high-frequency signal integrity work, power rail measurement, or any application where you cannot afford to disturb the circuit under test, active probes are not optional. Keysight's InfiniiMax active probe system extends this capability to the GHz range for the higher-bandwidth InfiniiVision models.

CalReady self-calibration

Keysight InfiniiVision instruments include a CalReady self-calibration function. This stores factory calibration data in the instrument and compensates for temperature coefficient drift within a working session. It should be run after the instrument has warmed up for approximately 30 minutes, and whenever the ambient temperature changes significantly. This is a meaningful feature: it means the instrument's readings are more consistent across a working day as the ambient temperature varies.

It is important to be clear about what CalReady does not do. It does not replace annual external calibration. It addresses short-term, temperature-related drift. It does not compensate for the long-term aging of the timebase crystal oscillator, component parameter drift over time, or the comprehensive parameter-by-parameter verification that external calibration provides. Both CalReady (within session) and annual accredited calibration (annual) are necessary and serve different purposes.

Software ecosystem and price

Keysight's BenchVue software platform and MATLAB integration provide a capable automation and data capture environment. Protocol decode options (covering I2C, SPI, UART, USB, CAN, LIN, and many others), are licensed separately on the InfiniiVision platform. Price range: approximately SGD 4,000 for a 1000X series entry model to SGD 30,000 and above for a 6000X series high-performance model.

Rigol DHO/DS Series. The Value Contender

Rigol's current product line for bench use centres on the DHO series (DHO1072, DHO1102, DHO1204) as the updated architecture successor to the older DS series. The DHO series features improved noise performance compared to the DS series, a refreshed user interface, and significantly deeper memory than competing products at similar price points.

Competitive specifications at compelling prices

The DHO1204 offers 200MHz bandwidth across four channels with 25 Mpts of memory at approximately SGD 1,200. This is a genuinely capable instrument for its price. Memory depth is actually a notable advantage of the DHO series , 25 Mpts is deeper than many more expensive instruments in Keysight's entry range. For capturing long protocol streams or slowly varying signals at high sample rates, this is practically useful.

Rigol instruments support standard SCPI control and Python automation. The interface is well-documented and widely used. There is a large community of third-party resources, open-source scripts, and tutorials, which matters for engineers building automated test setups on a budget. The DS1054Z in particular has accumulated an extraordinary body of community knowledge and tooling that makes it productively usable well beyond its nominal specifications.

Noise floor and vertical accuracy

Rigol's noise floor on the DHO series is typically in the range of 1.5–2 mV RMS at the minimum vertical scale. This is measurably higher than Keysight's noise floor, and translates directly to higher measurement uncertainty on small signals. Rigol's vertical accuracy specification is ±3% typical, compared to Keysight's ±1.5%. For most bench development work (debugging digital logic, power supply characterisation, audio circuits, sub-100MHz communication interfaces), this difference is irrelevant because the signals being measured are far larger than the instrument's noise floor and the measurement tolerances are far larger than ±3%.

The software bandwidth unlock issue

Some Rigol models (including certain DS1054Z and DS1104Z variants), have hardware that is electrically capable of operating at a higher bandwidth than the firmware allows. Third-party software tools exist that can modify the firmware to unlock this capability. This practice requires a careful understanding of what it does and does not achieve.

The hardware in these instruments is the same silicon across SKUs; the firmware simply limits the operating bandwidth. Unlocking removes this limit and gives access to the hardware's true capability. However, the factory calibration is performed at the nominal specification, not at the unlocked bandwidth. If you use the instrument at the unlocked bandwidth in a measurement context (particularly in any compliance, quality, or reference capacity), you are operating at a bandwidth for which you have no traceable calibration certificate. The measurement uncertainty at the unlocked bandwidth is uncharacterised. This is an ethically ambiguous practice that also likely voids the manufacturer warranty. It is not appropriate for any use case where calibration traceability matters.

Price range for Rigol: approximately SGD 480 for a DS1054Z to SGD 3,000 for a four-channel DHO1204.

Head-to-Head Comparison

When Keysight Is the Right Choice

Keysight is the correct choice when your use case has one or more of these characteristics:

• IEC/MIL-STD compliance testing: when the oscilloscope is cited as a reference instrument in a test report, for example, measuring rise/fall times for EMC pre-compliance, or verifying signal timing against IEC 61000 or MIL-STD-461 requirements. In these contexts, the test report is submitted to a customer, a notified body, or a regulatory authority. The instrument must have a current traceable calibration certificate, and the measurement uncertainty must be formally assessed against the test limits. Keysight's ±1.5% accuracy class provides more usable margin against the test limit than Rigol's ±3%.
• Semiconductor characterisation and high-speed digital design above 500MHz: once you are working with DDR memory interfaces, high-speed serial buses (PCIe Gen 3+, USB 3.x, 25G Ethernet), or RF signals, the bandwidth, noise floor, and probe loading effects become the limiting factors in measurement quality. This is Keysight's design environment.
• Production test where false passes would be costly: the waveform update rate advantage of Keysight's MegaZoom ASIC is critical in production test. A glitch that occurs once every few seconds will be captured routinely at 1,000,000 waveforms/second; it may be missed almost entirely at 30,000 waveforms/second. A false pass on a production unit due to a missed glitch has downstream warranty and liability implications far exceeding the instrument price differential.
• Defence and aerospace requirements: DSTA and DSO National Laboratories environments, aerospace MRO facilities, and defence electronics contractors in Singapore typically have contractual requirements for instrument traceability to national standards. This means the instrument must have a calibration certificate from a SINGLAS-accredited laboratory with explicit traceability to NMC. Both Keysight and Rigol can be calibrated to this standard, but Keysight's accuracy class means the uncertainty contribution from the oscilloscope is smaller.
• When you are measuring what you need to measure. Not hoping the measurement is close enough: this is perhaps the most honest framing. Keysight instruments are specified conservatively; what the datasheet says is what the instrument reliably delivers. When the measurement margin is tight, that conservatism has value.

When Rigol Is the Right Choice

Rigol is the honest, right choice in a larger number of everyday engineering scenarios than the brand positioning might suggest:

• Educational laboratories: bandwidth and accuracy are more than sufficient for teaching oscilloscope concepts, characterising standard circuits, and developing laboratory skills. The price point means an institution can equip more workstations for the same budget. The large community and extensive tutorial ecosystem makes Rigol instruments particularly well-suited to self-directed learning.
• SME bench development: if you are debugging digital interfaces below 200MHz, characterising power supply switching waveforms, analysing audio circuits, or troubleshooting embedded system timing, Rigol's DHO series gives you everything you need. The ±3% vertical accuracy is not a constraint when the signal you are measuring has ±20% tolerance or the pass/fail threshold is comfortably far from the nominal value.
• Second instrument or backup scope: a Rigol DHO1204 as a second bench oscilloscope (for probing multiple points simultaneously, or as a backup when the primary scope is in calibration), costs less than a calibration service visit for a Keysight. This is a genuinely rational purchase in any serious electronics lab.
• Situations where the instrument's measurement uncertainty is not the limiting factor: if you are measuring a tolerance component (±10%, ±20%), debugging a prototype where the specification is still being defined, or performing qualitative waveform observation, the difference between ±1.5% and ±3% accuracy is invisible in the output. Do not pay for performance you cannot use.
• Budget-constrained startups: cash deployed in a Rigol instead of a Keysight at company formation is cash available for components, prototyping, testing services, and team. The tradeoff is appropriate early in a company's lifecycle, when instrument accuracy is not the binding constraint on product development.

Calibration. What Both Brands Require

Regardless of brand, oscilloscopes used as measurement instruments in any quality-managed environment require annual calibration by an accredited laboratory. The calibration procedure verifies the same set of parameters for both Keysight and Rigol instruments:

• Timebase accuracy: the accuracy of the time axis, determined by the stability and calibration of the internal crystal oscillator reference. Crystal oscillators drift with temperature and age. Timebase accuracy must be verified at calibration, not assumed to be stable from factory. Timebase error directly propagates to frequency measurement, period measurement, and rise/fall time measurement errors.
• Vertical scale accuracy: the accuracy of the amplitude measurement at each V/div setting. This is verified across the full range of vertical settings because the accuracy can vary by range. An instrument may be within ±1.5% at 500mV/div but be closer to ±3% at 1mV/div. A thorough calibration checks each range that will be used in practice.
• Bandwidth verification: the -3 dB frequency at which the instrument's amplitude response falls to 70.7% of the input amplitude. This requires a calibrated sine wave source at the relevant frequency. An instrument rated at 200MHz may, after drift, be responding correctly only to 175MHz. Enough to cause meaningful rise-time measurement errors on fast digital signals and to shift frequency response curves in EMC testing.
• Trigger level accuracy: the accuracy of the trigger level control, which determines where the instrument references its trigger edge. Trigger level errors can cause systematic timing offset in time-correlated measurements.

For oscilloscopes used in compliance testing, the calibration certificate should explicitly state the bandwidth-verified figure and the vertical accuracy at each range calibrated. A certificate that records only "pass/fail" without stating the measured value and uncertainty is not acceptable for most compliance and quality applications.

Bandwidth calibration requires a precision signal source with calibrated amplitude and frequency. Typically a sine wave source that has itself been calibrated against a traceable reference at the relevant frequency point. Unitest's scope calibration scope covers timebase accuracy, vertical scale accuracy at all standard ranges, and bandwidth verification. The three parameters that matter for compliance use.

The Software Unlock Question

The Rigol bandwidth unlock deserves a more detailed treatment than a brief dismissal, because engineers frequently encounter it and the decision has real measurement implications.

Some Rigol DS series instruments (most famously the DS1054Z), use hardware that is physically identical to higher-bandwidth variants. The difference between a DS1054Z (50MHz) and a DS1104Z (100MHz) in the same production run can be purely a firmware lock. Third-party tools exist to modify the firmware and unlock the hardware's full capability. In informal use and hobby applications, this is a widely-used practice with a large body of community documentation.

The critical issue from a measurement quality standpoint is this: the factory calibration of a DS1054Z is performed at 50MHz. The calibration certificate covers the instrument as a 50MHz oscilloscope. If you unlock the instrument to operate at 100MHz and use it in a measurement role (including any quality, compliance, or reference application), you are operating at a bandwidth for which no calibration exists. The vertical accuracy, noise floor, and frequency response at 100MHz are uncharacterised. Your measurement uncertainty at the unlocked bandwidth is formally unknown.

For informal bench use with wide margins and no compliance implications, this may be a reasonable tradeoff. For any application where the measurement result matters (production test, compliance testing, anything that goes into a quality record), it is not appropriate. The correct path, if you need the higher bandwidth, is either to operate within the nominal specification or to send the instrument for calibration at the unlocked bandwidth specification. Unitest can accommodate calibration at non-standard specifications when requested. This gives you a traceable certificate covering the actual bandwidth at which you are operating.