Siglent ODP6100B 1 GHz Optical Isolated Probe
Siglent ODP6100B is a 1 GHz optical isolated voltage probe with ±60 kV isolation and ultra-high CMRR. Ideal for accurate high-voltage floating and gate signal measurements in Si, SiC, and GaN power designs. Features battery operation, auto-zeroing, and multiple attenuators for flexible test setups.
A gigahertz of bandwidth on a node that is floating at kilovolts
The hard measurement in a half-bridge is the high-side gate. The signal you want is a few volts of gate drive. The node it sits on is slewing through hundreds of volts in a nanosecond. An ordinary probe sees both, and the common-mode transient swamps the few volts you came for.
The ODP6100B breaks the electrical connection entirely. The front end sits on the circuit, runs from its own battery, converts the signal to light and sends it down two metres of optical fibre to a receiver plugged into the oscilloscope. There is no copper between the device under test and the instrument, so there is no path for the common-mode current that causes the problem.
It is the 1 GHz member of the ODP6000B series, with a 0.45 ns typical rise time and ±60 kV of isolation. The 500 MHz ODP6050B is otherwise identical and costs €4,030 less — so the section below sets out, in arithmetic rather than adjectives, exactly when the extra bandwidth is worth paying for.
Three parts, and no copper between them
Power over fibre and an optical signal path give complete galvanic isolation between the measurement system and the device under test.
Front end (E/O transmitter)
Takes the attenuator on an SMA input, runs from a replaceable lithium battery, and converts the differential voltage to an optical signal. 102 × 45 × 33 mm, and it sits on the circuit at whatever potential the circuit is at.
Optical fibre
Two metres of fibre, and the only thing joining the two halves. This is where the isolation comes from — light carries the signal, so no current can follow it. Keep the bend radius above 10 cm.
Rear end (O/E receiver)
Plugs directly into an oscilloscope BNC and takes USB 5 V / 2 A. Carries the auto-zero button and the status indicator. Presents a 50 Ω terminated output at ±0.5 V full scale.
Is the extra €4,030 worth it? Here is the arithmetic
The ODP6100B costs €8,530 against €4,500 for the ODP6050B — 90 % more money for twice the bandwidth and a 0.45 ns rise time instead of 0.7 ns. Everything else about the two probes is identical. Whether that is worth paying depends on one thing only: how fast your edges are.
A probe does not measure a rise time, it measures the combination of its own and the signal’s. To a good first approximation those add in quadrature — the measured rise time is √(tsignal² + tprobe²). Put the two published rise times through it and the decision makes itself.
| Your signal’s real rise time |
ODP6050B — 0.7 ns | ODP6100B — 0.45 ns | ||
|---|---|---|---|---|
| shows | error | shows | error | |
| 5 ns | 5.05 ns | +1 % | 5.02 ns | +0.4 % |
| 2 ns | 2.12 ns | +6 % | 2.05 ns | +2.5 % |
| 1 ns | 1.22 ns | +22 % | 1.10 ns | +10 % |
| 0.5 ns | 0.86 ns | +72 % | 0.67 ns | +35 % |
Edges slower than ~2 ns
The ODP6050B is within about 66 % and the#37; and the extra €4,030 buys you very little. Buy the 500 MHz one and spend the difference on attenuators.
Edges around 1 ns
This is where the upgrade earns itself. The 500 MHz probe reads 22 % slow; the ODP6100B is within 10 %. Typical of SiC gate drive.
Edges near 0.5 ns
⚠ Neither probe is accurate here — the ODP6100B is still 35 % out. Buy it for the isolation, not expecting sub-nanosecond fidelity.
Root-sum-square is the standard first-order approximation for cascaded responses. The figures above are arithmetic from the two published rise times, not a measurement of either probe — your oscilloscope’s own rise time adds into the same sum and makes the real error slightly larger again.
Why a gigahertz, and why now
Wide-bandgap devices moved the goalposts. Silicon and IGBT designs switch in around a hundred nanoseconds, and a 100 MHz differential probe resolves that comfortably. Silicon carbide is an order of magnitude faster, gallium nitride another, and the rail voltages went up at the same time rather than down.
Faster edges need bandwidth, and higher rails need common-mode rejection at that bandwidth — which is the part that gets quietly dropped. A probe’s CMRR headline is almost always its DC figure, and DC is not where a GaN half-bridge does its damage.
The ODP6100B exists for the fast end of that table. It is also where the honest limits below start to matter most, because at a gigahertz the attenuator you fit decides almost everything.
How fast, and how high
| Device | Rise time | Differential voltage |
|---|---|---|
| Silicon / IGBT | ~100 ns | up to ~700 V |
| SiC | ~10 ns | up to ~3.3 kV |
| GaN | 1 ns and below | up to ~10 kV |
Indicative trend across device technologies, from Siglent’s ODP6000B technical material — not a specification for any particular part.
Read this before you order: the 1 GHz is a small-signal figure
The ODP6100B is a 1 GHz probe and it will measure up to ±5000 Vpk. It will not do both at once, and on this model the gap is at its widest. The whole table below sits below the probe’s bandwidth: its highest row is 200 MHz, and that is with the 50X tip at ±25 Vpk. At full voltage you have 700 kHz.
| Maximum test voltage | Maximum test frequency | Attenuator |
|---|---|---|
| ±5000 Vpk | 700 kHz | 10000X |
| ±2500 Vpk | 800 kHz | 5000X |
| ±1000 Vpk | 2 MHz | 2000X |
| ±500 Vpk | 3 MHz | 1000X |
| ±250 Vpk | 20 MHz | 500X |
| ±100 Vpk | 50 MHz | 200X |
| ±50 Vpk | 100 MHz | 100X |
| ±25 Vpk | 200 MHz | 50X |
Source: ODP6000B series instruction manual, EN01D. To use the bandwidth you paid for, the signal has to be small — which is exactly what gate-drive measurement is, and exactly what a kilovolt rail is not. If your work is high voltage and high frequency, design your test around this table rather than around the headline bandwidth.
CMRR at 1 GHz is where this probe is decided
Siglent publish 160 dB (DC–10 MHz), and that is a real figure — roughly a hundred million to one. But it belongs to a low-ratio attenuator at low frequency. Every attenuator starts near 180 dB at DC and falls as frequency rises, and the high-voltage attenuators fall fastest. On a 1 GHz probe that last column is the one that matters, and it is brutal.
| Attenuator | DC | 1 MHz | 10 MHz | 100 MHz | 1 GHz |
|---|---|---|---|---|---|
| 5:1 | ~180 | ~168 | ~150 | ~130 | ~90 |
| 50:1 SUPPLIED | ~180 | ~150 | ~125 | ~100 | ~90 |
| 200:1 | ~180 | ~135 | ~115 | ~85 | ~80 |
| 1000:1 SUPPLIED | ~180 | ~100 | ~85 | ~55 | ~30 |
| 2000:1 SUPPLIED | ~180 | ~90 | ~80 | ~50 | ~20 |
| 5000:1 SUPPLIED | ~180 | ~85 | ~75 | ~45 | ~15 |
| 10000:1 | ~180 | ~80 | ~70 | ~40 | ~10 |
Values in dB, read from the CMRR curves in the ODP6000B instruction manual and rounded to the nearest 5 dB — Siglent publish them as a graph rather than a table, so treat them as indicative, not guaranteed. Rows marked SUPPLIED are the four attenuators that ship with the probe.
Read the last column carefully. At 1 GHz the 50X tip still gives around 90 dB, but the 5000X gives around 15 dB — a common-mode rejection of less than six to one. Put 600 V of common mode on that and roughly 100 V of it lands in your result. The gigahertz and the kilovolts are available from the same probe, but not from the same measurement.
Our application note optically isolated probes and high-CMRR measurement works this through in volts rather than decibels, and gives a five-step procedure for choosing the attenuator so the numbers above land on your side of the trade.
What common-mode rejection is worth, in volts
CMRR is quoted in decibels, which makes it easy to nod at and hard to feel. Convert it to volts and it stops being abstract. Take the measurement this probe exists for — the high-side gate of a half-bridge running from a 600 V rail, where the gate signal itself is about 15 V.
A classical high-voltage differential probe has roughly 25 dB of common-mode rejection at 100 MHz. That is a ratio of about 18:1. So 600 V of common mode arrives at the screen as 600 ÷ 18 ≈ 33 V of error — on a signal that is only 15 V tall. The error is more than twice the thing you are measuring, and nothing about the trace tells you which part is which.
| Common-mode voltage | 600 V |
| Differential signal | 15 V |
| Classic HV differential probe, 100 MHz | 25 dB ≈ 18:1 |
| Error it contributes | ≈ 33 V |
| ODP6100B, 50X tip, 100 MHz | ≈ 100 dB |
The 25 dB figure is from Wolfspeed’s application note Dynamic Characterization and Measurement Methods, quoted by Siglent. The ODP6100B row is read from the manual’s CMRR curves.
The attenuator is the range — and, at 1 GHz, the performance
The probe amplifier itself takes ±0.5 V. Every measurement range below is simply that figure multiplied by the attenuation ratio, which is why the specification sheet’s “±25 V” is the 50X number rather than a limit on the probe. Four are supplied — 50X, 1000X, 2000X and 5000X. The other seven are ordered separately.
| Model | Ratio | Measurement range | Max non-destructive | Input impedance | Connector | Price ex VAT |
|---|---|---|---|---|---|---|
| CK-AT5X-2-SMA | 5:1 | ±2.5 Vpk | 2 kVpp | 1 MΩ ‖ 28 pF | MMCX | €173 |
| CK-AT10X-2-SMA | 10:1 | ±5 Vpk | 2 kVpp | 1 MΩ ‖ 6 pF | MMCX | €173 |
| CK-AT20X-2-SMA | 20:1 | ±10 Vpk | 2 kVpp | 5 MΩ ‖ 6 pF | MMCX | €173 |
| CK-AT50X-2-SMA | 50:1 | ±25 Vpk | 2 kVpp | 10 MΩ ‖ 4 pF | MMCX | supplied |
| CK-AT100X-2-SMA | 100:1 | ±50 Vpk | 3 kVpp | 10 MΩ ‖ 2 pF | MMCX | €173 |
| CK-AT200X-2-SMA | 200:1 | ±100 Vpk | 3 kVpp | 10 MΩ ‖ 2 pF | MMCX | €173 |
| CK-AT500X-2-SMA | 500:1 | ±250 Vpk | 5 kVpp | 10 MΩ ‖ 2 pF | HVMCX | €288 |
| CK-AT1000X-2-SMA | 1000:1 | ±500 Vpk | 5 kVpp | 20 MΩ ‖ 2 pF | HVMCX | supplied |
| CK-AT2000X-2-SMA | 2000:1 | ±1000 Vpk | 6 kVpp | 20 MΩ ‖ 2 pF | HVMCX | supplied |
| CK-AT5000X-2-SMA | 5000:1 | ±2500 Vpk | 6 kVpp | 40 MΩ ‖ 2 pF | HVMCX | supplied |
| CK-AT10000X-2-SMA | 10000:1 | ±5000 Vpk | 12 kVpp | 40 MΩ ‖ 2 pF | 5.08 mm socket | €368 |
Prices from Siglent’s Annex 1 price list 2026_1, ex VAT. The optional attenuators do not yet have their own pages on this shop — ask us for a quotation and we will order them with the probe. Note 2 of the manual: where an MMCX connector is specified, SSMB may be used instead. There is a step-by-step method for choosing between them in our application note on high-CMRR measurement.
Specifications
| Bandwidth (−3 dB) | 1 GHz |
| Rise time (typical) | 0.45 ns |
| Terminal load | 50 Ω |
| Output voltage range | ±0.5 V |
| Host noise (Vrms, typical) | 1.5 mV |
| DC accuracy | ≤ ±1 % |
| Isolation voltage (DC + peak AC) | ±60 kV |
| Attenuator (50X) + host delay | 15.3 ns, with the 2 m optical fibre |
| Power — front end | Battery, approx. 8 hours operating, approx. 30 days standby |
| Power — rear end | USB 5 V / 2 A |
| Auto calibration | Yes, one button on the receiver |
| Operating temperature | −10 °C to 60 °C |
| E/O transmitter (front end) | approx. 102 × 45 × 33 mm |
| O/E receiver (rear end) | approx. 106 × 49 × 23 mm |
| Attenuator length | approx. 200 mm |
| Optical fibre length | approx. 2 m |
| Weight | approx. 400 g |
ODP6100B or ODP6050B?
Bandwidth and rise time are the only differences. Isolation voltage, the attenuator set, CMRR behaviour, the power arrangement and the contents of the case are identical on both.
| ODP6050B | ODP6100B | |
|---|---|---|
| Bandwidth | 500 MHz | 1 GHz |
| Rise time (typ.) | 0.7 ns | 0.45 ns |
| Isolation | ±60 kV | ±60 kV |
| Price ex VAT | €4,500 | €8,530 |
When a differential probe is the better buy
Optical isolation is expensive, and it is not always what the measurement needs. Below roughly 400 MHz, with moderate common-mode voltage, a high-voltage differential probe measures the same signal for a fraction of the cost. Siglent’s own DPB6150 pair reaches ±1500 Vpk with more than 80 dB of CMRR at DC.
Where the optical probe earns its price is common-mode rejection at high frequency — the fast switching edges of SiC and GaN devices, where a differential probe’s CMRR has already fallen away.
Manual & documentation
Both links go to Siglent’s own document server and always fetch the current revision, so they cannot go stale. Nothing opens in a new window.
Application note — optically isolated probes and high-CMRR measurement
siglent.eu · 16 min read
Written by us, and the practical companion to this page: what common-mode rejection is worth in volts, the CMRR curves tabulated by attenuator, the voltage-versus-frequency limit drawn as a boundary, a five-step attenuator selection procedure with a worked example, and the setup mistakes that quietly corrupt the result.
ODP6000B Series Instruction Manual
PDF · approx. 1.0 MB · EN01D
The complete specification: electrical and mechanical tables, all eleven attenuators with their measurement ranges and non-destructive limits, the CMRR curves, the maximum-voltage-versus-frequency table, the packing list and the fibre handling rules. Every figure on this page is read from it. It covers both models in the series.
Siglent’s application note — optical isolated probes and high CMRR
PDF · approx. 1.0 MB · factory original
The manufacturer’s own note, and the source of the differential-probe comparison. Worth reading for its side-by-side captures of the same gate edge measured with an optical probe and with a conventional differential probe.
There is no separate datasheet
Siglent publish one document for the ODP6000B series, and it is the instruction manual. There is no standalone datasheet — we checked the factory’s full document index and its own product page. That is not a problem in practice, because the manual carries everything a datasheet would: the electrical and mechanical specification tables, the complete attenuator list, the CMRR curves and the voltage-versus-frequency limits. If you were looking for a datasheet, the manual is it.
Firmware and documents for the oscilloscope you will use this probe with are on our downloads page, which is searchable by model number. If you need something we have not listed, just ask.
In the box
The full packing list from the instruction manual — identical to the ODP6050B’s. Two batteries and a twin charger mean one pack charges while the other works, which is the point of a battery-powered front end.
| Voltage probe unit | 1 | E/O transmitter, 2 m fibre and O/E receiver |
| CK-AT50X-2-SMA | 1 | 50X attenuator — ±25 Vpk |
| CK-AT1000X-2-SMA | 1 | 1000X attenuator — ±500 Vpk |
| CK-AT2000X-2-SMA | 1 | 2000X attenuator — ±1000 Vpk |
| CK-AT5000X-2-SMA | 1 | 5000X attenuator — ±2500 Vpk |
| CK-327 | 2 | MMCX to Dupont cable |
| CK-28 | 5 | MMCX-K |
| CK-24 | 9 | HVMCX-K |
| CK-326 | 3 | HVMCX extension cable |
| CK-328 | 2 | MMCX to RG178 |
| CK-325 | 1 | Output extension cable |
| CK-314A | 1 | USB Type-C power cable, 1.5 m |
| CK-605 | 2 | Power supply adaptor, 5 V / 2 A |
| CK-690B | 1 | E/O transmitter support bracket |
| CK-691 | 1 | Battery charger set |
| Documents | — | Instruction manual, warranty card, test report |
Using it properly
Six things from the instruction manual that decide whether the measurement is right — and whether the probe survives.
Set the scope to 50 Ω
Whenever the measured signal’s rise time is under 3.5 ns — which on a 1 GHz probe is essentially always — set the oscilloscope input to 50 Ω and set the channel’s attenuation ratio to match the attenuator fitted. Get either wrong and the amplitude on screen is wrong.
Zero it in place
One press of the auto-zero button calibrates and zeroes the probe in about 20 seconds, without disconnecting it from the circuit. Green flashing means it is working; three beeps and steady green mean it succeeded.
Two batteries, user-replaceable
Standard card-type camera lithium cells, about 8 hours per charge and 30 days of standby. When they age you buy replacements yourself — no return to the factory. ⚠ Remove the attenuator first: fitted, it mechanically blocks the battery door.
It sleeps by itself
The transmitter wakes when the receiver is powered up and sleeps five seconds after the receiver is switched off, so battery is not wasted between measurements.
Respect the fibre
Minimum bend radius 10 cm. No knots, no twisting, nothing heavy resting on it, and inspect it before each use. The fibre is the isolation — if it fails, the probe does.
Power down before unplugging
The front end connects directly to high voltage on the device under test. Always switch the circuit off before removing the probe, and keep the transmitter box suspended and away from high-voltage pulse circuitry while measuring.
What it is used for
High-side gate drive in half- and full-bridge inverters
The measurement the probe exists for. VGS on the high-side device is a small differential signal riding on a node that swings the full DC link in nanoseconds.
SiC and GaN device characterisation
Double-pulse testing of wide-bandgap devices, where the switching edge carries the information and a 100 MHz differential probe rounds it away. This is where the extra bandwidth over the ODP6050B is actually spent.
Floating measurements under fast common-mode transients
Anywhere the reference point moves — motor drives, frequency converters, renewable-energy inverters.
Switching supplies, LED drivers and inverter appliances
R&D, debugging and maintenance across power electronics generally, including small differential signals sitting on large common-mode voltages.
Questions
Is the ODP6100B worth €4,030 more than the ODP6050B?
It depends entirely on your edge rate, and the arithmetic is on this page. At 2 ns edges the 500 MHz ODP6050B is within about 6 % and the extra money buys you very little. At 1 ns it reads 22 % slow while the ODP6100B is within 10 %, and that is the point where the upgrade earns itself. Below about 0.5 ns neither probe is accurate — the ODP6100B is still 35 % out — so do not buy it expecting sub-nanosecond fidelity.
Can I use the full 1 GHz at high voltage?
No, and this is the most important thing to understand before ordering. The manual’s maximum-test-voltage table tops out at 200 MHz, and that is with the 50X tip at ±25 Vpk. Every high-voltage attenuator is limited to a few megahertz or less — ±2500 Vpk is capped at 800 kHz. The 1 GHz bandwidth belongs to small-signal work.
What is the CMRR at 1 GHz?
It depends on the attenuator, and the spread is enormous. Reading the manual’s curves at 1 GHz: around 90 dB with the 5X, 50X or 20X tips, but around 30 dB with the 1000X and around 15 dB with the 5000X. So the two things this probe is bought for — a gigahertz of bandwidth and rejection on a high-voltage node — pull hard against each other.
What is the maximum voltage it can measure?
±5000 Vpk with the optional CK-AT10000X-2-SMA attenuator, but only up to 700 kHz. The four attenuators supplied cover ±25 V to ±2500 Vpk. The measurement range is always the attenuation ratio multiplied by ±0.5 V, because that is the probe amplifier’s own input range.
What does ±60 kV isolation actually mean?
It is the common-mode voltage the optical link will stand off — the potential the probe tip may float at relative to the oscilloscope’s ground. It is not the signal you measure. The differential signal is set by the attenuator; the isolation figure is what makes it safe to take that measurement on a floating node.
How is it powered, and how long does it run?
The front end runs from a replaceable lithium battery for about 8 hours, with roughly 30 days of standby. Two batteries and a twin charger are supplied, so one charges while the other works. The receiver at the oscilloscope end takes USB 5 V / 2 A.
Do I need to set anything on the oscilloscope?
Set the channel’s attenuation ratio to match the attenuator fitted, and set the input impedance to 50 Ω whenever the signal’s rise time is under 3.5 ns — which, on a probe bought for gigahertz work, is essentially always.
Where is the datasheet?
There isn’t one. Siglent publish a single document for the ODP6000B series — the instruction manual — and it carries everything a datasheet would: the full electrical and mechanical specifications, all eleven attenuators, the CMRR curves and the voltage-versus-frequency limits. It is linked in the documentation section above.
Should I buy this or a high-voltage differential probe?
If your work is below about 400 MHz with moderate common-mode voltage, a DPB6150A or DPB6150D does the job for a fraction of the price. Optical isolation earns its cost when you need genuine common-mode rejection at high frequency — fast SiC and GaN gate-drive measurement is the case it was built for.
Not sure you need the gigahertz?
Tell us your edge rate, the voltage you are measuring and the frequency you need to see it at. We will tell you whether the ODP6100B earns its €4,030 premium over the ODP6050B for your measurement, which attenuator that combination actually allows — or say plainly if a differential probe would serve you better for a fraction of the cost. We are the official Siglent distributor for the European Union and we ship from the Netherlands.