Application Brief · AWG Migration

Still fighting the double pulse on a Tektronix AWG7000?

If you run a Tektronix AWG7082C, you know the workaround: a series resistor or a capacitor tacked onto the output to tame a secondary pulse and the ringing that rides with it. The Berkeley Nucleonics Model 685 puts a clean edge on the bench without the band-aid, and it produces a true double pulse only when you ask for one. The AWG7000 series is also discontinued, so that workaround is now your long-term support plan. It does not have to be.

Berkeley Nucleonics Model 685 ArbRider AWG
Berkeley Nucleonics Model 685 arbitrary waveform generator
The problem

The artifact you have learned to live with

You program one clean pulse. The output hands you two. There is the pulse you wanted, then a smaller ghost of it a few nanoseconds later, wrapped in ringing. The fix that gets passed around the lab is always the same: a series resistor at the output, or a capacitor in line, to damp the reflection and swallow the second pulse. It mostly works. It also reshapes your edge, costs you amplitude, and drops one more uncalibrated part between the instrument and the device under test.

That is not how a bench should feel in 2026. And on the AWG7082C it is now permanent, because the AWG7000 series has been discontinued. New units are gone. What is left is the used market, the rental fleet, and a soldered-on workaround standing in for a support contract.

Legacy AWG7000 output Model 685 output Wanted pulse, then a ghost pulse and ringing One clean edge, flat baseline, no conditioning
A single programmed pulse: the artifact teams damp with external passives, next to a clean DC-coupled edge.
Root cause

Why the second pulse shows up

The AWG7000 series reaches its top bandwidth through a low-amplitude, AC-coupled direct-DAC path. Two things follow from that design. Fast edges into anything short of a perfect 50 ohm termination reflect, and a reflection arriving back at a fast output reads on your scope as a secondary pulse. An AC-coupled path also carries its own baseline behavior, so droop and overshoot ride along with the edge. Teams running these units in the field land on the same answer over and over: add a series resistor or a blocking capacitor to absorb the reflection and quiet the ringing.

The passive band-aid treats the symptom. It cannot give you back a clean, full-amplitude, DC-coupled edge, because the output stage was never built to deliver one.

The fix

The Model 685 puts a clean edge on the bench

The Berkeley Nucleonics Model 685 drives a DC-coupled, 50 ohm, 5 Vpp output directly. Rise and fall times come in under 300 ps, jitter under 2 ps, and the baseline sits flat where you set it through a ±2.5 V offset range. No series resistor. No blocking capacitor. No ghost pulse to chase. The edge you program is the edge that reaches your device.

And the double pulse? The 685 makes one on purpose. A clean double pulse is a native waveform, alongside sine, square, ramp, pulse, Gaussian, noise, and fully arbitrary playback, so the shape that used to be a defect becomes a setting you dial in. Underneath it all sits 16-bit vertical resolution, which is where the 685 pulls decisively ahead: more levels, lower quantization noise, and pulse and step shapes the 8 to 10-bit AWG7000 path simply cannot render.

Side by side

AWG7082C and Model 685

SpecificationTektronix AWG7082CBNC Model 685
Product statusDiscontinued; used and rental onlyCurrent and supported
Vertical resolution8 to 10-bit16-bit
Output stageLow-amplitude AC-coupled direct DAC, or amplified path; field workarounds for ringingDC-coupled, 5 Vpp into 50 Ω, ±2.5 V offset, no conditioning
Rise / fall timeFast on the direct path, AC-coupled< 300 ps, DC-coupled
JitterNot specified here< 2 ps
Sample rate8 GS/s (16 GS/s interleaved)6.16 GS/s
Analog bandwidthTo ~3.2 GHz (to ~6 GHz with the wide-bandwidth option)2 GHz
Channels22, 4, or 8
Memory depth32 MptsUp to 4 Gpts/channel
Double pulseUnintended artifact to suppressNative, clean, on demand

Model 685 figures are from the published Berkeley Nucleonics specification. AWG7082C figures are from published Tektronix documentation for the discontinued AWG7000 series. The AWG7082C leads on raw sample rate and direct-path bandwidth. For the pulse, timing, and signal-emulation work most AWG7000 owners actually run, 16-bit fidelity and a clean DC-coupled output matter more, and they end the double-pulse workaround for good. If your application genuinely needs more than 6 GS/s on a single channel, the wider ArbRider line has options worth a conversation.

What you gain

More than a clean pulse

Next step

See it on your own pulse

Bring us the waveform that gives your AWG7000 trouble, the one you have been damping with passives. We will reproduce it on a Model 685 and put both edges on a scope side by side. Let's look at the results together. Ready to retire the workaround?

Request a Model 685 demo   or explore the Model 685 data sheet.