Review: Picotest J2150B comb injector

This handy tool lets you test active and passive components at frequencies up to 3 GHz. I tried an early unit, using it for radiated and conducted EMI emissions testing.

After almost ten years, Picotest has launched a new version of its Harmonic Comb Injector (generator). The J2150B ($495) is one of the pieces I’ve wanted for my lab, and so I purchased one.

Comb generators produce harmonic artifacts as a feature of their nonlinear behavior. They consist of a diode and amplifier to create a broadband spectrum of higher-order harmonics that engineers can use to characterize devices. Because of their nonlinear performance, they are generally limited to full characterization of active devices. You can, however, characterize passive devices as well. This review is based on my use and experience. I am not linked in any way to Picotest.

Figure 1. The J2150B arrived in this box.

The J2150B arrived in a small package (Figure 1). Its dimensions are 6.6 cm × 2.5 cm × 1.4 cm, connectors included (Figure 2). The enclosure was 3D printed and not sealed, which means I could open it to see the inside. The circuit board wiggled a bit inside the case, and I think that’s because the case does not seem to be the final, refined version. The previous model seemed better from the pictures, but knowing the quality of Picotest, I’m confident that this is a pre-release unit.

Editor’s note: According to Picotest, the company purposely did not glue the case together, which provided access to the circuit inside. This unit is a prototype.

Figure 2. The J2150B is small enough to fit in the palm of your hand or in your pocket.

On my unit, the push button is separated from the rest of the enclosure, and it doesn’t have a membrane or something similar. As a result, it sometimes doesn’t work quite well. That’s understandable given that this is an early unit.

Characteristics

Table 1 shows the J2150B’s electrical and RF characteristics. The data sheet shows that the harmonics can cover up to 1.5 GHz of frequency range; we will check that later.

Table 1. J2150B specifications (image: Picotest)

Other characteristics are the small form factor and the USB-powered capability (even using a battery pack). The output is DC-coupled, so a DC-blocker is needed to avoid damaging the front-end of a spectrum analyzer (DC is not a friend of RF receivers). Figure 3 shows the comb generator plus DC-blocker.

Figure 3. A DC blocker protects the spectrum analyzer’s input.

Table 2. J2150B modes of operation. (image: picotest)

Table 2 explains the comb generator’s five working modes. Mode 1 is cycling between modes 2, 3, and 4. Mode 5 is a simple 10 kHz square wave, 50% Duty Cycle, good for oscilloscope probe calibration. Modes 2, 3, and 4 are generating pulses at 1 kHz, 100 kHz, and 10 MHz spacing. The latter is the difference between this and the previous model, which had a frequency of 8 MHz in Mode 4. At least, this is what is stated in the instruction manual and technical specifications: the result of a measurement with a 600 MHz oscilloscope showed that the frequency is still 8 MHz. As it seems a pre-series product, they may have forgotten to update the instruction manual or the firmware on the comb generator itself. I then tried to measure it with my 100 MHz Rigol DS1054, and I got the same result, shown once in Figure 4.

Figure 4. Acquisition of pulses in Mode 4, with 600 MHz

As soon as the comb is plugged into a USB socket, Mode 1 is the default (cycle between the three frequencies of Modes 2, 3, and 4). Switching between the other modes is achieved by pressing the MODE button for at least 1 sec. Three LEDs will change color according to the table in Table 2.

Stability and rise time

A comb generator could be used as a reference source of “known pulses,” and for this reason, it can have many uses. As a known source, it should be stable in both frequency and amplitude. An oscilloscope can check the pulse shape and other parameters such as risetime and frequency (or period, the spacing between two pulses in time).

Figure 5. The yellow trace shows the initial state of the peak immediately after switching on. The purple trace shows the Max-hold trace, while the cyan trace shows the Min-hold trace.

To check stability, I used my spectrum analyzer and the three traces shown in Figure 5. The yellow trace is the trace measured as soon as the comb generator is powered up from a USB cable with an integrated switch. This trace is immediately frozen to catch the initial state. Purple trace was set as max-hold, and cyan trace was set as min-hold. By having a single pulse on the screen, I was able to monitor changes in amplitude using markers: max-hold marker minus min-hold marker gives the total excursion over time. The comb generator was kept on and working for about 45 minutes, and the results are shown in Figure 5. Figure 6 shows the setup used.

Figure 6. I used this setup with a spectrum analyzer to evaluate the stability over time of the comb generator, in Mode 4.

Now let’s investigate the rise time. Picotest declares something between 470 ps and 270 ps: I measured less than 900 ps with a 600 MHz, 10 GS/s oscilloscope. It does make sense because if a signal has a 470 ps rise time, it means that it has frequency components up to 2 GHz (inverse of the rise time). Figure 7 shows the acquisition of the pulse in Mode 4 (8 MHz).

Figure 7. A 600 MHz oscilloscope at 10 Gsa/sec captured a single pulse in Mode 4.

Applications

Radiated emissions in SAC or FAR: check of the table influence

*Figure 8. The top-view table setup for conducting radiated emissions tests inside an anechoic chamber. (Image: CISPR 16-1-1)*

EMC test engineers involved in full-compliance testing must consider the influence of the table used when measuring radiated emissions. In fact, the material of the table should match the precise characteristics in terms of relative dielectric constant, ε_r. CISPR 16-1-4 explains the method. Figure 8 shows the top-view setup for table evaluation according to CISPR 16-1-1, while Figure 9 shows the setup side view as prepared inside an anechoic chamber. The comb generator is connected to a suitable antenna, which should be small in size to respect the distances explained in the standard. The difference between two measurements, with and without the table, provides the influence of the table, frequency by frequency, expressed in dB.

Figure 9. The side-view table setup for conducting radiated emissions tests inside an anechoic chamber. (Image: CISPR 16-1-1)

Having a small comb generator makes this check easier, and by adding a battery pack and a short RF cable, the setup is done. I would use Mode 4, as the frequency span is big, usually from 30 MHz to 1500 MHz. But we can go even beyond, I would say up to 3 GHz. Figure 10 shows the setup for wide-range frequency emissions of the comb generator, and Figure 11 shows the frequency range of 30 MHz to 3000 MHz in Mode 4. I used my Tektronix RSA306B to see a very spectacular real-time emission of the comb generator in Mode 4: at 3 GHz, the level is around -60 dBm, that is 47 dBµV, with a bandwidth of 100 kHz.

Figure 10. This setup shows the comb generator connected to a Tektronix RSA306B real-time USB spectrum analyzer.

Figure 11. The spectrum plot shows the output of the comb generator from 30 MHz to 300 MHz in Mode 4.

Conducted emissions test: check of the measurement chain

A similar application uses the comb generator to check the setup for conducted emissions. In this case, using the comb generator makes it easy and fast because the comb generator is a well-known source of RF peaks.

The measurement is quite straightforward. I needed a cable to connect an electric plug to an RF connector (from BNC up). The cable has an electric plug on one side and a connector suitable for the comb generator on the other. EUT mains from the line-impedance stabilization network (LISN) should be disconnected for safety reasons. Doing so avoids a direct connection between the mains voltage and the RF output of our comb generator, which would permanently ruin it and possibly cause electrocution.

By comparing a previous measurement against the actual, we are able to spot failures in the measurement chain. I found that, especially for laboratories where inverters and high current devices are tested, it is very important to check the LISN, because high current pulses can damage either the LISN itself or the Limiter, but without any exterior sign of burning. Figure 12 shows the measurement chain verification setup for conducted emissions. I don’t have an AC LISN, so I used a DC LISN, but the principle is the same. From the left: battery pack, comb generator, DC block, DC LISN by Tekbox. The coaxial cable is connected to the Rigol DSA 815 spectrum analyzer through a transient limiter (with 10 dB attenuator included) by Leo Bodnar.

Figure 12. This test setup let me verify the LISN for conducted emissions.

Cable shielding and resonances

The final application for a comb generator is to find resonances and check cable shielding. Many articles have been written about that; under References, I collected the most interesting, in my opinion.

I was trying a way to couple the pulses of the comb generator not only on coaxial cables, but also on any kind of cable, by conducting the pulses between the inner wires of a shielded cable and the shield itself. In a video by Dr. Min Zhang (see references), you can see a box doing that. In Figure 13, I used several breakout boards, modifying them by adding an SMA connector. Doing that lets me connect the output of the comb generator to different types of cables. Figure 14 shows the J2150B connected to USB, serial, and HDMI cables for testing. All these cables are shielded.

Figure 13. Examples of breakout boards modified to add an SMA connector for easy coupling with the comb generator’s output.

Figure 14. The modified breakout boards let me connect the comb generator to different cables.

Summary

The J2150B comb generator is a small but powerful tool that makes it handy for carrying in your pocket. I didn’t cover an entire range of applications, regarding low-frequency signal injection in DC-DC converters, mainly because I don’t have the necessary board to experiment with. You can find interesting articles by visiting the Picotest Insights page.

Main advantages of the comb are:

The power supply can be a simple power bank.
It uses a traditional USB type A connector.
It’s a very small form factor
You can change frequencies according to the application (from 1 kHz to 8 or 10 MHz in three steps).

The remaining doubts I have, related to:

The 3D printed case is not final and needs refining, especially the push-button for the choice of the Modes.
The manual states that the J2150B can operate up to 10 MHz, but actually it is 8 MHz.

I asked Picotest about these two points, and they replied that the comb generator is still in its early stage, and they are willing to invest more time and effort to finalize the product if they receive signs of interest.

References

Kenneth Wyatt Review: Picotest J2150A harmonic comb generator
Kenneth Wyatt, Measuring Cable Resonance with a Comb Generator
Min Zhang, How Good are Shielded Cables? – A Practical Demonstration, YouTube video

Claudio Stazzone holds a degree in Telecommunications Engineering. He began his EMC experience in 2008, performing measurements on both civil and automotive devices and components. After one year, he became interested in EMC troubleshooting and problem-solving. He has performed hundreds of “in situ” measurements on any sort of industrial machine. He has experience in chambers characterization measurements (NSA, SVSWR, SE) and laboratory management. He is interested in pre-compliance measurements, sharing technical issues with other engineers all over the world, writing technical articles, spice simulations, and EM waves open-source simulators. He currently covers the position of EMC Technical Coordinator for an important TIC company. In his everyday activity, there are EMC measurements with customers, and general management of the EMC activity of the lab from a technical point of view. He has been Responsible for the EMC and RED directives for Notified Body activity.