CS328A Issues

Recently bought a isolated signal generator for loop analysis. Installed the signal generator on my CS328A. I spent quite a bit of time troubleshooting since some things did not make sense.
I will list them:
1. Gain/phase was really noisy (picture shown in attachment)
2. Signal coming out of the signal generator after 10Mhz could not be measured or distorted by CS328A. Looked at the same signal using a Rigol scope and it was distorted (see attachment).
3. Signal generated from Rigol and measured by CS328A was very distorted at 25Mhz. The same signal generated from Rigol was measured with a Rigol scope showing a clean Signal.

Rigol Scope used was a 100Mhz 1GSa/s which also acts as a signal generator (Model: 1104Z). All measurement was made with a coaxial cable and no termination. I have also tried terminating the cable (50 ohm) with no signal integrity at high frequency. Same issue as before. I avoided the probes that came with the scope since they were limited to 10Mhz.

Not sure if I am missing something or if I have a CS328A not working properly. Need Help!

Recently bought a isolated signal generator for loop analysis. Installed the signal generator on my CS328A. I spent quite a bit of time troubleshooting since some things did not make sense. I will list them: 1. Gain/phase was really noisy (picture shown in attachment) 2. Signal coming out of the signal generator after 10Mhz could not be measured or distorted by CS328A. Looked at the same signal using a Rigol scope and it was distorted (see attachment). 3. Signal generated from Rigol and measured by CS328A was very distorted at 25Mhz. The same signal generated from Rigol was measured with a Rigol scope showing a clean Signal. Rigol Scope used was a 100Mhz 1GSa/s which also acts as a signal generator (Model: 1104Z). All measurement was made with a coaxial cable and no termination. I have also tried terminating the cable (50 ohm) with no signal integrity at high frequency. Same issue as before. I avoided the probes that came with the scope since they were limited to 10Mhz. Not sure if I am missing something or if I have a CS328A not working properly. Need Help!

Hello Rashed,
I took a look at your signals, and I see no problems. Your Gain response was flat from 0 to about 30 MHz, and then had a slight lift because you are using unterminated coax. You have set the FRA to 20 steps per decade, and so the sig gen will step with 20 frequency points in each decade. As you have the option 'Connect Measure Points' turned off, the in between frequencies are shown as -120 dB. To see a continuous response set the option to enable (click on the led).

The scope samples at 100 MSPS. It has very low jitter however (<1ps), and so the sampling points are accurate. Based on Fourier analysis they can still be used to measure the amplitude of each frequency component. We use frequency folding (ie undersampling) above the Nyquist limit of 50 Mhz. If you are sampling a 25 MHz signal, there will be only 4 samples per cycle, and so in the time domain the signal will not look like a sine wave. This is of not concern to the FRA system, as we use frequency coherent sampling. If you wish to see the time domain signal, use the Repetitive Display Method.

The CS701 signal generator does not have a reconstruction filter in it's output, as it introduces amplitude and phase variation as you approach the filter cutoff frequency. For this reason it looks severely distorted on the Rigol. However this is of no concern to the FRA system, as we are coherent. If you wish to use the sig gen in the time domain, please attach the Minicircuits BNC filter supplied with the CS701, and the signal will then be a nice sine wave.

So to summarize:
1. Use the Connect Measurement Points led on.
2. Use 50 coax and 50 ohm terminations for best performance at >10 Mhz.
3. If you use scope probes, set them to 10x and be careful about compensating them first.

regards

Bart Schroder

Hello Rashed, I took a look at your signals, and I see no problems. Your Gain response was flat from 0 to about 30 MHz, and then had a slight lift because you are using unterminated coax. You have set the FRA to 20 steps per decade, and so the sig gen will step with 20 frequency points in each decade. As you have the option 'Connect Measure Points' turned off, the in between frequencies are shown as -120 dB. To see a continuous response set the option to enable (click on the led). The scope samples at 100 MSPS. It has very low jitter however (<1ps), and so the sampling points are accurate. Based on Fourier analysis they can still be used to measure the amplitude of each frequency component. We use frequency folding (ie undersampling) above the Nyquist limit of 50 Mhz. If you are sampling a 25 MHz signal, there will be only 4 samples per cycle, and so in the time domain the signal will not look like a sine wave. This is of not concern to the FRA system, as we use frequency coherent sampling. If you wish to see the time domain signal, use the Repetitive Display Method. The CS701 signal generator does not have a reconstruction filter in it's output, as it introduces amplitude and phase variation as you approach the filter cutoff frequency. For this reason it looks severely distorted on the Rigol. However this is of no concern to the FRA system, as we are coherent. If you wish to use the sig gen in the time domain, please attach the Minicircuits BNC filter supplied with the CS701, and the signal will then be a nice sine wave. So to summarize: 1. Use the Connect Measurement Points led on. 2. Use 50 coax and 50 ohm terminations for best performance at >10 Mhz. 3. If you use scope probes, set them to 10x and be careful about compensating them first. regards Bart Schroder

Thank you for the reply. It helped a lot.

I was following the video on the website and I am guessing the software in the video is a little outdated.

I have a few more questions regarding FRA.
1. After probe calibration, phase is always off by tens of degrees at greater than 30Mhz. Only way to get phase within a degree was to use the filter (mini-circuits) that came with the unit. I did not see the use of the filter in the video for probe calibration. Does this mean I need to use the filter to get within a degree of phase correction? Is there some knobs in the software that allows me to do probe calibration without using the filter. My setup used 50 ohm termination for each channel and 50 ohm coax cable (RG58). see attachment.
2. In the video, I do not see the software filter being used fro each channel. Every time I use FRA, the filter turns ON automatically in the software for both channel. Is that the default moving forward?

Regards

Thank you for the reply. It helped a lot. I was following the video on the website and I am guessing the software in the video is a little outdated. I have a few more questions regarding FRA. 1. After probe calibration, phase is always off by tens of degrees at greater than 30Mhz. Only way to get phase within a degree was to use the filter (mini-circuits) that came with the unit. I did not see the use of the filter in the video for probe calibration. Does this mean I need to use the filter to get within a degree of phase correction? Is there some knobs in the software that allows me to do probe calibration without using the filter. My setup used 50 ohm termination for each channel and 50 ohm coax cable (RG58). see attachment. 2. In the video, I do not see the software filter being used fro each channel. Every time I use FRA, the filter turns ON automatically in the software for both channel. Is that the default moving forward? Regards