Jump to content
  • 0

Network Analyzer for Loop Gain


AcHmed

Question

I want to use the Network Analyzer for extracting the power stage and loop transfer function of switching and linear regulators. In order to do this it would be helpful if the sweep amplitude can be varied at selected frequency intervals. As an example I might want a 400mV amplitude from 10Hz to 400Hz then 50mV from 401Hz to 100kHz. Does the discovery already have this feature and I can't find it or not? If the discovery presently can't do this would it possible to add this feature?

 

I'm also wondering what's the difference between Waveforms 2015 and v2.7.5?

Link to comment
Share on other sites

10 answers to this question

Recommended Posts

Hello,

You can find the list of main differences on the following page: https://reference.digilentinc.com/waveforms3#documentation
To have both installed at the same time, v2 and v3 (2015), don't forget to update the earlier one to v2.9.4.

In WF 2015 you can perform the measurement separately. Make phase one and add as reference channel, make phase two, then adjust the frequency start/stop to see both parts.

To increase the precision of such low frequency measurement, increase the number of "Min periods" under Wavegen gear options.
In case you are working with resonant circuit (like speaker) increase the Settle time too.

Link to comment
Share on other sites

4 hours ago, attila said:

Hello,

You can find the list of main differences on the following page: https://reference.digilentinc.com/waveforms3#documentation
To have both installed at the same time, v2 and v3 (2015), don't forget to update the earlier one to v2.9.4.

In WF 2015 you can perform the measurement separately. Make phase one and add as reference channel, make phase two, then adjust the frequency start/stop to see both parts.

To increase the precision of such low frequency measurement, increase the number of "Min periods" under Wavegen gear options.
In case you are working with resonant circuit (like speaker) increase the Settle time too.

I was hoping that the feature was already available in the software. The problem is the two signals can be seperated by upto 40dB to 60dB which would put one signal well below the noise floor of even a high-end Agilent Network Analyzer. For example one signal may be 100mV the other 1mV or less. The wide signal amplitude difference occurs at the extremes, DC to maybe a couple hundred Hz then a decade or so after the crossover frequency. The best way to deal with this is user selectable amplitude at specific frequency ranges.

Before I spend time wadeing through the literature on the API would it even be possible to add this feature?

 

Thank you for your suggestion.

 

Link to comment
Share on other sites

4 hours ago, attila said:

Another possibility is selecting in Network Analyzer under Wavegen gear options Channel External.
Having this, you can use external generator or control the frequency sweep from WaveForms Wavegen instrument.

Thank you attila I will try that.

My original plan was to put a RM10 T38 wide band transformer, with some fast opamps OPA356 type for gain and BNC outputs in a small hammond enclosure. Just plug the AD in the back and good to go. I've neen using the poor mans loop injection method and thought I might try and get everything in one enclosure. I guess I could use one of the AD9838B DDS chips I have for the signal plus some buttons and LCD with uc. Thats getting a bit more involved then I wanted but it is, what it is.

Great job on the Analog discovery nice little kit for the pricepoint with good software.

Link to comment
Share on other sites

On 2016-10-05 at 11:32 PM, emarte said:

Achmed,did you find any work around to this? I have been trying the same with no luck yet.

No I didn't.

The best way I can think to do it would be the ability to set the amplitude over a sweep range.

For example.*

1Volt amplitude from 100Hz to 1kHz

Then from 1.01kHz to 10kHz say 500mV then from 10.01kHz to 100kHz 250mV etc..... You get the point.

That is the way dedicated loop gain instruments do it.

I was considering a hardware approach but it started to need more time then I had to work on it. The simplest approach is to add the feature in software the Hardware already exist in the discovery. I would really like this feature, unfourtanetly I am not familiar with how to add it in waveforms.

 

*Those were only examples to get the concept across, the actual amplitude settings would depend on the DUT.

Link to comment
Share on other sites

I Understand. Definenily that feature would be a great plus to the Waveform.

For now I have been doing some testing with a mixture of tips I have been reading from where you in most of them actually.

What I I think it may work is using The ADD Channel option (limited to 8). 

1-) Back to your example, lets say that we sweep 1Volt amplitude from 100Hz to 1kHz and then Save in Add Channel

2-) Then from lets say from 100khz to what... 250k a 250mV fixed.

3-) And then a Third sweep from 1Khz to 100KHZ using External and waveform to generate a sweep (in sweep option) from 1k to 100khz with an amplitude of 1V damping option to 250mv.

1 and 2 can and could be done easily. Getting to generate in waveform the middle interval is being my problem so far. It is no a perfect solution but I think once we can generate the sweep we want we could get satisfactory responses. Pretty  much like image below.

Fig-5.jpg

I already have a isolation transformer I got from ebay and also I will be building in the next week a kit for measuring output impedance in power supplies. I am trying to replicate Ridley Ap300 accessories.

 

a_intro.jpg

Well, I'll keep you posted of any updates.

Regards,

Edwin

 

Link to comment
Share on other sites

Hello Again,

I was able to generate a changing signal like the image above. I used 3 sweep  intervals and external and It does the job. It is not ideal but it works. I tested it with a simple low pass RC filter now I have to test it with a real power supply.

 

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...