Here's a couple Bode Plots of a simple RC Low Pass and High Pass filter ...
However, this took awhile to do, much longer than I would tolerate in a lab where my time was valued,...
Yes it takes time.
There is freq scale log and start from 10Hz to 120MHz.
I do not remember how long it takes and also I can not see what amount of data points you have used, if max 500 it takes time.
As we can see it use lot of time in inder 1kHz specially long time between 10Hz to 100Hz.
I have many times thinking, is it always necessary to use so much time there in low freq, how important this resolution is in these cases around 10Hz. What are freq steps there.
If user can define resolution steps it can perhaps do lot of faster. Yes user can set amount of steps but not how it is shared over sweep span.
Reason, explanation, for long time can see in these two images. Thinking this it can ask how much can ask more fast BodePlot.
One image can see current step is bit over 1kHz.
What more it tell: 10 bits, 1Mpts, 10ms/div, 10Msa/s.
Sampling itself one shot for this one individual BP data takes 100ms. time
Then other image tell bit more
Perhaps this is sweep starting point.
50ms/div, 1Mpts, 2Msa/s.
Sampling itself one shot for this one individual BP data takes 500ms. time
And this sampling alone is not whole time what one step take. There need wait generator after command, then fill pretrig buffer and wait trig. Look level and do front end level ajustment (if ALC is on) if need and start processing for result and after all is ready then command Generator to do next step.
But, is it optimal for time. Perhaps I do not need lot of datapoints between some very low frequencies but still want more resolution on other part of whole sweep.
1MHz datapoint is muck more fast to do than 10Hz data point.
Selecting between log and lin sweep is also question. Some times lin is technoically ok but user select log because.... eyes like. Some toimes I have thinked that perhaps in some cases is more wise to do linear sweep and then afterwards change it to log scale.
It need ask, do I need same amount of BP data points between 10-100Hz as I need between 10MHz to 100MHz. If user have freedom to set how steps are shared over sweep he can optimize sweep time perhaps better for just his need in one test.
Here in image I can see is used Ch1 for reference and CH2 for DUT out. These BP what I have tested, user have full freedom to select reference channels and select also what receive signal from DUT.
In some cases it is useful to select other than manufacturer set default.
Example user can select Ch1 for reference and then Ch3 for DUT out. In this case both channels use different ADC so both can use max interleaved speed. But also there can be other reasons for select... example one like more yellow or green BP then he can select Ch2 as reference channel. And if DUT have one or two or even three outputs then select input channels as like set colors for these.
Always in these cases need care that reference channel timing is perfect related to DUT out channels so that phase have even some meaning. Phase accuracy is... not easy case at all when we go out from "audio" freq range. There need calibrate cables and channels time difference and if work with generator 2 channel mode, one for ref and one for DUT in, also need care generator phase error.
One channel mode is other way difficult when we go out from "audio" LF freq range. Signal split to Ref and DUT need do right way and then also care phase calibration using right kind of dummy thru in place of DUT.
But, we have programmable level figure for sweep span in BP (during sweep it follow user defined level table), why we can not have programmable step interval where user can define how many steps it use in user defined slices in whole sweep. Yes it is not easy... but with this user can do some optimize for sweep time just for his need in some test case. He can put more fine steps in important positions and then less to some other positions.
Did you really need 50 data points between 10 and 100Hz. (if 500 was in use)
But perhaps 50 is not even enough between 10 to 100MHz in some cases.
It is also good to remember, what ever it plot to screen real base to all is the Table what BP produce. Display is just only post image from The Table and example if user select manually scale so that part of bode is lot of out from display still all is there in The Table what can then later display moving display position in table or changing displayed scale.
BP is whole "machine" other than oscilloscope and before is fully familiar with it, it need work, it need lot of work for find all things built in.
It was one evening when I do some work with this BP II (other scope model)
Then hit some produced some harmonics etc in DUT.
I have measured DUT out level with HP Power meter quite accurately and in separate freq points.
Then I look Siglent BodePlot and think what a crap shit they have done... lot of errors related my measurements with instruments what are enough reliable way checked they are in specs.
Thank god I have SA. With SA I look this DUT out in same setup used with BP.
SA show nicely this test signal from DUT output. But what... also its level is totally off. After some seconds SA was tuned to wide span and it was clear that reason or other my DUT produce totally high amount of harmonics and non harmonics. Game over, this was clear. Of course know that power meter used is wide band.
Then like child I think all my old "bode plots" in my life and with various methods using analog scopes etc...
Why this Siglent bode plot did not see whole level but it see nearly like SA see this test carrier level out from DUT. But is is oscilloscope BodePlot....
Well, inside some minutes I know this BodePlot is frequency selective. It listen only frequency what he have ask generator send. Yes of course it is not very narrow resolution band width. Later investigations also show that RBW is not fixed. It depends frequency. It have several width steps.
I have been tens of years fan for reading HP Journals (until "Lady Agilent" destroy HP) and I have been in situation where I have get these for read also when I was young.
If this BP is made by old HP in history there is perhaps very deep explanation how it works and what is inside it, how it produce final data from signals, including least bloc level working principle and math how it compute result etc etc and something about how it set default parameters and so on.
From Siglent. Just (nearly) Nothing.
Information level nearly like image where see button and in image can see it is labeled "Start" and then perfect User Reference Manual tell.. Push this for start.
Example for some lab (hobby or what ever) use it is important to know in some cases how this BP do. Example some knowledge about this frequency selectivity, even least tell it is.
Because DUT can be lot of more than just passive RLC circuit and if dut generate something during sweep there is not any indication about it. It is good that user know that this BP is swept tuned freq selective.
Now it is in level "Use if you like it and try find how it works"
Long time ago R&S SWOB have whole manual about theory and practice with all details and not even this level how HP explain they instruments. Yes they was expensive but it need remember volumes...
today instruments are low price but it can not explain why manuals and information is so poor.. because now also volumes are much higher so price of good and deep manuals etc is funded from more small sources but bit it looks more like "who cares" state than price is reason. Of course one reason is this continuous new products and hurry to push unfinished products to markets. No one can then keep manuals updated specially if manuals are very deep even with theory paragraphs deeply integrated with used FW and circuits.
Home
Help
Search
Login
Register
