Siglent 2104XP Lockup on Bode Plot

[Ray CI]

Hi Folks...

I can't seem to get Bode plots to work on a new Siglent 2104XP.  I have tried all seemingly logical configuration settings.  Upon starting the plot, the unit lights flash (while doing the auto-set) then, the scope locks-up indefinitely and must be manually power-cycled.

To simplify the explanation, an example (trivial) LP filter circuit is shown (470Ohm resistor and 1uF cap) connected to the waveform generator set to 1V pp.  Channel 1 (probe on the left) is at the input.  Channel 3 (probe on the right) is reading the output.  Manually energizing the ckt at different frequencies with the AWG indicates the ckt works and the probes are connected properly.

Enclosed are pics of the circuit and all relevant Bode config parameters.  The scan is running from 100 to 5kHz.   This is my first time using the Bode Analysis feature since getting this scope.  I don't see what I'm doing wrong and need some additional eyes on this.

Any thoughts on what might be going wrong here?

PS:  Yes, I have read the manual thoroughly (it's basically useless) and I've watched numerous YouTube videos that demonstrate this feature.  Still stuck!

EDIT:  When setting this up, the "Internal Interface Test" shows successful status.

Thanks
Ray

[tautech]

For newbies to Bode plot the simplest mistake is not placing a Tee into the DUT stimulus path to provide a reference signal into the scope.
FG Out should be tee'ed into ch 1 and then to the DUT and the through path into the selected channel provides the info for the scope to plot the result as it sweeps through the set frequency range.

Simply, without providing a reference signal you won't get a result.
Initially I would leave vertical scaling to Auto to be sure to get a result on the display.

This app note for X-E Bode plot might provide some further guidance.
https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/
However, do note they are using a standalone AWG and have synced outputs which negates the need to tee the stimulus.

[Martin72]

For newbies to Bode plot the simplest mistake is not placing a Tee into the DUT stimulus path to provide a reference signal into the scope.

See picture one...
Plus:

Upon starting the plot, the unit lights flash (while doing the auto-set) then, the scope locks-up indefinitely and must be manually power-cycled.

Me I´ve forgot one or two times to clamp on the source signal, but this never lead the scope to hang.
@Ray:
Do a default setup, then try again...

[Ray CI]

For newbies to Bode plot the simplest mistake is not placing a Tee into the DUT stimulus path to provide a reference signal into the scope.
FG Out should be tee'ed into ch 1 and then to the DUT and the through path into the selected channel provides the info for the scope to plot the result as it sweeps through the set frequency range.

Simply, without providing a reference signal you won't get a result.
Initially I would leave vertical scaling to Auto to be sure to get a result on the display.

This app note for X-E Bode plot might provide some further guidance.
https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/
However, do note they are using a standalone AWG and have synced outputs which negates the need to tee the stimulus.

Thanks tautech...  At this stage, I'll try anything.  Trust me on this, I've tried multiple circuits and literally dozens (if not a hundred) different config settings.  They all just hang the scope.  Once it's locked-up, it could sit there for an hour and not recover.  All the buttons are frozen and it needs a hard power reset to recover.   Is that consistent with not using a Tee?  I would think that an improper wiring setup would lead to a weird/inaccurate plot result.



Ray

[tautech]

For newbies to Bode plot the simplest mistake is not placing a Tee into the DUT stimulus path to provide a reference signal into the scope.
FG Out should be tee'ed into ch 1 and then to the DUT and the through path into the selected channel provides the info for the scope to plot the result as it sweeps through the set frequency range.

Simply, without providing a reference signal you won't get a result.
Initially I would leave vertical scaling to Auto to be sure to get a result on the display.

This app note for X-E Bode plot might provide some further guidance.
https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/
However, do note they are using a standalone AWG and have synced outputs which negates the need to tee the stimulus.

Thanks tautech...  At this stage, I'll try anything.  Trust me on this, I've tried multiple circuits and literally dozens (if not a hundred) different config settings.  They all just hang the scope.  Once it's locked-up, it could sit there for an hour and not recover.  All the buttons are frozen and it needs a hard power reset to recover.   Is that consistent with not using a Tee?  I would think that an improper wiring setup would lead to a weird/inaccurate plot result.

It can't make any sensible plot without a reference waveform, period !
Using the internal FG as you said you where, it must be tee'ed. The scope inputs need the stimulus values/info as the SW only drives the FG....internal connections to FG don't mean shit as far as getting stimulus info back to compute a plot.

It's less about settings and instead hookup as this must be correct.
So dedicate ch1 to the stimulus reference that also goes to the DUT and the result into any of the other channels which need be set to receive the result.
If we consider there must always be a channel needed for the reference (stimulus) waveform there are 3 more in these 4ch DSO's where we can also use 2 to measure points/stages within a DUT and one for the full through result.

[Ray CI]

For newbies to Bode plot the simplest mistake is not placing a Tee into the DUT stimulus path to provide a reference signal into the scope.
FG Out should be tee'ed into ch 1 and then to the DUT and the through path into the selected channel provides the info for the scope to plot the result as it sweeps through the set frequency range.

Simply, without providing a reference signal you won't get a result.
Initially I would leave vertical scaling to Auto to be sure to get a result on the display.

This app note for X-E Bode plot might provide some further guidance.
https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/
However, do note they are using a standalone AWG and have synced outputs which negates the need to tee the stimulus.

Thanks tautech...  At this stage, I'll try anything.  Trust me on this, I've tried multiple circuits and literally dozens (if not a hundred) different config settings.  They all just hang the scope.  Once it's locked-up, it could sit there for an hour and not recover.  All the buttons are frozen and it needs a hard power reset to recover.   Is that consistent with not using a Tee?  I would think that an improper wiring setup would lead to a weird/inaccurate plot result.

Ray

OK, Martin...  you get to wear a golden star all day long!      Pressing the default button did the trick.

Here's the filter circuit output and as a bonus, the plot of the amplifier with input and output filters that I originally intended to measure.    (And yes, I wanted an Fc of  around 15 and 15k Hz). 

[Ray CI]

For newbies to Bode plot the simplest mistake is not placing a Tee into the DUT stimulus path to provide a reference signal into the scope.
FG Out should be tee'ed into ch 1 and then to the DUT and the through path into the selected channel provides the info for the scope to plot the result as it sweeps through the set frequency range.

Simply, without providing a reference signal you won't get a result.
Initially I would leave vertical scaling to Auto to be sure to get a result on the display.

This app note for X-E Bode plot might provide some further guidance.
https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/
However, do note they are using a standalone AWG and have synced outputs which negates the need to tee the stimulus.

Thanks tautech...  At this stage, I'll try anything.  Trust me on this, I've tried multiple circuits and literally dozens (if not a hundred) different config settings.  They all just hang the scope.  Once it's locked-up, it could sit there for an hour and not recover.  All the buttons are frozen and it needs a hard power reset to recover.   Is that consistent with not using a Tee?  I would think that an improper wiring setup would lead to a weird/inaccurate plot result.

It can't make any sensible plot without a reference waveform, period !
Using the internal FG as you said you where, it must be tee'ed. The scope inputs need the stimulus values/info as the SW only drives the FG....internal connections to FG don't mean shit as far as getting stimulus info back to compute a plot.

It's less about settings and instead hookup as this must be correct.
So dedicate ch1 to the stimulus reference that also goes to the DUT and the result into any of the other channels which need be set to receive the result.
If we consider there must always be a channel needed for the reference (stimulus) waveform there are 3 more in these 4ch DSO's where we can also use 2 to measure points/stages within a DUT and one for the full through result.

Hi Tautech...  I know that with RF, cable terminations and impedance matching is critical.   In my case, I'm showing my son how to make a class A audio amp (last one I made was in 1979 so, I'm pretty happy about this outcome).  I've got a bunch of BNC connectors and a partial spool of RG316 -but darn! I cannot find the crimper. 

Anyhow, thanks to both you and Martin for helping me out.  Man, I spent the better part of 8 hours over 2 days, trying to figure-out what I was doing wrong.


Thanks!

Ray

[rf-loop]

For newbies to Bode plot the simplest mistake is not placing a Tee into the DUT stimulus path to provide a reference signal into the scope.
FG Out should be tee'ed into ch 1 and then to the DUT and the through path into the selected channel provides the info for the scope to plot the result as it sweeps through the set frequency range.

Simply, without providing a reference signal you won't get a result.
Initially I would leave vertical scaling to Auto to be sure to get a result on the display.

This app note for X-E Bode plot might provide some further guidance.
https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/
However, do note they are using a standalone AWG and have synced outputs which negates the need to tee the stimulus.

Thanks tautech...  At this stage, I'll try anything.  Trust me on this, I've tried multiple circuits and literally dozens (if not a hundred) different config settings.  They all just hang the scope.  Once it's locked-up, it could sit there for an hour and not recover.  All the buttons are frozen and it needs a hard power reset to recover.   Is that consistent with not using a Tee?  I would think that an improper wiring setup would lead to a weird/inaccurate plot result.

It can't make any sensible plot without a reference waveform, period !

We can work also without reference when we know what we are doing and if we think this kind of FRA result meet our need.

Here is scalar plot (no Phase)  without any kind of reference.

In this setup DUT out come to CH1.   CH4 is set for reference DUT In but there is nothong connected to CH4.
Generator OUT is connected to DUT input.
Naturally in this case we can not get Phase (it generate Phase data but it is bullshit and Phase trace is better to shut off.
Also we loose measured information about DUT input level.

For this kind of application we need select display amplitude mode for Vout  instead of  Vout/Vin (because now Vin is unknown!)

In this case I have also turned off Automatic Gain Control (= Hold). (reduced dynamic range but faster sweep.)
Because AGC is off user need set oscilloscope input V/div carefully beforehand when BodePlot is turned off because when BodePlot is on, user can not adjust these (Why channel adjustments are not enabled - perhaps because <censored>). Channels V/div need adjust beforehand so that highest point do not clip but enough close to full scale for get best possible dynamic range over noise level. For this adjustment, make all connection ready and use then SDG manually for find DUT out top level and set V/div so that it can not clip. (keep it so that p-p do not go out of display even if there is some room before ADC really clip. (In some models just before clip it may have also reduced linearity)

I have used maximum points (501) (because span is narrow I use linear freq axis).
One sweep in this case 1min 35sec. (it may also be different with different SDG and connection. I have used here LAN between Scope and SDG1000X)



For this plot there is no reference.  CH4 is in this case DUT In and there is nothing connected. I have set BodePlot channels just only for color because here I want yellow plot. (setup can see in other image). Note this noise. Channel Gain mode is Hold! (AGC Off)  Turning it to Auto give more dynamic range.

[KungFuJosh]

I don't use any T-adapters. In my pickup test setup, I have a test coil that has the AWG (mini grabbers) and CH3 (10X probe) connected directly. The DUT output is connected to CH1. You can see phase is measured, but I have the trace disabled.