I started using the Bode plotter tool extensively. Unfortunately it is restricted to 1 MHz maximum frequency, even though the PCGU could go up to 2 MHz, and the scope can go even higher; how luck will have it the range a bit above 1 MHz just happens to be the one that is interesting for my project…
(1) could you update the Bode plotter software so it can go up to 2 MHz? That should be a small change.
(2) You could define a sine5x waveform that has 5 sine wave periods in it, and use it for the Bode scan above 2 MHz: When you scan the generator frequency up to 2 MHz using this waveform, the actual output signal will be a sine wave of up to 10 MHz. Of course the generator’s oversampling is less and its waveform fidelity (or jitter) may increase as a result, but this is still better than the current restriction to 1 MHz. The scope should just be able to handle that speed in real time mode. Still looks like a pretty limited software change to me (add another range limit (10 MHz), use a different, predefined waveform whenever scanning above 2 MHz), and it would add a lot of value.
I just tried with a 4x sine wave and went up to 8 MHz, looks pretty good. Maybe a calibration function is helpful (i.e. acquire a reference scan where the generator is connected to the scope with just a short cable) to account for limited bandwidth of the generator output and the scope input. The user would acquire that waveform and then tell the Bode tool to plot all waveforms relative to this one (i.e. divide the other waveforms with this one). But that could come later, my first priority would be to just be able to acquire data up to higher frequencies.
Any chance getting a beta version of such an upgraded Bode tool? I’m sure there are others out there who need a bit more bandwidth for their Bode plots!
Thanks Wolfgang for this idea to improve the Bode Plotter.
Indeed, it may not a big job to do this modification to the software.
I think this change will be done in the next version of the software
(At the moment I’m on vacation and I have no access to the PCGU1000/PCSU1000 hardware to make any test. The beta test version may be available earliest in two weeks from now.)
You may now also try to use the sweep function of the PCGU1000 and the spectrum analyzer with the FFT Option “Maximum” to get frequency response plots beyond the 1MHz limit of the Bode plotter.
That sounds great - can’t wait to see this new version!
Also thanks for the workaround suggestion - I’ll try that in the meantime. One similar setup I tried was to set the generator waveform to “white noise” and then use the scope in FFT mode set to “average”. That works, too, but is very limited in dynamic range (meaning, it’s not really possible to observe larger attenuation very accurately).
The beta test version of the software is now ready for testing.
You may download it from: vel255.diinoweb.com/files/PCSU1000GU_v3.11_beta.zip
There is only the PCSU1000GU.EXE file modified yet. This file is for the combination PCSU1000 + PCGU1000.
Just replace the old file with this one in the folder: C:\Program Files\Velleman\PcLab2000SE
The Bode plotter is now expanded to 5MHz. Seems to work quite well.
To get very accurate results please use the option “Normalize Output by Input”.
The option “Normalize Output by Input” was used.
When this option is used the CH2 of the scope must be connected to the second signal output of the PCGU1000 generator. This is similar connection as used to get the phase plot.
I tried to use the new version (V3.11 beta) today, but no luck. It always reports “No response from signal generator”. This accurs both when I try to launch the Bode plotter as well as when I try to start the signal generator from the scope window. The scope (PCSU1000) works fine, the problem is only with the generator.
Interestingly enough, the generator works perfectly fine when I launch software version 3.05 or 3.082. Only 3.11 beta has an issue.
I’m sorry about the confusion.
This version I send was based on recently released version 3.1.
This was the reason to the problem. The solution is: You have to download and install v3.1 and then replace this PCSU1000GU.EXE file to get this modified version running.
The main issue is that the version 3.1 needs new driver for the PCGU1000 to install.
There is also new driver for the PCSU1000 available though it runs with its old driver too.
For more details how to upgrade to v3.1 please see: pclab2000se_31_upgrade_drivers.pdf
(On Velleman downloads page.)
Ok, installed the new software version and updated the instrument drivers. The Bode plotter is working great now - thanks a lot! The “Normalize output by input” option is a helpful addition as well - I measured a short cable and the linearity of the plot was +/-0.1dB. This puts the PCSU/PCGU in a league well above their actual price range.
Two more suggestions:
First, would it be possible to allow selecting different voltage ranges for the measurement channel and the reference channel? That would be useful e.g. when analyzing an amplifier with a lot of gain - the input signal would be small (so scope channel 2 should be set to a smaller volt/div setting) and the output would be large (so CH1 should be at a larger setting).
Second, for very small signals any small noise affects the measurement accuracy, making the measured pk-pk signal appear larger than it really is. So could you add an option to turn on averaging for the Bode measurement, i.e. measure the output amplitude on the averaged signal instead of the single shot signal? Of course that will slow down the scan to some extent (not too much for higher frequencies), but would increase the accuracy.
It’s good to see that the modified plotter is working fine.
You got really amazing results !
[quote]First, would it be possible to allow selecting different voltage ranges for the measurement channel and the reference channel? That would be useful e.g. when analyzing an amplifier with a lot of gain - the input signal would be small (so scope channel 2 should be set to a smaller volt/div setting) and the output would be large (so CH1 should be at a larger setting).[/quote]Actually you can do it even now. You may freely select the channel CH2 Volt/div setting to get adequate waveform on the screen. - I think this option will be put to the documentation in the next version.
[quote]Second, for very small signals any small noise affects the measurement accuracy, making the measured pk-pk signal appear larger than it really is. So could you add an option to turn on averaging for the Bode measurement, i.e. measure the output amplitude on the averaged signal instead of the single shot signal? Of course that will slow down the scan to some extent (not too much for higher frequencies), but would increase the accuracy.[/quote]You are right - there may be some variation in the results on low levels.
Now the RMS value of a single shot data is used as an output to the plot.
I have succeeded to measure about 70dB dynamic range by using the option “Automatic Voltage Scale”.
Indeed, it looks good idea to use the existing averaging function of the scope for the Bode Plotter too!
There is one problem maybe: The averaging needs triggering.
Triggering may be a problem if the signal amplitude varies a lot. Especially if there is DC + AC at the output of the DUT.
Thanks for the hint with the voltage scale. So that’s solved.
For the averaging, I understand the need for a stable trigger. So maybe you just add a popup window that tells the user that trigger has be be turned on whenever you average over more than one waveform, just like you need trigger in equivalent time sampling mode. That way as long as the user does not employ averaging nothing would change in the handling.
I made some Bode Plotter test with averaging.
Indeed, averaging reduces noise and improves the dynamic range of the plot:
The advantage is about 10dB. With more noisy system it will be surely more than that.
This option will be for advanced users only. The user must know what to do and why.
The software modifications are minimal: There will be added to the Bode Plotter a menu option “Allow Averaging”. When selected, user can set the triggering on and select the number of averages.
The following popup will appear:
[code]You may now use the averaging option of the oscilloscope. Averaging reduces noise and increases the dynamic range of the Bode Plotter.
Connect the second output of the PCGU1000 to CH2 of the oscilloscope.
Select the Volt/div setting of CH2 to get maximum swing of the trace on the screen. Avoid clipping if you are using the options “Phase Plot” or “Normalize Output by Input”.
Put triggering on.
Select trigger source CH2.
Select proper number of averages using the Options/Average menu option of the oscilloscope.
Select proper wait time using Options/Wait Time menu option of the Bode Plotter.
The wait time must be long enough to allow the waveform stabilize on the oscilloscope screen after every frequency step.[/code]Not mentioned in the popup, but you can any time during the Bode Plot run click the “Pause” button, change the average settings and continue by pressing the “Pause” again.
There is a popup added to the option “Normalize Output by Input” too:
[code]This option normalizes the Bode Plot magnitude by the input magnitude of the DUT.
You may use this option to remove the effect of the input signal variations from the resulting magnitude plot.
This option is useful especially when measuring low impedance RLC circuits, whose input impedance varies with the frequency.
To use this option please do following:
Connect the second output of the PCGU1000 to CH2 of the oscilloscope.
Select the Volt/div setting of CH2 to get maximum swing of the trace on the screen but avoid clipping.[/code]
I hope these pop-ups are informative enough.
It works fine! Thank you!
Maybe you can go up to 10megs or higher?
I suggest new measurement mode “impedance” for circuit analyzer - the input of the circuit under test connected to generator via serial resistor with known value, CH1 connected to generator, CH2 connected to the input of the circuit (so we can measure voltage dropout on serial resistor) and software calculates impedance of the circuit under test and displays “impedance/frequency” graph…
[quote]Maybe you can go up to 10megs or higher? [/quote]Maybe not with the PCG10 but maybe with the PCGU1000… - Thank you for this idea!
In the PCG10 there seems to be about 0.8dB signal level decrease at 5MHz.
Using the Bode Plotter option “Normalize Output by Input” you can eliminate this slight generator output voltage decrease at 5MHz.
In the PCGU1000 there is much wider output range. The sweep function works fine up to 25MHz at the moment.
[quote]I suggest new measurement mode “impedance” for circuit analyzer - the input of the circuit under test connected to generator via serial resistor with known value, CH1 connected to generator, CH2 connected to the input of the circuit (so we can measure voltage dropout on serial resistor) and software calculates impedance of the circuit under test and displays “impedance/frequency” graph…[/quote] Thank you for this idea. This would be quite specific and a quite “professional” option for the circuit analyzer. - Interesting idea anyhow…
