Using sweep generator and a scope to display filter response. FY6800 suitable?

[strayr]

I saw a video of someone using a signal generator to output a sweep and a ramp signal and then set his scope to trigger on the ramp and there was the most useful realtime picture of a filter envelope. https://youtu.be/-A_DxsxPdeI?t=300

I want to do this in the audio domain, 40hz to 10kHz required, 10Hz to 40kHz nice. I've got a few scopes, and I can do very slow (but decent quality) VNA with my syscomp and dump the bode plot to .csv and do further analysis. But I really want to be able to tune filters in realtime. I'm looking at this KKmoon which i understand to be a rebranded feeltech FY6800.

Will I be able to get a suitable trigger signal out of the fy6800 to do this? I assume it would by output from the sync out but the manual is not particularly helpful.

[strayr]

is the poor man's option a square/pulse on channel 2 set to the period of the sweep on ch1? Would that be viable? is there a better way to do this? Am I overthinking this?

[retrolefty]

I think the thing lacking with that setup would be a lack of a logarithmic response as filter response traits really requires a fairly large amplitude response.

[strayr]

Page 35 of the manual suggests you can do a log or linear frequency sweep if that's what you mean?

[aneevuser]

I've tried and failed to do this with a FY6800. I was unable to find any way to generate a suitable trigger signal. I believe that I tried (and failed) with the second channel idea, but I don't recall the details now. If you find a suitable method, please let us know.

[rstofer]

A better way to do it would be the Digilent Analog Discovery 2
https://store.digilentinc.com/analog-discovery-2-100msps-usb-oscilloscope-logic-analyzer-and-variable-power-supply/

[strayr]

That's quite expensive, is that actually going to do what I'm asking and let me tune a filter in realtime, or is this going to step through the frequencies and produce a nice high quality bode plot after a (not so) brief wait which I can already do with my syscomp CGM? granted it looks like it's got better software than the syscomp, but I'm specifically looking for an affordable sweep generator with some way of syncing up my scope to view the envelope of a sweep.

[retrolefty]

No, in amplitude, scopes usable only offer linear vertical response not logarithmic.

[strayr]

Ok, am I trying to do something that can't be done in realtime, or is this just very expensive? Or something people don't do very often?

[retrolefty]

Ok, am I trying to do something that can't be done in realtime, or is this just very expensive? Or something people don't do very often?

 On the contrary you simply have to add an inexpensive log amp stage between the filter output being analyzed and the input of your scope. Here is one similar  to what I have used in the past.

https://www.ebay.com/itm/AD8307-RF-Power-Detector-Module-DC-to-500MHz-Transmitter-Power-Test-92dbm/193398869569?hash=item2d07788e41:g:oIQAAOSw9D1effwJ

[rstofer]

That's quite expensive, is that actually going to do what I'm asking and let me tune a filter in realtime, or is this going to step through the frequencies and produce a nice high quality bode plot after a (not so) brief wait which I can already do with my syscomp CGM? granted it looks like it's got better software than the syscomp, but I'm specifically looking for an affordable sweep generator with some way of syncing up my scope to view the envelope of a sweep.

Yes, it's pricey.  I haven't used the feature but it is possible to sync the two channels of the Arbitrary Waveform Generator so I don't see why it wouldn't work.  Generate a square wave for trigger and whatever for the actual signal.

Otherwise, you're right, it takes a while to generate the Bode' Plot.  How long it takes depends entirely on the lowest frequency of interest because there are some particular number of samples per decade.  So, 1..10 Hz takes just as long as 10..100 Hz or 100..1000 Hz and so on.

I did a quick experiment to prove to myself that I could sync the sweep to the square wave.  Attached are the plots.  I didn't spend much time on refining the output but it does work.  It does need a little refining and which waveforms to display needs a little more thought.  Probably the square wave and the filter output.  The sweep is what it is...

[ozcar]

...
Will I be able to get a suitable trigger signal out of the fy6800 to do this? I assume it would by output from the sync out but the manual is not particularly helpful.

I have done this with a FY6800, but not by using the SWEEP function. Instead, set channel 2 to generate ramp for the sweep, then set MODulation for channel 1: MODE FM, SOURCE CH2, and PARA (BIAS) to set the sweep range. You can then trigger off CH2 (I think I tried Y-X mode but that was a bit of a disaster).

[rstofer]

It seems to me that this sweep method doesn't provide a decent X axis.  What frequency does each unit on X represent?   

With the AD2 Network gadget, the Bode' Plot is properly annotated and in terms of dB versus log(frequency).  The units in Y are selectable but default to dB

There is also an Impedance gadget that I haven't used very much but it sweeps frequency and creates a plot of Phase and Impedance, similar to the Bode' Plot.  Again, properly annotated.

It's also possible to generate a piece-wise waveform so the first 't1' seconds is at a low frequency, the next 't2' seconds is at a higher frequency and so on.  At least for any given piece, we know what the X axis frequency should be.

About half way down:
https://reference.digilentinc.com/learn/instrumentation/tutorials/ad2-waveform-generator/start

I don't see any satisfaction in any of these ad hoc methods.  The AD2 Network tool does a fine job and presents the Bode' Plot properly, fully annotated.  It's reasonably quick if the starting frequency isn't too low.  0 Hz is a really bad idea.  Starting at 1000 Hz for an audio low pass filter might be entirely adequate.

[strayr]

I have done this with a FY6800, but not by using the SWEEP function. Instead, set channel 2 to generate ramp for the sweep, then set MODulation for channel 1: MODE FM, SOURCE CH2, and PARA (BIAS) to set the sweep range. You can then trigger off CH2 (I think I tried Y-X mode but that was a bit of a disaster).

I've gone and bougght one of these because i could actually get hold of one and it does a few other things that I want and I'm not finding a better solution (yet). I can make something rudimentary and useful happen with VCO, trying to get my head around mod, am I right that if channel 1 iud set of F1 and chanel 2 has an amplitude of V2 and the bias is set to B, I'm going to have a frequency range of B x V2? and with no offset distributed equally each side of F1?

[strayr]

On the contrary you simply have to add an inexpensive log amp stage between the filter output being analyzed and the input of your scope. Here is one similar  to what I have used in the past.

That could be handy, the guitar pickup testing community is big on an op-amp integrator and I need to see if I can get good results with mine here, but for proper results i've bbeen dumping bode p[lots to CSV and processing with R

[strayr]

...AD2 Network gadget...

The AD2 does look quite cool, I'm a bit concernbed about its durability and it kind of covers some stuff I already have.

I'm trying to use the sweep gen to make set up pots a pedal and other levels to a sane value before I spend time running a bode plot to get real numbers, similarly if I'm swapping components, it's the quick test for sanity before I spend time taking readings. Becassue I'm post processing and keeping data to compare (particularly for guitar pickups) I'm making bode plots at fairly high resolution, and doing tricks like switching in a capacitor of a known value and measuring the resonant peak to calculate the inductance and feeding that number back through to measure the capacitance of the unloaded system. It works reasonably well.

[strayr]

Winding back to the initial problem, at audio frequencies, I'm running at 200m/s /div or slower timescale so the scope is scrolling anyway, triggering isn't really a thing. N00b oversight on my part. But i'm getitng a visual I can use.

[ozcar]

I have done this with a FY6800, but not by using the SWEEP function. Instead, set channel 2 to generate ramp for the sweep, then set MODulation for channel 1: MODE FM, SOURCE CH2, and PARA (BIAS) to set the sweep range. You can then trigger off CH2 (I think I tried Y-X mode but that was a bit of a disaster).

I've gone and bougght one of these because i could actually get hold of one and it does a few other things that I want and I'm not finding a better solution (yet). I can make something rudimentary and useful happen with VCO, trying to get my head around mod, am I right that if channel 1 iud set of F1 and chanel 2 has an amplitude of V2 and the bias is set to B, I'm going to have a frequency range of B x V2? and with no offset distributed equally each side of F1?

I was not sweeping over audio range.

For what you want I think you'd have to set the PARA/BIAS to the full range, so 40kHz - 10Hz = 39.990 kHz, and the channel 1 frequency to the middle of the range = 20.005 kHz. From memory I don't think it uses the amplitude set on channel 2 - could be wrong though. I do remember messing about a bit to figure how it worked, like temporarily setting channel 2 to to a slow square wave, so I could see channel 1 alternating between the min and max frequencies.

[geggi1]

Google soundcard audio spectrum analyser software. You will get plenty of hit.
If you have labview you can even make the software your selve. But you need a quality soundcard.