Conventional vs arbitrary function generators

[japlytic]

These days, I recommend arbitrary function (waveform) generators over conventional function generators given arbitrary function generators are easier to use (especially with a waveform display), more flexible, consistent in output and offer better value for money considering that the design (and calibration) of an arbitrary function generator can be less expensive than a conventional function generator (depending on how many waveform types a conventional unit includes) nowdays.

Many arbitrary function generators include a number of built-in waveforms, some of which are complex e.g. damped sinewave, which can be difficult (and for certain waveforms, impossible) to generate on a conventional function generator.

I noted that the following calibrations have been dispensed with in arbitrary function generators (most of them have been replaced with software defined functions):
Typical Calibrations Dispensed With: Distortion, Output Harmonics, Square Wave Needle Suppression, Square Wave Rise/Fall Time, Square Wave Duty Cycle Adjustment, Internal Waveform Modulation (AM/FM etc.), Modulation Low Frequency Suppression, Burst Start/Stop Phase
Typical Calibrations Still Required: Output Amplifier Offset, Output Amplifier Frequency Compensation, Master Clock, External Waveform Modulation Level

[eccerr0r]

Re: Conventional vs arbitrary function generators

Of course an arbitrary generator can do anything a fixed function generator, problem being, sometimes it's real tough to generate an "arbitrary" function at a high frequency without distortion at low cost...

So it depends on the situation.

[japlytic]

Re: Conventional vs arbitrary function generators

Sometimes it's real tough to generate an "arbitrary" function at a high frequency without distortion at low cost...

It depends on what filters you switch in and how good the filters are at a given frequency.

[eccerr0r]

Re: Conventional vs arbitrary function generators

and depends on how good the filters are for a specific waveform, not to mention adding the filters increases costs.

[Longbow]

Re: Conventional vs arbitrary function generators

I disagree with your entire premise that an AWG is inherently a better type of instrument than a simple AFG. The 2 features are almost exclusive in their practical use, to say the least.

Obviously it won't make sense to state a distortion figure for a truly arbitrary waveform because there won't be a reference waveform for comparison. Offering specialized built-in waveforms would correctly be referred to as arbitrary functions, whereas a user-defined output would be available in an arbitrary waveform generator.

Arbitrary waveforms are almost always used in manufacturing testing and burn-in facilities that subject a piece of equipment to a particular testing regimen. It wouldn't make sense to spend the extra bucks on this capability if your needs are simpler. At least, manufacturers justify a higher cost for added features. Last time I checked, the built-in waveforms in an AWG were no different than the built-in waveforms in an AFG.

The real reason that modern digital units dispense with calibration procedures is that the waveforms themselves, already exist in the unit's ROM, or exist as instructions to a math processor on how to produce them. Calibration procedures that you refer to would only be necessary in analog devices such as the familiar analog sine, square, triangle generators that were the norm until recently.

I personally would recommend an analog sine, square, triangle generator for 90% of bench uses. Spend the extra money where it is needed such as on a nice scope. The extra features on a DDS generator are cool, but usually unnecessary for typical equipment repair and testing. Just a personal opinion.

[llonen]

Re: Conventional vs arbitrary function generators

Conventional for me

[stj]

Re: Conventional vs arbitrary function generators

you could probably make an arbitrary one cheap.
you can get a Nucleo dev board running at 400MHz for $20
that has usb & ethernet.

just add a very fast DAC and use dma to stream an uploaded wave-table to the dac from ram.

[eccerr0r]

Re: Conventional vs arbitrary function generators

I don't really have a waveform generator at all :o

I do have an AVR software based DDS that I hacked up for the heck of it if I need it (using an R-2R ladder for DAC; overclocked to 14.3MHz so it has a maximum of a bit more than 1.5MSPS from the wavetables), and a RF signal generator. With an analog oscilloscope I can easily see the waveforms steps in each of the waveforms on the DDS... Alas I don't have a filter on the output, and I need to finish it by adding offset/amplitude control as well as decreasing the output impedance. Not much you can do with a 90S2313 DDS.

[llonen]

Re: Conventional vs arbitrary function generators

Yes I could probably construct an arbitrary one cheap enough I don't make that much use of the conventional one I have to justify the work load.

[japlytic]

Re: Conventional vs arbitrary function generators

For new units on eBay, there are more listings for arbitrary units than conventional units.

[stj]

Re: Conventional vs arbitrary function generators

probably because conventional ones mostly used a special chip that is getting hard to find,
microcontrollers on the other hand, are getting cheaper and faster at an amazing rate.