Your M2 will happily generate sinewaves out to 44kHz. You can even confirm their 2nd harmonic level, as it is within the 96kHz bandwidth. The REW method to measure impedance of a part uses such a test signal that extends to 96kHz - ie. the impedance of a part is measured out to 96kHz.
Your particular test setup may, or may not, benefit from a separate signal generator. If your DUT has a common input and output ground then there is no concern imho. If you are testing a device that has a convenient input socket, and output socket, and the grounds of each socket are internally subject to a noisy ground interconnect then that can show up in a measurement, or can be alleviated by disconnecting the output ground going to the soundcard, to avoid noise from an internal ground link. I do that a lot for amplifier testing. One advantage of a spectrum response is that often any ground loop related noise is clearly identifiable as such, and likely has no impact on your particular measurement result, other than aesthetic displeasure of seeing other signals in the spectrum (imho most audio people like to post spectrum plots on forums that are 'clean'). All test setups need awareness and checking to ensure confidence that your particular measurement outcome is valid.
Ground loop noise related to grounding of your M2 and PC is a different matter - and if noise artifacts arise from the PC connection to mains earth, and are a practical concern, then use a laptop (I sometimes use a laptop for that reason - horses for courses).
Afaik, xrk971 could have used a soundcard for that test in post #20 that uses a 60kHz sinewave.
Your particular test setup may, or may not, benefit from a separate signal generator. If your DUT has a common input and output ground then there is no concern imho. If you are testing a device that has a convenient input socket, and output socket, and the grounds of each socket are internally subject to a noisy ground interconnect then that can show up in a measurement, or can be alleviated by disconnecting the output ground going to the soundcard, to avoid noise from an internal ground link. I do that a lot for amplifier testing. One advantage of a spectrum response is that often any ground loop related noise is clearly identifiable as such, and likely has no impact on your particular measurement result, other than aesthetic displeasure of seeing other signals in the spectrum (imho most audio people like to post spectrum plots on forums that are 'clean'). All test setups need awareness and checking to ensure confidence that your particular measurement outcome is valid.
Ground loop noise related to grounding of your M2 and PC is a different matter - and if noise artifacts arise from the PC connection to mains earth, and are a practical concern, then use a laptop (I sometimes use a laptop for that reason - horses for courses).
Afaik, xrk971 could have used a soundcard for that test in post #20 that uses a 60kHz sinewave.
Hi,
"I'm more interested in accurately characterising discrete components than thd of solid state amplifier ....."
I'm surprised to hear that now, but then ... You didn't define what You needed a function generator for in first place.
This requirement though rather asks for something like a Curve tracer or an SMU.
In that case You may want to look for Digilent's Analog Discovery.
This is a small USB connected multi-functional device that allows working as Oscope, Curve tracer, Spectrum analyzer, Bode plotter, etc.
You can for example measure/analyze the impedance response of caps or inductances.
Utilizing a 14Bit ADC and DAC it's resolution is typically better than from an Oscope.
The maximum Bandwidth of 30MHz iIrc outranges any classic audio analyzer.
You probabely need to buy or build a few 'breakout boards' to achieve proper connectivity.
You can find video tutorials at Digilent, as well as Analog Devices (their ADALM2 is the same class device, actually the predecessor).
Or maybe You define Your requirements again for us and Yourself, new and clear ;-)
jauu
Calvin
"I'm more interested in accurately characterising discrete components than thd of solid state amplifier ....."
I'm surprised to hear that now, but then ... You didn't define what You needed a function generator for in first place.
This requirement though rather asks for something like a Curve tracer or an SMU.
In that case You may want to look for Digilent's Analog Discovery.
This is a small USB connected multi-functional device that allows working as Oscope, Curve tracer, Spectrum analyzer, Bode plotter, etc.
You can for example measure/analyze the impedance response of caps or inductances.
Utilizing a 14Bit ADC and DAC it's resolution is typically better than from an Oscope.
The maximum Bandwidth of 30MHz iIrc outranges any classic audio analyzer.
You probabely need to buy or build a few 'breakout boards' to achieve proper connectivity.
You can find video tutorials at Digilent, as well as Analog Devices (their ADALM2 is the same class device, actually the predecessor).
Or maybe You define Your requirements again for us and Yourself, new and clear ;-)
jauu
Calvin
@Calvin Sorry for the confusion.
I'm building a couple of power amplifiers and more importantly a power supply for them. I'd like to be able to examine what's going on inside the circuits - so I decided I'd better get an oscilloscope - and I'd like to measure some of the components - hence the signal generator.
Both of those would be around the same price as a new QA403 - $720 from the US - a bit more than the Analog Discovery $520 which is similar to a good soundcard solution - Cosmos E1DA ADC + Topping DAC say $500 - with REW.
Hope that helps clarify my requirements.
I'm building a couple of power amplifiers and more importantly a power supply for them. I'd like to be able to examine what's going on inside the circuits - so I decided I'd better get an oscilloscope - and I'd like to measure some of the components - hence the signal generator.
Both of those would be around the same price as a new QA403 - $720 from the US - a bit more than the Analog Discovery $520 which is similar to a good soundcard solution - Cosmos E1DA ADC + Topping DAC say $500 - with REW.
Hope that helps clarify my requirements.
You seem to want to use close to the best available tools, when it appears you haven't yet used such tools before to adequately appreciate what level of performance you need or could use. I say this as you are bringing up equipment costs, and perhaps don't want to spend money, but then again do want to spend money. If money is no problem then just go ahead and buy equipment, as you won't know what you need until you get down to bench testing.
If money is a concern then I'm not sure why you haven't been doing bench tests on your target equipment using the M2 tool you already have, as that would fast-track your own assessment of what level of performance and convenience you need.
If money is a concern then I'm not sure why you haven't been doing bench tests on your target equipment using the M2 tool you already have, as that would fast-track your own assessment of what level of performance and convenience you need.
I agree:
As to cost, yes I do consider cost, but not at the expense of performance. I expect to use the kit I buy for many years - it's cheaper that way - and I'll buy used if the opportunity presents itself.
It's been very interesting hearing what kit people use, and what works well.
Calvin's suggestion of the Analog Discovery looks like it would be a good fit.
Otherwise, it seems like the SDS800X HD oscilloscope + SDG2042X AWG would also work, although I note Calvin's preference for an audio analyzer below 96kHz, presumably for it's greater bit-depth. And use the M2 for lower frequencies.
Whilst I prefer the usability of dedicated instruments, if the Analog Discovery can replace two Siglents then I guess that's the one to go for.
which I think you're giving me a heads up with. But I won't get down to testing until I have some kit. And I'd like to be able to put a square wave into an amplifier with a frequency response up to say 400kHz and measure the results.
As to cost, yes I do consider cost, but not at the expense of performance. I expect to use the kit I buy for many years - it's cheaper that way - and I'll buy used if the opportunity presents itself.
It's been very interesting hearing what kit people use, and what works well.
Calvin's suggestion of the Analog Discovery looks like it would be a good fit.
Otherwise, it seems like the SDS800X HD oscilloscope + SDG2042X AWG would also work, although I note Calvin's preference for an audio analyzer below 96kHz, presumably for it's greater bit-depth. And use the M2 for lower frequencies.
Whilst I prefer the usability of dedicated instruments, if the Analog Discovery can replace two Siglents then I guess that's the one to go for.
I'm in the process of getting an oscilloscope - probably a Siglent SDS804X HD - and thought it'd be handy to add a waveform generator.
I'm fine with doing analysis on my Linux laptop, and was looking at the Uni-T UTG-962G ~£147, and the Siglent SDG1032X ~£277.
So what are you using, and how do you like it?
@Giallograle : I have the SIGLENT SDG-810 function generator, among some other SIGLENT units : SDS-1202-XE oscilloscope and SDM-3045-X digital multimeter. These are fine quality, reliable instruments, quite easy to set up quickly...
T