Function generator with simultaneous modulations

[maxwell3e10]

I am looking for a function generator that can simultaneously be controlled by burst count and external amplitude modulation. It seems most of them allow only one at a time but this is poorly documented.  DS345 from Stanford research systems can do it, but its pretty old and only allows 30000 burst counts, while I need 10^6, as most modern ones do. Any suggestions? Perhaps someone has used this mode or would be willing to test theirs.

[capt bullshot]

Just a quick test result:
On the Agilent 33522B modulation and burst is exclusive, activation of either one turns off the other one

[kultakala]

I assume it will blow your budget...  but while looking at some oscilloscopes and function generators on the Rohde & Schwarz site i stumbled over the HMF2525.
Its about 1700 $ 

But in the manual it says:

- AM modulation: Only one type of modulation may be active at any time.
- Burst mode: The BURST mode is available for each waveform, also for any symmetry setting.

In the german version its says even more clear that burst mode works for any kind of signal.

So it seems that this puppy could do it, but maybe i am misinterpreting it ?

[maxwell3e10]

Perhaps it would work. At least logically those two functions shouldn't interfere. One small problem is that it says that the burst mode is limited to 50,000 cycles. But it has a synchronous gated mode, which I think might have similar functionality.

[tautech]

A bit of a fiddle with a SDG1062X gets this:



Ch 1, REF=Internal, 50 KHz Sinewave @ 5V p-p, Burst 50 KHz, 1 cycle (50 us). Start Phase=270^o

Into: Aux In/Out

Ch 2, 1 MHz Sinewave @ 1V, MOD AM, Source=External..............to scope.

[maxwell3e10]

Thanks, this is interesting. From what I can tell SDG1062X has only one input BNC, so one couldn't put in external modulation and externally trigger the burst. On the other hand SDG1025 has separate modulation and trigger inputs, so perhaps it can do both?

[tautech]

Thanks, this is interesting. From what I can tell SDG1062X has only one input BNC, so one couldn't put in external modulation and externally trigger the burst. On the other hand SDG1025 has separate modulation and trigger inputs, so perhaps it can do both?

Correct.

I only had a 1062X and my aging 1010 on hand that had 2 channels to try this.
The older 1 k series I think are slowly being phased out to be replaced with 1kX as our basic 2 channel AWG.

I did achieve the same result with the SDG1010 but the quality of the waveform was inferior to the 1062X and I didn't want to post it, but despite having less inputs a 1062X was able to do this job.
On the other hand the SDG2042X that seems to be a forum favourite has the additional inputs and the waveform quality would have been much better still. Sold my last one a week ago so couldn't use that.

If you have more than basic needs for an AWG, the SDG2042X is the one to go for. Check out the thread here on the blog, lots of happy owners. 

[nugglix]

Hi!

I repeated the experiment tautech did w/ a SDG2042X.

Channel 2: 1kHz triangle wave, burst, 4 cycles -> AuxIn
Channel 1: 50kHz sine, AM-DSB modulated from AuxIn

Seems to work well.
And learned something.

[maxwell3e10]

This looks nice. What I think is going on here is that when the burst is off, the Aux-in signal is zero. This makes the Ch1 output zero (in AM-DSB) mode or small (in AM mode). So one can't say the same channel has burst and analog modulation.

Its actually a bit different from what I had in mind (which was not very clear): The high frequency  signal (say 10 kHz) is in burst mode, has say 10000 cycles. While its on, its amplitude is modulated at a lower frequency (10 Hz). So the same channel is being AM modulated and bursted.


Its too bad the updated function generators from Siglent are stingy with inputs. What I really would want is to have two separate analog modulations for the 2 channels. There are several function generators that have two modulation inputs, I found BK4075B, GW/Instek AFG3022, Rigol DG5072 and Textronix AFG3022. But I suspect that they may not be able to do analog modulation and burst on the same channel.

[tautech]

So one can't say the same channel has burst and analog modulation.

Correct
But there is more than 1 way to skin a cat. 

Its actually a bit different from what I had in mind (which was not very clear): The high frequency  signal (say 10 kHz) is in burst mode, has say 10000 cycles. While its on, its amplitude is modulated at a lower frequency (10 Hz). So the same channel is being AM modulated and bursted.

Which is hard to demonstrate on a scope at such low frequencies.

Its too bad the updated function generators from Siglent are stingy with inputs.

Only the new base model, 1kX and like I said that's the best I had to try. 
There's a # of tricks you can do with BNC patch cables and Tee's. 

What I really would want is to have two separate analog modulations for the 2 channels. There are several function generators that have two modulation inputs, I found BK4075B, GW/Instek AFG3022, Rigol DG5072 and Textronix AFG3022. But I suspect that they may not be able to do analog modulation and burst on the same channel.

This is where you'll struggle to find something that can enable both options (Mod and Burst) in the user UI.
SCPI code is where you'l probably have to go for this to have 2 channels doing Burst and Mod but I can't advise if even that can achieve what you want, others might have a try with this for you.

[MarkL]

If you're unable to find a unit that does both burst and AM, you might want to take a look at doing the modulation externally and leave the burst count up to the function generator.

For a fast solution, you could grab an analog multiplier eval board and use that as your modulator.  You didn't say anything about your voltage, AM depth, or distortion requirements, but here's a range of possibilities from Analog Devices:

  http://www.analog.com/en/products/linear-products/analog-multipliers-dividers.html

[maxwell3e10]

Yes, already have an external analog multiplier to do this, but was looking for something with more accuracy. I think old function generators would implement AM with an analog multiplier, while newer ones digitize the external modulation and do the multiplication digitally. I am not sure how DS345 does it.

If the basic functionality is there on some of these models, just not accessible through UI, that would be OK. Its just hard to tell without investing in the instrument.

[joeqsmith]

I take it that it's not something that repeats where you could just upload the modulated wave and trigger it multiple times.  You want to modulate the burst live.   

What other requirements do you have? 

[MarkL]

If you don't care about the precise number of cycles in the burst, you could do it with a function generator that has a gate input and an AM modulation input.  You would use a second function generator to feed the gate input on the first to create the needed burst size.  The second function generator could also be triggered to start the burst.

The old HP 8116A and the older Tek FG504 both have separate inputs for gate and modulation.  I tried it on both and they can both do gate and AM at the same time.  But if you were looking for a more modern unit, these definitely don't qualify.


EDIT:  I should also note that both units are also analog implementations, so they are not particularly accurate.  The FG504, however, has a phase-lock input so you could feed it a reference if you wanted an exact and stable output frequency.

[maxwell3e10]

Yes, its possible to use arbitrary waveform, but its clunky because if the two frequencies are not easily commensurate, it needs to be very long. And to scan one of the frequencies requires updating the waveform every time.

The gate input could work, the only possible issue is that its typically not phase-synchronous, so can cause some little problems.

I wonder if I contact the manufacturers, if I would get a response one can actually trust, as its a relatively obscure feature.

[tautech]

I wonder if I contact the manufacturers, if I would get a response one can actually trust, as its a relatively obscure feature.

Siglent have application engineers in Ohio, give them a yell.
http://www.siglentamerica.com/contact_us

[Siglent America]

I am looking for a function generator that can simultaneously be controlled by burst count and external amplitude modulation. It seems most of them allow only one at a time but this is poorly documented.  DS345 from Stanford research systems can do it, but its pretty old and only allows 30000 burst counts, while I need 10^6, as most modern ones do. Any suggestions? Perhaps someone has used this mode or would be willing to test theirs.

Hello,
We were informed of your application so we scratched our collective heads to see if there was a way to do this using two units, the channel summing mode, etc. The basic problem is that we cannot do burst and modulation at the same time. So we have two possibilities on this, if you are willing to do some extra work:

1) Connect to a PC and program the instrument to output the desired AM signal. Before enabling the output, calculate the time required for the desired number of bursts. Enable the output for the time calculated previously. After the time expires, set the amplitude to 0V (or as close as you can). Shutting off the output will stress the relay and could lead to premature failure if it’s pushed too hard by excessive switch counts.

2) Use CH 1 to generate the desired AM signal. Run that to an external switch or gate You could use a simple FET gate using only a few components. Now use CH 2 from the generator to drive the gate input of your FET switch/gate circuit.

Interesting application. Please let us know what you decide to do.
Good luck!

[MarkL]

The gate input could work, the only possible issue is that its typically not phase-synchronous, so can cause some little problems.

What do you mean by "phase-synchronous"?

If you're referring to the start and stop of the waveform, the gate controls the number of full cycles output of the selected waveform.  The gate is not a simple on/off of the output.  For example, if your output waveform is a sine, the sine will start and stop at 0 degrees.  Some function generators allow you to select the starting phase from -180 to 180 degrees, but it will still always be an integer number of cycles as controlled by the gate input.

If by "phase-synchronous" you mean that you want the carrier to be in phase with itself from burst to burst, then perhaps what you want *is* a simple on/off of the output.  In which case a CMOS SPST switch on the function generator output, controlled by the burst signal, might work for you.

Or are you looking to somehow synchronize the burst or carrier start phase to the AM input?

[maxwell3e10]

Thank you for attention to the forum. There are a number of ways to do it using external components, either to generate AM modulation with a multiplier or to generate burst with a switch. My hope would be to find an internal solution. This should be possible in principle, but depends on the method in which the burst mode is implemented, as an (internal) arbitrary wave or as a continuous wave with a separate counter.

Based on my reading of various manuals, my best guess is that BK Precision 4078B might be able to do it, as the modulation and running mode are separate functions. Also it has 10 ! inputs on the back, so all features are available for both channels at all times.

[maxwell3e10]

Regarding gate modulation, I am not sure if its usually an integer number of cycles or simple on-off. I think some function generators have an option for either, but I haven't checked which is more common in general.
What I would prefer is the starting phase of the carrier to be the same at the gate trigger.

[MarkL]

Regarding gate modulation, I am not sure if its usually an integer number of cycles or simple on-off. I think some function generators have an option for either, but I haven't checked which is more common in general.
What I would prefer is the starting phase of the carrier to be the same at the gate trigger.

The two I mentioned, 8116A and FG504, both have an integer number of cycles with a selectable start phase.  I checked before I posted.  But I can't speak for other models.

Good luck on your search.

[Siglent America]

Hi MarkL.
The Siglent SDG1000X and 2000X families do have a calibrated adjustable Start Phase.

[Plasmateur]

Thanks, this is interesting. From what I can tell SDG1062X has only one input BNC, so one couldn't put in external modulation and externally trigger the burst. On the other hand SDG1025 has separate modulation and trigger inputs, so perhaps it can do both?

Correct.

I only had a 1062X and my aging 1010 on hand that had 2 channels to try this.
The older 1 k series I think are slowly being phased out to be replaced with 1kX as our basic 2 channel AWG.

I did achieve the same result with the SDG1010 but the quality of the waveform was inferior to the 1062X and I didn't want to post it, but despite having less inputs a 1062X was able to do this job.
On the other hand the SDG2042X that seems to be a forum favourite has the additional inputs and the waveform quality would have been much better still. Sold my last one a week ago so couldn't use that.

If you have more than basic needs for an AWG, the SDG2042X is the one to go for. Check out the thread here on the blog, lots of happy owners. 

I could not agree more. The SDG2042X is the best bang for your buck on the market. I am very happy with my purchase so far.....

Now if they could just get both outputs to trigger from the same software trigger....well, I'd be super super happy

[maxwell3e10]

So I am looking again into the issue of simultaneous modulation, for a different application. While many signal generators (analog synthesized sine wave) have simultaneous modulation capabilities (e.g. AM+FM), most arbitrary waveform or function generators (direct D/A output) don't have it. I couldn't find even one example of an AWG that clearly advertises simultaneous modulation. Of course one could create an arbitrary waveform that will have modulations, but it is cumbersome. I don't think there is a technical obstacle to have simultaneous modulation, so maybe some such generators do exist?