waveform generator

[mastershake]

looking to spend ballpark around 400 ish give or take on a decent one. i know this will not get me top of the line. id really like to find something 80-100mhz. they all seem to have one or more things that kind of ruin it overall like one example has 100mhz 16bit etc but square wave is only 2mhz and triangle is 10. not sure how often ill use those but still then others that have say 50 square wave are like 60-maybe 80 overall and usually a chap quality (of course) for that kind of money.

is there anything that gets really recc here in that 4-500 price range that does most things really well without to much noise etc and has most things that would be used. this will be passed onto a student once they start to need or use it so im hoping to find something somewhat decent.

i can get a GREAT deal on the 100mhz hantek 3000 series right now but i know how most feel about them here and the specs are really only meh once you really look at them. just curious if one model tends to stick out in that price range and gets recc here more often.

[tautech]

An 'improved' SDG1032X will get you to 60 MHz which is good value.
OTOH an 'improved' SDG2042X can get you to 120 MHz, the limit of Bode plot BW in Siglent DSO's however that is at the top end of your desired spend.

[mastershake]

by improved do you mean those are modded? im trying to read all i can but with my eye sight its very slow going now.

[tautech]

by improved do you mean those are modded?

Yes.
Others can provide guidance. 

[RoGeorge]

those are modded?

Not modded, but fully unlocked.  Usually, a company produces only the top of the line instrument, then lock it down by software and sell it cheaper, in the hope the buyer will buy later more software "features", which can unlock the instrument to its maximum performance.  The hardware is the same, only the sticker with the model is different (and the software locks) across a given line of instruments.  For example one can buy online a "frequency upgrade" for a low end generator, which is nothing but a key number that will unlock the software, so the same instrument can go upper in frequency.  Nothing changes in hardware.

There are ways to generate these keys and undo all factory locks for free, so one can buy the cheapest version of an instrument, and unlock it later, so to get the same performance as if the instrument were the top of the line.

[Performa01]

It's the same old story again: some T&M gear - an AWG in this case, that costs next to nothing, yet it has to be "decent" nonetheless.

"Decent" in this case appears to be "80-100mHz". I very much doubt this. I suspect it should rather be 80 - 100 MHz. Is it really too much asked to use correct terminology in an EE-forum? "m" = milli, "M" = Mega.

These 80 to 100 MHz apply to sine waves. Of course square and triangular waves have much lower frequency limits. A 100 MHz AWG has to have a 100 MHz reconstruction filter (ideally brickwall) at its output(s). Guess what happens if we tried to output e.g. a 50 MHz square under those conditions. Of course it would be a pure sine.

So it's up to the manaufacturer, how much deviation from the ideal waveform they think the customer will be tolerating. They'd have to limit the square wave output to max. 20 MHz in a 100 MHz AWG, in order to include at least the 5th harmonic, thus getting sort of barely acceptable signal fidelity. Yes, there are rare exceptions, like the Siglent SDG1000X series, which have a dedicated circuit for square wave output - but that's quite some effort, rarely seen anywhere else and even in the higher class AWGs from Siglent, because while it's easy to convert a sine to a square by means of a comparator, it's not so straight forward to recover the amplitude.

It is similar with triangular waves. What amount of nonlinearity are we willing to tolerate? 1 %, 2 %, 5 %...? The whole reason to have triangular waves (and ramps) traditionally was for linearity tests. That's why we cannot tolerate high non-linearities, hence triangular waves and ramps are limited to low frequencies - usually 1 MHz or even lower. On the other hand - what would be the use of a high frequency triangle wave?

The above hints apply to quality gear. Some manufacturers might not bother and present you quare waves that are almost indistinguishable from sines and you get heavily bent triangular waves. Yes, the user still has the option to limit the use of those waveforms to lower frequencies, but the fact remains that inexperienced users might get fooled by the false claims in the specs of such products.

[TurboTom]

The choice of an AWG depends on many details, even if it's just for hobby use. For example, if you already own a Siglent scope that's capable of doing bode plots, you may want to get the AWG from Siglent as well so you can combine it with your scope to actually enable the system to do these types of measurement. You may want to consider the type and quantity of auxiliary inputs of your AWG. If you need to have two channels with individual external modulation (or sync output or what auxiliary function you may consider), none of Signlent's AWGs within your price range will check the box for you since their AWGs only feature one combined AUX BNC.

If you just need the AWG without any affiliation to a "brand family", and you consider "liberating" (hacking/unlocking) an option, the current best-value-for-money buy would probably be Rigol's DG811 (~250 USD), improved to (almost) DG992 features, if you can live with the ridiculous form factor that is  .

[coromonadalix]

i have "modded" mine, and works very well, modded other Siglent stuff too,  all works wonders and i saved $$$$

the know how  are on this forum .. received some help for an option ...


love the fact  Siglent provide additionals stuff like labview examples / tools ....   in one case  we will try to automate some tests

i managed to create a simple program to control a generator, it was my 1st tryout at this

Rigol are "moddable" too, but a lack of theses extras made us change to Siglent