Which function generator is better? Agilent 33120A or Gw Instek AFG 2025 ?

[electronic_guy]

Hi,

I'm looking to buy a function generator for my home lab and I'm bit confined with available budget to around $400 shipping included.
I read lot of posts here that speak high of Agilent 33120A. I'm personally biased to buy HP Agilent instruments my self. But locally I have brand new Gw Instek AFG 2025 for about $385 also. So I'm bit stuck if I want to go with my intuition to buy Agilent 33120A or go with a brand new Gwinstek AFG 2025.
Also I don't know of there are any other good function generators to buy for my application. I'm okay with a clean signal upto 15 MHz.

(I haven't used any of Rigol and siglent, if I'm not wrong they sound like bit blown big by some fans here, I could be wrong but I sensed a pattern of continuous recommendations to those two brands. I'd keep an open mind about this)

My applications are beginner stuff like amplifiers, OPAMPs and digital circuits related to filters and signal processing.
My target is to go for RF stuff in the range of 1GHz to 3 GHz. For an RF signal generator I already have a CMU200. But I need a function generator in low side to try out the fun stuff of being a beginner.

Thanks.

[artag]

My inclination is also to go for the older equipment. I have the 33120 and am happy with it, but haven't compared anything else except the analog HP3312.

The only thing I have against the 33120 is that the base version doesn't have the ability to lock to an external reference - if that's important, you need to find one with that option. It's quite a complex component and rarely seen as an upgrade but I have seen a moderate number with it already fitted.

[GerryR]

33120  ; I had one for years and sold it when things slowed down, which I regret.  HP older stuff were built like tanks; the new stuff, not so much.  I do have a BK precesion 20 MHz Arb which performs well enough, but it annoys me when I go to press a button, and the whole unit moves, because it is so light.

[nctnico]

I'd choose neither Agilent 33120A or Gw Instek AFG 2025. Take a look at the Feeltech, Uni-T or Siglent SDG1000 series generators.

[TheSteve]

Never forget the 33120A has an isolated output - you won't find that very useful feature on any lower end AWG(and even many higher end ones).

[bdunham7]

Never forget the 33120A has an isolated output - you won't find that very useful feature on any lower end AWG(and even many higher end ones).

True, but it is only low-voltage isolated and that can be fairly closely approximated for the lower frequencies with a slightly modified Feeltech generator or  (gasp!) an isolation transformer to float the AWG.  I keep an FY6600 around for this very reason.  A battery-powered hand-held like the UNI-T works as well.

[mawyatt]

Like the old HP and Tek stuff also, but the newer AWG offerings have the advantage of employing the newest DDS & DAC chips from AD and TI. These chips offer great performance and functionality at a moderate cost.

We selected the Siglent 2042X and it has very good output waveform fidelity due to the 16 bit DACs and output chain, UI isn't the best but manageable and I'll take the output fidelity over UI and why we selected such. You can also "expand" the 2042X to 120MHz output with software upgrades outlines by the great folks here on EEVblog. If you are using a Siglent DSO then this interfaces with the DSO to give the nice Bode plot feature which is quite useful if you are into analogish type stuff!!

Saelig here in US is Siglent distributor and has EEVblog discount.

Anyway, hope you find something you like, they are all very good options

Best,

[bdunham7]

(I haven't used any of Rigol and siglent, if I'm not wrong they sound like bit blown big by some fans here, I could be wrong but I sensed a pattern of continuous recommendations to those two brands. I'd keep an open mind about this)

It's a bang-for-buck issue.  You stated that you have a budget of $400.  So compare the specs of an SDG1032X (or SDG1062X which is a free upgrade if you hack it) to your other options.  Dual output, direct entry of harmonics, much higher sample rate and bit depth--these are not trivial or subjective issues.  I'm not as familiar with the Rigol models, but the B4$ is likely to be similar.  Even the 'cheap toy' FeelTech FY6900 will outperform your other choices in many ways. 

[2N3055]

Gw Instek AFG 2025 doesn't even have a sweep or any modulations.  Skip that one.
400 $ for it is 10x times too much. For 50$ it would be OK.

Agilent 33120A is old. Apart from isolated ground outputs, it has nothing going for it. It does have basic modulations and sweep but very basic instrument. That said, ground isolation is not fully floating certified for high voltage. But it is useful to break ground loops.. That is a plus, but only one. And it is a single channel instrument.
If you could get it very cheap go for it. Cheap here means much less than what you can get new in this class of price.

Rigol 800 series and Siglent 1000X series generators can be had new, with warranty, for less than 400 $.
For 2 independent channel outputs (that can be synchronized). For much higher frequency ranges, features etc..
Just download datasheets for Rigol and Siglent in that price range and read.

You will get why people speak enthusiastically about them. And all that even without hacking some mention here.
With Rigols, you can copy data from Rigol scopes. With Siglent, you get nice FRA, and by using PC also waveform copy/edit.

They are lot of instruments for the money.

[tautech]

With Rigols, you can copy data from Rigol scopes. With Siglent, you get nice FRA, and by using PC also waveform copy/edit.

Most brands offer simple waveform replication from scope captures via a USB stick waveform capture or as you correctly state for Siglents via the free EasyWaveX SW utility via USB or LAN connection.

For ease of compatibility AWG choice is often driven by the brand of scope already in the lab.

[2N3055]

With Rigols, you can copy data from Rigol scopes. With Siglent, you get nice FRA, and by using PC also waveform copy/edit.

Most brands offer simple waveform replication from scope captures via a USB stick waveform capture or as you correctly state for Siglents via the free EasyWaveX SW utility via USB or LAN connection.

For ease of compatibility AWG choice is often driven by the brand of scope already in the lab.

I personally prefer scope -> PC software -> Edit -> AWG  type of workflow.
It allows you to filter and fine tune captured data for AWG.

That is useful when you need to simulate sensors and input signals. You capture original signal and use AWG to simulate. It allows you to change frequency, filter, and noise ...  Useful stuff.

[nctnico]

For ease of compatibility AWG choice is often driven by the brand of scope already in the lab.

Only if the AWG can't deal with other formats. My Tektronix AWG accepts the formats of many brands of oscilloscopes without problems. No need to mess with crappy PC software.

[TheSteve]

I'd also suggest if you did get a 33120A aim for spending in the $150 region for it.

[2N3055]

For ease of compatibility AWG choice is often driven by the brand of scope already in the lab.

Only if the AWG can't deal with other formats. My Tektronix AWG accepts the formats of many brands of oscilloscopes without problems. No need to mess with crappy PC software.

PC software allows you to filter with arbitrary filters, convolute with other waveform shape, use waveform math, edit parts, normalize, combine....
All without saving anything to any sticks and moving it to other instrument..

[nctnico]

For ease of compatibility AWG choice is often driven by the brand of scope already in the lab.

Only if the AWG can't deal with other formats. My Tektronix AWG accepts the formats of many brands of oscilloscopes without problems. No need to mess with crappy PC software.

PC software allows you to filter with arbitrary filters, convolute with other waveform shape, use waveform math, edit parts, normalize, combine....
All without saving anything to any sticks and moving it to other instrument..

But that is beyond the usually crappy/limited software that comes with an AWG for free. Bottom line: in the end you still need an AWG which isn't particular about the data you feed it.

[2N3055]

For ease of compatibility AWG choice is often driven by the brand of scope already in the lab.

Only if the AWG can't deal with other formats. My Tektronix AWG accepts the formats of many brands of oscilloscopes without problems. No need to mess with crappy PC software.

PC software allows you to filter with arbitrary filters, convolute with other waveform shape, use waveform math, edit parts, normalize, combine....
All without saving anything to any sticks and moving it to other instrument..

But that is beyond the usually crappy/limited software that comes with an AWG for free. Bottom line: in the end you still need an AWG which isn't particular about the data you feed it.

While I think it could be easier to use, EasyWaveX from Siglent does all that... And more, like math equation waveforms, etc etc. You can also open data files in csv format.
I combine that with my SDG6000X and make all kinds of interesting stuff..
And if you have Siglent scope, you can pull data directly from the scope from inside of application, edit it and download to AWG directly. You can also pull data from AWG directly back into software, edit and return it.


It is not as versatile as being able to open directly datafiles from other manufacturers directly, but CSV files from one manufacturer can be easily converted to needed format. Script to do so is not too hard to make. For one off files, editing by hand in Notepad++ is not a problem.

If I could choose, if AWG would read a specific file type directly, Matlab data file format would be interesting. Because it would allow to make or process something in Matlab/Octave/Scilab and load it directly.

[Bud]

Think about resell. Nobody will want your Instek 2 years from now.

[2N3055]

Think about resell. Nobody will want your Instek 2 years from now.

Nobody wants that Instek now. It doesn't even have sweep and AM modulation. It has 4k AWG.

[eeviking]

Gw Instek AFG 2025 is only 10bit resolution.
Siglent 1000X is 14bit and 2000X is 16bit.

If you pair a Siglent generator with a new Siglent scope you can also do bode plot.

[rvalente]

I'd go with a FY6800 or 6900 for low budget or a DG811 from Rigol, which is hackable to DG922

70MHz Function Generator, 2 Channel, 250MSa/sec, 16Bit Resolution, 16M Memory

70 MHz function / arbitrary waveform generator
16 Mpts memory depth per channel for arbitrary waveforms
Dual-channel performance is equivalent to two independent signal sources
High frequency stability: ±1 ppm; low phase noise: -105 dBc/Hz
Built-in high-order harmonic generator (at most 8-order harmonics)
Built-in 7 digits/s, 240 MHz bandwidth full featured frequency counter
160 built-in arbitrary waveforms, covering common signals in engineering, medical and automotive electronics, math processing, and more
Sample rate up to 250 MSa/s, vertical resolution 16 bits
Arbitrary waveform sequence editing function available; arbitrary waveforms also can be generated through PC software
Various analog and digital modulation functions: AM, FM, PM, ASK, FSK, PSK, and PWM