Rigol RF Porn!

[olsenn]

Check out the new product Rigol is making!

http://www.rigol.com/prodserv/DSG3000/

This is in Chinese

[dr.diesel]

Link is dead already, 

[olsenn]

Link is dead already

That's my bad; you need to go to www.rigol.com, and on the upper-right side of the screen, select Chinese as the default language/region.

It is a $10k (entry model) 9kHz - 3GHz RF signal generator!

[tinhead]

It is a $10k (entry model) 9kHz - 3GHz RF signal generator!

hmm 55k RMB, that's indeed 9k USD. That will be in a year somethimg 7k USD, not bad price indeed.

[saturation]

Interesting price range, compare the competition:

http://www.home.agilent.com/en/pd-748073-pn-N9310A/rf-signal-generator-9-khz-3-ghz?&cc=US&lc=eng

In the $10k range, there's are several Agilent devices to consider.  Can Rigol compete with them head on  ?

It is a $10k (entry model) 9kHz - 3GHz RF signal generator!

hmm 55k RMB, that's indeed 9k USD. That will be in a year somethimg 7k USD, not bad price indeed.

[tinhead]

Interesting price range, compare the competition:

no, that Agilent gen is not a competition, check the SSB phase noise.

[G0HZU]

Wow, the phase noise on that Agilent generator is awful.  I looked up the datasheet and it shows a plot at 1GHz. -95dbc/Hz at 20kHz offset

It's worse than my old Marconi 2019A at 1GHz.

However, the phase noise on the Rigol is still a bit disappointing in terms of it having such a wide loop cutoff. In some areas the Rigol phase noise is excellent but it isn't very good on the 100MHz plot.  Some of the other spec points on the Rigol look extremely good. Especially the amplitude flatness and accuracy at -127dBm.

However, will it still calibrate to be that accurate and flat  in 10 years' time?

We have loads of HP and Agilent generators at work. Some of our Agilent generators really are quite poor in terms of noise floor and some have poor phase noise on the ranges below 1000MHz.
None are as bad as that N9310 though! The best generators we have at work are the PSG range and they have really low noise. But they are very expensive...

[rf-loop]

SSB Phase Noise at 20 kHz Offset

This Rigol phase noise looks terrible if we are talking good RF generators.

New Rigol
f = 100MHz <-120dBc/Hz
f =1000MHz <-108dBc/Hz
f =3000MHz <-102dBc/Hz

20 years old (just like new "pre aged") Hewlett-Packard
125 MHz: -144 dBc/Hz
200 MHz: -141 dBc/Hz
500 MHz: -137 dBc/Hz
1000 MHz: -134 dBc/Hz
2000 MHz: -125 dBc/Hz


Just for fun:

SSB phasenoise, 100kHz offset

Rigol RF generator have around -125dBc/Hz  (100MHz)

Siglent ordinary function generator datasheet tell -116dBc/Hz (typical)
(I have tested 80MHz but due to lack of accuracy for phase noise measurement I can only say that it looks something like -110dBc/Hz at 80MHz and something like  -105dBc/Hz if offset is 20kHz.)

One old HP RF generator with 100kHz offset and 1000MHz carrier,  -150dBc/Hz

After 10, 20  etc years these old HP's are working like new (or better becouse aged and drifts less), but I do not know about this Rigol how it is after 20 year.

These old gears can buy five or ten with this one new rigol price.

[olsenn]

I honestly don't see why so many people seem to think 20-year-old, rusted, large, CRT-screen, analog HP equipment is a viable substitute for brand new, ditital, compact, modern, programmable, feature packed devices! Seriously, get your head out of your ass and come to grips with the fact that phase noise isn't the ONLY thing that matters in this type of equipment.

That being said, $10k sounds expensive to me and I won't be buying one. But I suppose this kind of equipment isn't used by hobbyists. Just like the DSA815 spectrum analyzer (which I own), this device will be in a class of its own.

[grego]

I honestly don't see why so many people seem to think 20-year-old, rusted, large, CRT-screen, analog HP equipment is a viable substitute for brand new, ditital, compact, modern, programmable, feature packed devices! Seriously, get your head out of your ass and come to grips with the fact that phase noise isn't the ONLY thing that matters in this type of equipment.

That being said, $10k sounds expensive to me and I won't be buying one. But I suppose this kind of equipment isn't used by hobbyists. Just like the DSA815 spectrum analyzer (which I own), this device will be in a class of its own.

What people see is stability, reliability and repeatability.  Which is key.  New does not always imply "better".  There are advantages in the newer gear just like there are disadvantages.

The point rf-loop was making was that a 20 year old piece of equipment is more stable -- and he (rightly) pointed out that he doesn't think the Rigol will hold up over time comparatively.

[rf-loop]

I honestly don't see why so many people seem to think 20-year-old, rusted, large, CRT-screen, analog HP equipment is a viable substitute for brand new, ditital, compact, modern, programmable, feature packed devices!

Seriously:

I have one of this serie 13 year old one model from this serie in my workshop still in active use for works (rusty, oldfashion analog, CRT etc...? really?

And this ESG-A series economy class signalgen  is partially crap quality if we talk about serious high quality signal RF gen...look example 8644B or even very old 8642B..  I have both of these and no rust, no CRT-screen.. They even do not use led display becouse LED display is too noisy...  read -85 HP journal.

This 8644B can use for check / measure E series phase noise as Agilent ESG cal manual recommend.  I have one of this, 18 years old also in my workshop, for work.  Is it rusty old useless museum equipment... ?
But yes, it is 1GHz version.

And this 8642B but it also not so old. Only 26 years. Not rusty, light year better signal than this Rigol. Works like charm and is still accurate and in full working perfect condition. Also not rusty and no CRT and not even LED display becouse HP tell that leds displays produce too much noise.

(HP journal -85)
(on the front page is "heart" of 8642B)

But yes, Rigol RF gen is nice for decorate lab and perhaps better than some other cheap use and recycle gens.

[tinhead]

But yes, Rigol RF gen is nice for decorate lab

hehe, on the other side that's only 1.5k more than the Agilent crap N9310A, when one have choose between these two then Rigol is definitely better choice.

[gregariz]

I honestly don't see why so many people seem to think 20-year-old, rusted, large, CRT-screen, analog HP equipment is a viable substitute for brand new, ditital, compact, modern, programmable, feature packed devices! Seriously, get your head out of your ass and come to grips with the fact that phase noise isn't the ONLY thing that matters in this type of equipment.

That being said, $10k sounds expensive to me and I won't be buying one. But I suppose this kind of equipment isn't used by hobbyists. Just like the DSA815 spectrum analyzer (which I own), this device will be in a class of its own.

What people see is stability, reliability and repeatability.  Which is key.  New does not always imply "better".  There are advantages in the newer gear just like there are disadvantages.

The point rf-loop was making was that a 20 year old piece of equipment is more stable -- and he (rightly) pointed out that he doesn't think the Rigol will hold up over time comparatively.

Probably very little that Agilent makes today will hold up to the equipment (often Mil Spec) that they made back in the 90's in terms of mechanical construction. Agilent and the marketplace have changed. Nowadays Agilent make a range of cheap, often chinese made, gear with low but often adequate specs to compete with people like Rigol. At the high end of Agilents range you can usually still find something better than anything Rigol can make if you are prepared to pay. Otherwise try Anritsu, alot of their stuff is still designed and made here in the US.

[G0HZU]

This Rigol phase noise looks terrible if we are talking good RF generators.


And this 8642B but it also not so old. Only 26 years. Not rusty, light year better signal than this Rigol. Works like charm and is still accurate and in full working perfect condition. Also not rusty and no CRT and not even LED display becouse HP tell that leds displays produce too much noise.

Ah yes, but if you look closely at the Rigol phase noise at 1GHz at close to carrier phase noise it is way better than the 8642B.

That's why I cautiously said the Rigol's phase noise was a bit disappointing in some places but excellent in other places.

The issue I have with the 8642B is that it is very big and weighs 33kg (!)  and is fairly limited in terms of close to carrier phase noise at UHF. Eg if you wanted to use it to downmix a 1GHz signal to baseband and look at very close to carrier noise on a soundcard (eg less than 1kHz offset) then its performance is very poor. The Rigol would kill it for this test. eg it's 25dB better than the 8642B at 100Hz offset.

The Rigol is nice and small and has lots of modern connectivity and has IQ mod as an option. I'm not sure I believe the calibration accuracy claims on that web page but it is a nice generator.  It does look to be overpriced though.

However, I would recommend the Marconi 2024 as a used high end sig gen for hobby use. These can be bought for a fraction of the cost of the Rigol and despite being nearly 20yrs old can offer good performance to 2.4GHz from a small and light unit.

[KJDS]

I'l stick to using older stuff, even if it is a little larger and heavier, though I have just sold an 8664A, mostly because it was too heavy at 33kg to carry upstairs comfortably.

[1design]

If you want a decent CW source the R&S SMA100A would be a better option than any Agilent..and I tried both the E8663D and the SMA100A. But this is just my humble opinion.
Also an SSA(E5052B) measurement of the DSG3000 was made and is bang on the specifications for phase noise, so it is also a very promising option for those on a tight budget.

BR,
Miha