Spectrum Analyzer - Rigol DSA815

[Rufus]

The first bug you've described I can confirm. The analyzer is dead during the 1000s and as you described accurately after power-on it comes back to this state.

The second bug I cannot not reproduce. I'm also using firmware 1.05.

Thanks for that. I asked Batronix to pass on the TG version of the bug to Rigol Germany. They have seen the report but couldn't reproduce it. They contacted me directly and I suggested they tried the auto then zero span version because it is simpler to describe. They couldn't reproduce that either and at the moment they are saying my analyzer must be faulty and send it back for repair.

I don't believe it is faulty. I'm sure when there is a permanent mixer overload the sweeping part of the firmware sends many requests to the GUI part to beep and display the "Intermediate frequency out of range" warning and the GUI locks up trying to handle them or something like that is going on.

Please try the zero span thing. Make sure the centre frequency is on your signal and reduce the attenuator till you get beeps and IF out of range warnings as it sweeps past the centre.  Strangely if you turn on the TG the beeping and warnings stop - which is another bug. Regardless of the TG being on or off if you now set zero span it will stop updating the trace and the UI will completely lock up until you remove the signal.

[olsenn]

Can somebody please post the 1.05 firmware on this site?

[olsenn]

Dave, are you going to review/teardown your demo DSA815? I'd love to see what its insides look like! Cheers.

[SeanB]

He said he cannot break those cal seals ( even though Mike has given a demo on how) and show it. A request from the distributor and a pretty reasonable one. He can find bugs and break it because of that though

[RCMR]

He said he cannot break those cal seals ( even though Mike has given a demo on how) and show it. A request from the distributor and a pretty reasonable one. He can find bugs and break it because of that though

Shame -- I'm in the market for a new SA and the Rigol was on my list -- but I won't pull the trigger yet -- I want to see what's inside first.

[EEVblog]

He said he cannot break those cal seals ( even though Mike has given a demo on how) and show it. A request from the distributor and a pretty reasonable one. He can find bugs and break it because of that though

I never said I can't break the cal seals?
Yes, I intend to teardown and eventually review these units.
Although Mike has already done great teardown on the function gen, so is there a point of me doing that one?

Dave.

[kg4arn]

My vote is to omit repeating the tear down of the function generator.

I'd rather the time spent on demonstrating the functionality of the Fgen and how clean its output is.

[KTP]

My vote is to omit repeating the tear down of the function generator.

I'd rather the time spent on demonstrating the functionality of the Fgen and how clean its output is.

100% agree.  Seeing the inside is great and all, but there is only so much of "it's heavy" that potential buyers find useful.

I want to see how well it works.  Jitter, slew rate, interface bugs...

[olsenn]

Any updates Dave?

[Dread]

Well, here's the response from Rigol on the DSA815:

See attached pdf from Rigol.

Looks pretty decent to me - what do the rest of you think?

Slow....

[Lukas]

Photos of the inside: https://www.mikrocontroller.net/topic/261382#2828679

[olsenn]

Photos of the inside: https://www.mikrocontroller.net/topic/261382#2828679

Thanks for the pictures!

As far as sweep speed is concearned, I do not find it to be too bad at all. Granted, I have no high-end SA's to compare to, but there are certainly no better alternatives in the same price range. If I need to see a precise frequency responce I will lower the RBW, and if I need a fast changing signal, I will increase it.  I can usually get all the information I need with a sweep time of (much) less than a second.

[Dread]

Photos of the inside: https://www.mikrocontroller.net/topic/261382#2828679

Thanks for the pictures!

As far as sweep speed is concearned, I do not find it to be too bad at all. Granted, I have no high-end SA's to compare to, but there are certainly no better alternatives in the same price range. If I need to see a precise frequency responce I will lower the RBW, and if I need a fast changing signal, I will increase it.  I can usually get all the information I need with a sweep time of (much) less than a second.

At least not Brand New but on Ebay you can find some deals on older stuff thats faster.

[David]

I'm considering trying to purchase this at work for some EMC pre-compliance testing. Anyone actually used it for testing conducted emissions etc?

[videobruce]

Seems this has died out.
Any new owners or reviews??

[RCMR]

My finger is poised over the "buy" button but I'd still like to find a local source -- rather than direct-importing.

[saturation]

Most other reviews I've run across have been good, if not high praise, and these are not from eevblog [ which I think are still the best as its most technical and critical, see archives.]  The cost of a used quality eBay SA, plus tracking generator and high accuracy frequency counter, all of questionable quality after receipt, will set you back about 75% of the cost of a new Rigol, but without the uncertainty of status.  For countries with a Rigol branch office to avail of warranty issues [ ~ none so far] its a far better deal; for those without a Rigol office nearby, the issues of warranty would have to be caveat emptor versus the big names.




http://tmworld.com/electronics-blogs/the-emc-blog/4389791/First-Impressions--Rigol-DSA815TG-Spectrum-Analyzer

There are multiple video reviews from this user:

RIGOL DSA-815 RF Spectrum Analyzer Frequency Accuracy - Revisited AGAIN!


http://forums.qrz.com/showthread.php?353552-Spectrum-analyser-resolution-bandwidth-for-amateur-radio-usage/page2

[videobruce]

saturation;
I did see that review though it wasn't as complete as I would of liked.
There were a couple members here that either had one or were going to but one. I wonder if they could come forward and post their experiences?
In another forum in a thread that was closed (no idea why) there were a few comments about this scope, but the member who had the most input has his PM's turned off. The main issue to him was phase noise.

I downloaded the Rigol specs in one of their glossy brochures and it does look like the phase noise is going to be quite poor. In one of their videos it shows -87dBc/Hz at 10kHz offset at the lower end of the analyser range.
 If you are not worried about close to carrier noise performance then I guess this isn't too much of a limitation but it is probably the noisiest analyser I have ever experienced in this respect. It's so noisy I wonder if the analyser uses the same LO/mix strategy as the higher frequency versions.
 In terms of signal sensitivity, the data suggests to me that the analyser noise figure with 0dB attenuation is about 40dB and this is about 15dB deafer than many bench analysers.
 The input IP3 appears to be +12dBm with 0dB attenuation and this initially looks good but the NF is high so the 3rd order IMD dynamic range will be several dB worse than a half decent 'regular' analyser.
 It does have a preamp to get the NF down to about 20dB but I would expect the input IP3 will suffer a lot with this switched in. Another plot shows the phase noise to be about -105dBc/Hz at 100kHz offset. Good enough for most testing but markedly inferior to a regular analyser.
 It would be interesting to see how the IP3 holds up on narrow spans as the glossy brochure shows this test at a 10MHz span. I suspect they chose this wide span to hide the fact the phase noise will spoil the appearance on narrow plots. eg an IMD test on a 20kHz span would look very noisy.

Anyone know what he means by "IP3"??

[videobruce]

My finger is poised over the "buy" button but I'd still like to find a local source -- rather than direct-importing.

I'm in the US and the importer is in a Cleveland Ohio suburb which is only a couple hundred miles away.
You don't state where you are.

[saturation]

hi videobruce,

815 is the entry level DSA of their series, and if you compare just Rigol's offering, incremental improvement of the specs increases the cost of the SA fairly greatly.  So, the more you spend the quieter it be, and the more the amps have dynamic range.  So, the question is, is the 815 adequate for your needs?

On IP3:

http://eetimes.com/design/microwave-rf-design/4018951/Understanding-RF-power-amplifiers

I see the third order intercept given on the Rigol users guide TOI this is aka intercept point 3, IP3.  IP2 = SHI on the manual

http://en.wikipedia.org/wiki/Third-order_intercept_point

Note the difference between the harmonic calculation versus the IM product calculation.

saturation;
I did see that review though it wasn't as complete as I would of liked.
There were a couple members here that either had one or were going to but one. I wonder if they could come forward and post their experiences?
In another forum in a thread that was closed (no idea why) there were a few comments about this scope, but the member who had the most input has his PM's turned off. The main issue to him was phase noise....Anyone know what he means by "IP3"??

[G0HZU]

saturation;
I did see that review though it wasn't as complete as I would of liked.
There were a couple members here that either had one or were going to but one. I wonder if they could come forward and post their experiences?
In another forum in a thread that was closed (no idea why) there were a few comments about this scope, but the member who had the most input has his PM's turned off. The main issue to him was phase noise.

I downloaded the Rigol specs in one of their glossy brochures and it does look like the phase noise is going to be quite poor. In one of their videos it shows -87dBc/Hz at 10kHz offset at the lower end of the analyser range.
 If you are not worried about close to carrier noise performance then I guess this isn't too much of a limitation but it is probably the noisiest analyser I have ever experienced in this respect. It's so noisy I wonder if the analyser uses the same LO/mix strategy as the higher frequency versions.
 In terms of signal sensitivity, the data suggests to me that the analyser noise figure with 0dB attenuation is about 40dB and this is about 15dB deafer than many bench analysers.
 The input IP3 appears to be +12dBm with 0dB attenuation and this initially looks good but the NF is high so the 3rd order IMD dynamic range will be several dB worse than a half decent 'regular' analyser.
 It does have a preamp to get the NF down to about 20dB but I would expect the input IP3 will suffer a lot with this switched in. Another plot shows the phase noise to be about -105dBc/Hz at 100kHz offset. Good enough for most testing but markedly inferior to a regular analyser.
 It would be interesting to see how the IP3 holds up on narrow spans as the glossy brochure shows this test at a 10MHz span. I suspect they chose this wide span to hide the fact the phase noise will spoil the appearance on narrow plots. eg an IMD test on a 20kHz span would look very noisy.

Anyone know what he means by "IP3"??

Hi That was my post above

I haven't seen a decent technical review of this analyser yet which is kind of frustrating. My background is that I'm an RF designer and for the last 20+ years I've always had access to the very best spectrum analysers at my place of work.

The products we design require the very best analysers available and so I'm quite clued up on how to evaluate a spectrum analyser in terms of its basic performance requirements. But sadly none of the guys these get sent to for review seem to do these basic tests. So I'm only left with the manufacturer's datasheet.

However, I guess any regular RF tests would show that the Rigol has some very real limitations in terms of spurious free dynamic range and  maybe this isn't a fair comparison on an instrument that only costs $1500 new. Quite a remarkable price!

The areas I would be most concerned about would be the spurious free dynamic range (above DANL) on a basic IP2 and IP3 test across various RBW and attenuator/preamp settings.

Also I'd be slightly concerned about internal spurious performance and rejection of external spurious and the quality of the RF attenuators at the front end.

Obviously it won't offer lab grade performance in these areas but for most home users and field technicians this analyser would probably be a better choice than an older HP analyser. Few people will really need or appreciate the difference in performance against a high end analyser and it does have the advantage of a modern user interface and is highly portable and presumably has a reasonable warranty period

[videobruce]

I already have a Aeroflex 2399a (made by LG Innotek, originally IFR). I was looking for a somewhat portable model. Below are specs from their data sheet for comparison. The three included options are shown at the bottom;

Tuning Range: 9 kHz to 3 GHz
Resolution: 1 Hz
Frequency Span Width: 100 Hz/div to 300 MHz/div in 1, 2, 5 step selections (auto-selected)
Span Accuracy: ±3% of indicated span width
Readout Accuracy: ±(span accuracy + frequency standard accuracy + 50% of RBW)
Stability Residual FM: <100 Hz p-p at 1 kHz RBW, 1 kHz VBW, (p-p in 20 ms)
Noise Sidebands: -90 dBc/Hz at 10 kHz, -98 dBc/Hz at 100 kHz both offsets measured at 2.9 GHz

Measurement Range: +30 dBm to -110 dBm
DANL:.........................................Amp Out (typ) Amp In (typ)
50 kHz to 100 kHz -95 dBm amp out, -105 dBm -117 dBm
100 kHz to 3 GHz -105 dBm amp out, -115 dBm -127 dBm
300 Hz RBW, 10 Hz VBW
1 dB Compression Point -10 dBm minimum 100 kHz to 3 GHz at 0 dB attenuation
Displayed Range:
100 dB in 10 dB/div log scale
50 dB in 5 dB/div log scale
20 dB in 2 dB/div log scale
10 dB in 1 dB/div log scale
10 divisions with linear amplitude scale

Resolution bandwidth;
300 Hz, 1 kHz, 3 kHz, 10 kHz, 30 kHz, 100 kHz, 300 kHz, 1 MHz, 3MHz + 10 Hz, 30 Hz, 100 Hz (option 05)
Accuracy: ±20%
Selectivity: 60 dB/3 dB ratio <15:1
except 3 MHz filter 50 dB/3 dB ratio <15:1
60 dB/6 dB ratio <12:1 for 9 kHz and 120 kHz Quasi-Peak filters
Switching Error: ±1.0 dB
Video Selection: 10 Hz to 1 MHz in 1-3-10 sequence plus full BW

Sweep Rate (full screen): 20 ms to 1000 s in 1-2-5 sequence, 25 µs to 20 s in Zero Span
Sweep Rate Accuracy: 20% for <100 ms, 10% for all other sweep rates

Freq stability;
Frequency: 10 MHz
Output Level: +5 dBm nominal
Temperature Stability: ±2 ppm
Aging Rate: ±1 ppm/year

Low noise preamp (option)
DANL: Improves DANL by approximately 12 dB above 50 MHz
Frequency Range: 50 kHz to 50 MHz -115 dBm typical
DANL: 100 kHz to 1.8 GHz -130 dBm typical
1.8 GHz to 3 GHz -129 dBm typical
Gain: Compensated internally

Tracking generator (option);
Frequency Range: 100 kHz to 3 GHz
Output Level: 0 to -70 dBm
Output Level Resolution: 0.1 dB step
Absolute Level Accuracy: ?±1.0 dB at 0 dB
Frequency Flatness: ?2.0 dB at -10 dBm
Signal Purity;
Harmonics <-15 dBc
Non-harmonics <-25 dBc
Sub-harmonics <-25 dBc
Leakage: <-90 dBm

Digital bandwidth filter (option);
Bandwidths: 100 Hz, 30 Hz, 10 Hz
Bandwidth accuracy: ±20%
Selectivity (-60 dB/-3 dB): <5:1
Maximum span: 1 MHz
Sweep times for 10 kHz span;
RBW:  100 Hz <0.9 sec.; 30 Hz <3 sec.; 10 Hz <4.5 sec.
Displayed Average Noise Levels (DANL) between 1 MHz and 2.4 GHz
reduces DANL by typically 5 dB from the values in the 300 Hz resolution bandwidth filter.

[videobruce]

So, the more you spend the quieter it be, and the more the amps have dynamic range.  So, the question is, is the 815 adequate for your needs?

I don't know w/o a hands on test since I don't have numbers to compare properly.

IP2 = SHI on the manual

Is that suppose to mean it isn't good? If so, what should I be looking for?

I haven't seen a decent technical review of this analyser yet which is kind of frustrating.

Dito. One review was Ok, the other useless.

The areas I would be most concerned about would be the spurious free dynamic range (above DANL) on a basic IP2 and IP3 test across various RBW and attenuator/preamp settings.
Also I'd be slightly concerned about internal spurious performance and rejection of external spurious and the quality of the RF attenuators at the front end.

Good luck finding that. 

[G0HZU]

The spec performance you have posted up for the Aeroflex shows an analyser that I would class as a semi portable technician's analyser.

i.e. the RF performance isn't great but it's probably typical of many middle of the road analysers.

I suspect the Rigol will give similar performance although I'd like to think the Aeroflex will score better on the more subtle spurious performance tests.

I'd imagine most users will use the Rigol with the preamp switched in and maybe 10dB front end attenuation. The preamp will probably degrade the frequency response below 1MHz but I guess few people will be bothered by this.

In this configuration it probably performs in a similar fashion to many mainstream analysers in terms of DANL for a given resolution bandwidth.

[videobruce]

There are only two places to get these in the US (for the most part), Rigol in Ohio, or T Equipment in NJ.
T Equip. has a 30 day money back guarantee.

One downside is only four markers. Another is a software option for VSWR readout; $440.