SDRPlay vs Airspy R2 Test Results
- Thread starter SCPD
- Start date
- Status
smason
Member
Cool! Thanks for posting.
Hi edmscan,
tnx!
Are you sure this list is suitable&ready for worldwide publication in the RadioReference forum?
Did you have a specific reason for the used gain settings in your investigation?
They seem a bit strange for reference measurement.
I would f.i. never engage a receiver AGC in a lab situation for sensitivity measurements, as it logically alters the gain of different amplifier stages, thus calibration of measurements.
Have you taken dynamic range or blocking into consideration?
Or will you follow up on this?
What brand is this "V2.3-b2274" calibrated test console? I'm not familiar with it.
73!
Paul
tnx!
Are you sure this list is suitable&ready for worldwide publication in the RadioReference forum?
Did you have a specific reason for the used gain settings in your investigation?
They seem a bit strange for reference measurement.
I would f.i. never engage a receiver AGC in a lab situation for sensitivity measurements, as it logically alters the gain of different amplifier stages, thus calibration of measurements.
Have you taken dynamic range or blocking into consideration?
Or will you follow up on this?
What brand is this "V2.3-b2274" calibrated test console? I'm not familiar with it.
73!
Paul
Last edited:
It is for people that want to make an informed decision before purchase.
However .. more importantly, you want to buy the unit that works the best with the software you want to use. These things are only as good as the software. I found that out very early on.
The v2.3-b2274 is the version of SDRConsole software that was used.
IMD was tested and the SDRPlay performed better than the Airspy R2. See the 2nd tab.
Yes I am well aware of the issues with these 2 models .. and expect this thread to be locked eventually due to this. But .. this has been posted for its potential benefit for purchasers.
This may help you .. PDF format of the data most likely to be of interest.
Interesting.
What was your intention with the Airspy gain settings? You turned up the mixer gain to max, and used AGC for the other gain stages. Isn't that a bit odd for SINAD measurements?
And what was your exact method for determining the (single) IMD value?
Did you enable the AGC for the same gain stages during this measurement?
Will you add essential dynamic behaviour measurements at a later stage?
Tnx
73
Paul
Last edited:
I am not going to answer why (or likely any other) questions .. as it will turn into a big fight about how they were tested and if that favors one over the other.
What is optimal for you is not likely optimal for me due to differences in the software. That is why experimentation in regards to settings is often very necessary with these devices.
People can take out of it what they like and decide from there. To be fair the testing criteria was listed so there was no mystery.
In the interest of not having this thread locked soon .. I will not be commenting further to any questions.
Last edited:
Don't worry
Accurate measurements generally leave very little room for interpretation.
At first sight, yours just seem to contain quite a few loose ends, hence my questions.
I mainly asked *why* you chose these gain settings, you must have had a specific intention with this as part of your study?
Rob Sherwood, ARRL and Adam Farson f.i. also publish measurements.
They do this signed and not anonymously, as they feel quite responsible for the information they provide.
They are also open for debate if errors somehow slip into measurements. That is how science works and evolves.
Occasionaly this leads to parts of their work being revised for the sake of optimal "reference level" accuracy.
This forum is called Radio*Reference*.
It would be quite interesting if you would explain your measurements a bit further to make them as complete and accurate as possible.
73!
Paul
Last edited:
I knew the experts were going to tear apart the testing methods...
That's why I haven't replied. But, to not answer questions about why certain methods were chosen implies either lack of knowledge or intent contrary to an unbiased test.
This comes mere weeks after you couldn't even get an SDR running on your PC, and now you expect us to believe you know the ideal test setups?
There are some obvious legitimate questions on the setup. You say the ideal setup varies from person to person. That is true. But you have to admit the setup also affects the results of testing as well. You also took time to test various frequencies in the sensitivity test, but only tested one in the IMD test. Why? (I know - you're not going to answer any questions, but from a basic test standpoint that looks very odd.)
That's why I haven't replied. But, to not answer questions about why certain methods were chosen implies either lack of knowledge or intent contrary to an unbiased test.
This comes mere weeks after you couldn't even get an SDR running on your PC, and now you expect us to believe you know the ideal test setups?
There are some obvious legitimate questions on the setup. You say the ideal setup varies from person to person. That is true. But you have to admit the setup also affects the results of testing as well. You also took time to test various frequencies in the sensitivity test, but only tested one in the IMD test. Why? (I know - you're not going to answer any questions, but from a basic test standpoint that looks very odd.)
I expected this as well .. as I know there would likely be questions as to how the test was done.
So to clarify .. I did not do the test.
It was done by one of my close friends. I am just posting the results .. but he did explain at length the testing and said that the settings chosen were used to maximize the sensitivity of the devices.
The settings DID NOT change throughout the testing including the IMD test.
They was an attempt to do a fair and impartial test of unit A vs unit B. Nothing more, nothing less.
That is not to say that there are not better settings particularly on HF.
We are mostly VHF/UHF people .. and HF was not tested as that does not interest us. Besides .. the Airspy does not go there by default.
I am only answering this as Voyager .. you are relentless. :wink::wink:
VE6RHS
Member
Testing, testing, and testing SDRs...
The BIGGEST problem with testing SDR receivers is consistency. With the many versions of software & hardware available, results of any given SDR hardware/software combination will vary wildly.
As long as the tests posted here were performed CONSISTENTLY (i.e. keeping all parameters the same throughout), the tests will in fact be a good comparison (apples to apples).
12dB SINAD is an industry accepted standard, and measuring the SINAD from the output of a sound card is a brilliant method to maintain consistency when performing tests such as these.
This gets me thinking that maybe it's time to pull out the Service Monitor and start testing some of these units on my own!
So far, after playing with several varieties of SDR "dongles", I've found that the hardware is only as good as the software. The best hardware in the world is useless without a good software "front end".
To date, I haven't found the "perfect" SDR software (front-end). We still have a long road ahead.....
The BIGGEST problem with testing SDR receivers is consistency. With the many versions of software & hardware available, results of any given SDR hardware/software combination will vary wildly.
As long as the tests posted here were performed CONSISTENTLY (i.e. keeping all parameters the same throughout), the tests will in fact be a good comparison (apples to apples).
12dB SINAD is an industry accepted standard, and measuring the SINAD from the output of a sound card is a brilliant method to maintain consistency when performing tests such as these.
This gets me thinking that maybe it's time to pull out the Service Monitor and start testing some of these units on my own!
So far, after playing with several varieties of SDR "dongles", I've found that the hardware is only as good as the software. The best hardware in the world is useless without a good software "front end".
To date, I haven't found the "perfect" SDR software (front-end). We still have a long road ahead.....
Even if the settings did not change, that does not mean they were optimal for each unit. And I didn't mean to imply the settings changed mid-test.
I can't say I know the best settings, either, but I know there are some folks around who would know. Some of them have posted above.
One post is "relentless"?
Let's look at the bright side, the bashing part hasn't started yet...
(and I'm not bashing you - I only questioned things not realizing you didn't have the answers, as it was not your test)
@ RHS: (the tester???) I agree 12 dB SINAD is a standard. But lots can be done that affect SINAD prior to the audio output stage. If brand A has a more pure output, the test settings can be distorted to make Brand B look better. Note that I said A and B as it does not matter which one A and B are. Not saying this was done. Just saying it's possible, so test setup questions are valid.
I can't say I know the best settings, either, but I know there are some folks around who would know. Some of them have posted above.
One post is "relentless"?
Let's look at the bright side, the bashing part hasn't started yet...
(and I'm not bashing you - I only questioned things not realizing you didn't have the answers, as it was not your test)
@ RHS: (the tester???) I agree 12 dB SINAD is a standard. But lots can be done that affect SINAD prior to the audio output stage. If brand A has a more pure output, the test settings can be distorted to make Brand B look better. Note that I said A and B as it does not matter which one A and B are. Not saying this was done. Just saying it's possible, so test setup questions are valid.
Last edited:
Ah, your friend did measurements and you just published them.
That explains!
Still it would be very interesting for this forum if you could ask about your friend's measuring methods.
As far as dynamic range goes, Leif Asbrink, SM5BSZ measured the dynamic range for several SDR HW.
Dynamic range is at least as important for RX performance as raw sensitivity figures.
http://www.rtl-sdr.com/comparison-several-sdrs-degradation-broadcast-fm-frequencies/
Especially the method of measuring in a consistant way throughout various devices shows his deep knowledge on the subject. It is essential to keep as many individual settings and software parameters out of the equation to come to a scientific approach.
Check out Leif's website www.sm5bsz.com
This man is a living SDR encyclopedia.
That explains!
Still it would be very interesting for this forum if you could ask about your friend's measuring methods.
As far as dynamic range goes, Leif Asbrink, SM5BSZ measured the dynamic range for several SDR HW.
Dynamic range is at least as important for RX performance as raw sensitivity figures.
http://www.rtl-sdr.com/comparison-several-sdrs-degradation-broadcast-fm-frequencies/
Especially the method of measuring in a consistant way throughout various devices shows his deep knowledge on the subject. It is essential to keep as many individual settings and software parameters out of the equation to come to a scientific approach.
Check out Leif's website www.sm5bsz.com
This man is a living SDR encyclopedia.
Last edited:
VE6RHS
Member
Anyone with experience making these measurements can tell you that the techniques used are a mine-field of error-prone procedures.
I was once asked to test a RF signal generator and was totally stumped when the signal I was measuring would never go below -112dBm (in a screened room), no matter what the attenuator setting on the generator was. It took me an hour to figure out that the signal generator had a plastic case, and was radiating enough RF through the case that the spectrum analyzer consistently displayed -112.
My point is that ANY measurements are prone to error - sometimes from factors that are well beyond the standard considerations (like using a plastic, unshielded case).
Knowing the tester in this case, I can tell you that the measurements were NOT made in a screened room, but settings/configuration consistency was maintained throughout the tests. Identical connectors/cables were used in all tests. All equipment was warmed up for at least 30 minutes before measurements. Software parameters were chosen to be similar as possible for the two units (although the AGC and gain/attenuator settings are vastly different between the units). Software settings were checked at every measurement (while waiting for the AGC/amplifiers to stabilize, all of the software parameters were double-checked).
I think I'm going to try some of these tests on other/similar dongles!
Off to pull out the Service Monitor and build a screened room.....
I was once asked to test a RF signal generator and was totally stumped when the signal I was measuring would never go below -112dBm (in a screened room), no matter what the attenuator setting on the generator was. It took me an hour to figure out that the signal generator had a plastic case, and was radiating enough RF through the case that the spectrum analyzer consistently displayed -112.
My point is that ANY measurements are prone to error - sometimes from factors that are well beyond the standard considerations (like using a plastic, unshielded case).
Knowing the tester in this case, I can tell you that the measurements were NOT made in a screened room, but settings/configuration consistency was maintained throughout the tests. Identical connectors/cables were used in all tests. All equipment was warmed up for at least 30 minutes before measurements. Software parameters were chosen to be similar as possible for the two units (although the AGC and gain/attenuator settings are vastly different between the units). Software settings were checked at every measurement (while waiting for the AGC/amplifiers to stabilize, all of the software parameters were double-checked).
I think I'm going to try some of these tests on other/similar dongles!
Off to pull out the Service Monitor and build a screened room.....
I am just interested in why use 2.5msps that's a sub optimal setting for the Airspy, it is only to be used if your computer has issues running at full 10msps.
For example if you wanted to restrict for testing to 2mhz, the optimal way would be 10msps, decimation of 4 to take full advantage of the Airspy HW.
I would love to see a test between the two SDRs but it only really makes sense if both units are configured correctly.
What I would love to see is a professional entity ie ARRL etc.. conduct an SDR review of all the models. I think this is the only way we will ever see a good review. Too bad no publication has yet decided on doing so.
If you want to be taken seriously you should have no issues answering basic questions regarding testing methodology.
What does he say about the claimed (hard to believe) 35dbm IP3 ? LOL from a 12bit RTL dongle receiver chain I suspect Youseff sometimes gets ahead of himself on some of his claims.
at best 25-26 IP3
But still, not too shabby for a 200 dollar SDR. but 35, that sounds a bit like a tall tale. This is why I would love to see some professional reviews taking a good look at these SDRs.
Where is this Bob grove review?
There is a link to it on the Airspy.com website (under reviews)
Here is a direct link: http://airspy.com/downloads/TSM_Nov2015_Airspy.pdf
Here is a direct link: http://airspy.com/downloads/TSM_Nov2015_Airspy.pdf
- Status
Similar threads
