I ask the forgiveness of moderators if i have placed this thread in wrong section; moreover i use the solid state section almost exclusively for all of my threads.
I would like to know if someone else (especially from my Italian neighbours) has builded and used this kit of NUOVA ELLETRONICA Italian magazine "LX.1690" and if it is worthy. For anyone that he not has idea for what i speak, i explain further the whole topic.
As for many of you my desire it is a relatively accurate PC based analyser, and for this purpose there are free shared softwares in internet. One of them which is very good (according to my personal evaluation based to my knoweledge about measuring instruments for many years) and it is the "Visual Analyser" of the named Silanum Software. Maybe the owner of it is a member in this forum. Maybe... I don't know. At this point may to clear that the sound cards are not accurate for measurements. I tried my X-Fi Platinum of Creative which says that it is of 24bit resolution, but my PC says 16bit (Intel C2D E6850, Intel MB D35JOQ, 8G DRAM, WinXP 64bit). Moreover when i inject square waves in input, the Visual Analyser shows a THD of 26% and above and this is obvious also in the oscilloscope screen where the bounches are very high and in all the duty cycle portion of square. So- for me - the use of a sound card for measurements it is only a toy. My mini disc MDS-JE780 of Sony which i bought before one year, has by far most accurate conversion (i checked this also with sq. waves which is the only signal for a true evaluation of the ADs) with a perfect by 90% square in output with only two verry small debounches. Can you imagine that this MD of 300Euro has most reliable conversion from many Hi-End class DA converters of 700Euros and above? i looked this in a comparative test in a Greek magazine before one year. And this is expected because this last generation of MDs produced by Sony, in his attempt to beat the MP3 algorithm by inserting his new ATRAC 3-S for a proportional capacity in MD incorporates in the device double 24bit AD-DA converters! Yes!
Well, in the webpage of Visual Analyser in the hardware menu, suggested a kit made from the magazine Nuova Elletronica. When i read the reference for this, i find it interesting because the two 16bit AD converters included. And the kit is sell for 95Euros only. Corresponding ready devices of Velleman or Pico of 16bit costs from 400Euro and above. Of course this kit it has a USB connection to PC. You can see also informations in the web page either of Visual analyser or of Nuova Elletronica.
For so small ammount, i have ordered already the complette kit included also a kit for calibration amd a kit of filters for THD measurements. Total 130,5Euros included also two BNC cables with crocodile clips. I am in waiting of delivery of this kit.
I am waiting if anyone has build this kit to give me informations about the quality of measurements and possible improvements.
Thanks to all
Fotios
I would like to know if someone else (especially from my Italian neighbours) has builded and used this kit of NUOVA ELLETRONICA Italian magazine "LX.1690" and if it is worthy. For anyone that he not has idea for what i speak, i explain further the whole topic.
As for many of you my desire it is a relatively accurate PC based analyser, and for this purpose there are free shared softwares in internet. One of them which is very good (according to my personal evaluation based to my knoweledge about measuring instruments for many years) and it is the "Visual Analyser" of the named Silanum Software. Maybe the owner of it is a member in this forum. Maybe... I don't know. At this point may to clear that the sound cards are not accurate for measurements. I tried my X-Fi Platinum of Creative which says that it is of 24bit resolution, but my PC says 16bit (Intel C2D E6850, Intel MB D35JOQ, 8G DRAM, WinXP 64bit). Moreover when i inject square waves in input, the Visual Analyser shows a THD of 26% and above and this is obvious also in the oscilloscope screen where the bounches are very high and in all the duty cycle portion of square. So- for me - the use of a sound card for measurements it is only a toy. My mini disc MDS-JE780 of Sony which i bought before one year, has by far most accurate conversion (i checked this also with sq. waves which is the only signal for a true evaluation of the ADs) with a perfect by 90% square in output with only two verry small debounches. Can you imagine that this MD of 300Euro has most reliable conversion from many Hi-End class DA converters of 700Euros and above? i looked this in a comparative test in a Greek magazine before one year. And this is expected because this last generation of MDs produced by Sony, in his attempt to beat the MP3 algorithm by inserting his new ATRAC 3-S for a proportional capacity in MD incorporates in the device double 24bit AD-DA converters! Yes!
Well, in the webpage of Visual Analyser in the hardware menu, suggested a kit made from the magazine Nuova Elletronica. When i read the reference for this, i find it interesting because the two 16bit AD converters included. And the kit is sell for 95Euros only. Corresponding ready devices of Velleman or Pico of 16bit costs from 400Euro and above. Of course this kit it has a USB connection to PC. You can see also informations in the web page either of Visual analyser or of Nuova Elletronica.
For so small ammount, i have ordered already the complette kit included also a kit for calibration amd a kit of filters for THD measurements. Total 130,5Euros included also two BNC cables with crocodile clips. I am in waiting of delivery of this kit.
I am waiting if anyone has build this kit to give me informations about the quality of measurements and possible improvements.
Thanks to all
Fotios
I have been looking at this exact kit just today Fotios! I downloaded the article featuring the LX.1690 but have had difficulty translating to english. I have been using Visual Analyzer for a few years now and have been limited by my soundcard capabilities the same as you.
With the exchange rate as bad as it is right now I don't see myself purchasing this kit, but I would love to hear your opinion of it when you get it running!
Good luck with the project,
David
With the exchange rate as bad as it is right now I don't see myself purchasing this kit, but I would love to hear your opinion of it when you get it running!
Good luck with the project,
David
Giaime said:
KP11520 said:
Hi Giaime
According to your suggestion, i have made a tranlation of the moderator's "MARIOVALVOLA" post who has made the concrette kit. Following it is the translation :
"I have found the tip just. I have mounted the card. It works. Accustomed to more large members, for me it is already a conquest. It does not have noise neither up neither low .bisognerebbe to shield it underneath. the positioning feels a lot. To short, I will try to make a test more detailed.
I can anticipate that it holds the 96KHz calmly is a lot to delineate (like answer in frequency). Most advised to who it does not have nothing in house. Indeed funny.
I hope of being able to make a comparison reasoned with the old one clio that I have tendentially in house (like analyzer of phantom) Null of serious. I wanted to only emphasize the noise absence…. For the price, it is a beautiful result. (the generator is external….)"
But after it follows a big post of you. I can't translate this because it is so long. Can you tell us in English (in English i am not so good, you can suppose in Italians!) what is your comment in this post with few words?
Anyway, this week i hope to have the kit in my hands. What to hell, i have ordered it from the past Thursday by Air Mail and the Greece is near. Hope to have ready and tested in Weekend and I promise to write a post with my impressions. Fortunately I have acquired recently a good generator HAMEG HM8130 of 10MHz. With this will become the tests. And with the use of Silanum Visual Analyser as much with the RMAA analyser. To be sure. In any case I believe that it will have better results from my X-Fi Platinum.
Do you believe Giaime that ths LX.1690 kit has the same effectiveness in comparisson with the Pico products of the same bit depth? That is the question!
Fotios
Dear Fotios,
to make the thing short, I believe you can get far better measurements (more resolution, lower noise floor) with pro-audio PC soundcards than from Nuova Elettronica's product. You won't get the input attenuator and all the fancy hardware (you have to build yourself), but the price is lower.
Please note that with a PC soundcard you are limited with about 100kHz input frequency (using a 192ks/s sound card), no problem because Nuova Elettronica's kit is limited with 48ks/s.
I strongly suggest getting an analog oscilloscope to do "oscilloscope" duty, and a PC soundcard to get baseband FFT (in audio frequencies).
I use an M-Audio 2496, not particulary a good soundcard but it's cheap.
to make the thing short, I believe you can get far better measurements (more resolution, lower noise floor) with pro-audio PC soundcards than from Nuova Elettronica's product. You won't get the input attenuator and all the fancy hardware (you have to build yourself), but the price is lower.
Please note that with a PC soundcard you are limited with about 100kHz input frequency (using a 192ks/s sound card), no problem because Nuova Elettronica's kit is limited with 48ks/s.
I strongly suggest getting an analog oscilloscope to do "oscilloscope" duty, and a PC soundcard to get baseband FFT (in audio frequencies).
I use an M-Audio 2496, not particulary a good soundcard but it's cheap.
Giaime said:
Dear Giaime
You may don't believe that i checked this kit with only the FFT analysers of Silanum or RMAA. All of these are toys for me. I will test the kit with square waves injected in its input via the oscilloscope portion of Visual Analyser. If it gives better results from my Creative X-Fi Platinum (that is lower TIM & THD) then i am not interested for the sampling rates and the bits. My DSO is a Hameg HM507 (50MHz) and has two 8bit flash A/D converters which offer a Random Sampling Rate of 2GSa/s or a Real Time Sampling Rate of 100MS/s. Also it has its software to connected in PC. In the software also included a FFT analyser but i must paid 80Euros to get a key from Hameg to unlock it. As you can understand, with this software i can do the job. But the setup it is very complex in comparison with Silanum. Also i am not interested for the screen resolution of Visual Analyser (for this purpose i think it is the sampling rate value) because i use for real time measurements only my DSO which has by far much better resolution with its smallest screen. The problem it is the quality of the A/D converter and this is indicated by the bounches in square waves. For example, my Hameg although it has only 8bit converters, they are of top quality. If you have ever looked for A/D converters of different constructors, you find that many of 8bit are more expensive from 16bit. All sound cards included the PRO have A/D converters for music purposes only (this means for easy processed sinusoidal form signals).
For this reason i decided to try the LX.1690 kit. I have to lost 95Euros if it is not goot responsible in the test of square waves.
And be sure that i will refer this in the magazine and i will prompt them to upgrade this project by changing the converters with such of better quality.
Fotios
I use Visual Analyzer with an EMU 0202 with fine results. IMO, slamming square waves of short rise and fall times into a digital system is more a test of anti-aliasing filters and the front end, than of any inherent goodness of the conversion system. It's not even a wise idea for analog systems, for a variety of reasons. Still, people insist on doing it, and drawing erroneous conclusions from the results. My favorite waveform generators have adjustable rise and fall times, but that's a somewhat rare feature.
Conrad Hoffman said:
Mr. Hoffman, my care it is focused only in one thing. In the front end circuit of A/D converter. There are three types of converters:
1) The parallel or flash A/D converters
2) The single or dual ramp A/D converters
3) The successive approximation A/D converters
The first type it is the most complicated but also by far the most reliable from any of other two types because includes enough comparators in its convertion bank. Such type uses almost all the good DSO. Also i believe the Hi-End SACD (and from luck my cheap MD player).
The second type it uses a ramp generatorand it is more simplest from the first type because two circuits are enough to make a ramp. But it is an unreliable type for accurate convertions and for this reason developed the dual ramp type to enforce the thing.
The third type it is a sequel for reducing the cost of first type by using the same method but instead a range of comparators (the comparators are absolutely analog circuits) it use a D/A converter! instead them and only one comparator. This type used in all of microcontrollers today, sound cards, computers etc because his low cost. You can build very easy a converter not of 16bit, not of 24bit, not of 32 bit but of 64bit! by increasing simply the sampling rate of the control circuit of the trial digital word and the storage register. The only analog circuit used moreover in this type it is a single comparator. Try to make a such 64bit parallel flash A/D converter; you may include 64 seperate analog comparators!If you can understand, the sampling rate it lost its value in such type A/D converters; it is a physical law. You can take as much sampling rate you want. My DSO as i said has two flash A/D converters of 8bit each and can achive up to 2Gsa/s. But its converters does not cause debounches in square waves (the debounches means as i said not THD but IM and TIM of inadmissible percentage (26% and above!) and this means consequently measurements for fun because IMHO the IM and TIM distortion are of biggest interesting from THD+noise. Because the two first have relation with the quality of designing from the designer of project, instead the THD+noise have relation with the quality of parts used.
These all for beging to place an end in this missinterpreting of the values of our measurement methods. I continue with more elements about this subject.
With Respect to the oppinion of all
Fotios Anagnostou
it would be best to use an oscope for any measurements using square waves. the only time i use square waves in audio testing are slew rate measurements, as the other things a lot of people use square waves for are better tested with a swept sine and spectrum analyzer or a noise source and spectrum analyzer. 96khz bandwidth is actually a good range for an audio spectrum analyzer, as most of the analog ones go typically to 100khz or 200khz (with the exception of some HP analyzers that go up to 1 or 2 Mhz). even 48Khz is usable for most audio work.
if you are using a sound card as an analyzer, the first thing you should do is verify that the card does FOR SURE go up to 96khz by setting the sampling rate to 192k and sweep an oscillator up to 96khz, paying attention at 24, and 48khz that you dont get a mirror image on your spectrum trace (which is an indication that the card isn't really sampling at 192k, but is sampling at 96k or 48k).
if you are using a sound card as an analyzer, the first thing you should do is verify that the card does FOR SURE go up to 96khz by setting the sampling rate to 192k and sweep an oscillator up to 96khz, paying attention at 24, and 48khz that you dont get a mirror image on your spectrum trace (which is an indication that the card isn't really sampling at 192k, but is sampling at 96k or 48k).
FWIW, for years I've had scopes at work with 8-bit converters and built in FFT. That FFT is pretty much worthless for all known applications. Even with a ton of averaging. We just got a scope with 10-bit converters, and FFT. That FFT is actually quite useful. Still, the worst sound card, and Visual Analyzer, seems to do a better job than scopes with high speed, but limited vertical resolution.
Conrad Hoffman said:
For this reason exactly - as i said in a previous post - i don't pay the 80 Euros in Hameg Instruments to give me the key for unlocking the FFT which is incorporated in his software offered with each DSO. From this software the only that works in PCs via a RS232 interface it is a bidirectional remote control of scope and the capturing of curves with documentation in the hard disk of computer. Nothing else.
Unfortunatelly, my sound card (what big words of creative;24bit!) a X-fi platinum by alone and in 96KHz with the internal generator of Visual Analyzer show in the FFT of Visual analyser only 0,018% THD with sine waves but if i place the notch filter for example at 1KHz with the same signal applied the THD increased at 0,2%. If i place a pink noise signal, the THD increased further at 1,5%. All of these executed internaly without the use of an external generator and thus the A/D converter of the card that means completelly digital process. I don't know, maybe the generator of Visual Analyser it is no so good.
Fotios
I bought an M-sound 2496 (USB Pro) card - I was not able to get the same card PMA uses. I get very good results as reported by RMMA (5.5 because 6 does not work with the M-Sound - be warned). however, if I try to run 24 bit 96k, I get worse measurement results (2nd and 3rd harmonic at -75db vs -90 db on 16 bit).
I think using a sound card is a bit of an art - besides, you also hav e to have a fine appreciation of fow th e filters, resolution and th e measurement software are interacting.
I agree with comments above about sticking square waves into digital systems - you ar e probably limited to low frequencies and rise times have to be considered.
Maybe PMA can jump in here and comment since I see he seems to hav e quite a lot of expereince in this area.
PMA, what's your advice here?
I think using a sound card is a bit of an art - besides, you also hav e to have a fine appreciation of fow th e filters, resolution and th e measurement software are interacting.
I agree with comments above about sticking square waves into digital systems - you ar e probably limited to low frequencies and rise times have to be considered.
Maybe PMA can jump in here and comment since I see he seems to hav e quite a lot of expereince in this area.
PMA, what's your advice here?
Bonsai said:
Then dear Bonsai, why my DSO does not suffers from such type affections? Why its response in the offensive square waves it is flat from few Hz up to its full bandwidth? A DSO is it a digital system or not?
Also my MD player MDS-JE780 of Sony why it has an acceptable behaviour in the same type of signals (with reduced bandwidth in relation with my DSO of course) with little overshoot in the corners of duty portion of the square and a very small inclination from the original shape of the rising and the falling edges of signal? Is it or not a digital system also this?
I insist in my reference on post #9 for the type of A/D converters.
And finally i accept your place that the behaviour of a sound card it is acceptable for sinusoidal waveforms. But its analog part obviously adds a signifficant amount of TIM distortion in the original signal (25% average). In a whole waveform the THD and the TIM are separated or merged in its presentation? And if they are merged then the resulted measurement how much in percent error contains?
I ordered this kit of Nuova Elettronica because its little cost. I haven't not received yet it and i have a distress if the A/D converter incorporated in the MCU. If it is such then it is of Successive Approximation type and then i can't make any improvement. If it is external then i have an opportunity to make an improvement by changing (in collaboration with the magazine) the A/D converters with flash type.
Regards
Fotios
Sincerelly i ask your forgiveness, because i write this post only AND ONLY to stay my thread in first page for obvious reasons. These are for viewing it easy anyone who has build this project already. I have an agony because i don't received yet the kit from Nuova Elettronica; and they already have been charge my card with 170 Euros for the kit and the shipping cost. Also the Greek Posts are run in the rate of a turtle! Sorry my country, it is only a pure true and nothing else.
I hope to understand my place dear moderators. WHERE IS ANATECH? Hey friend, you forgot me so fast?
Fotios
I hope to understand my place dear moderators. WHERE IS ANATECH? Hey friend, you forgot me so fast?
Fotios
Unexpected delay of Kit delivery
Today it is 3/3/2008 and when i asked Nuova Elettronica when they dispatched the kit LX.1690, they answeered me in 26/02/2008. Yet i've not received the kit. Sorry for the delayed report about the performance of it, but me i am not responsible for this. Be patient please. The worse it is that Nuova Elettronica has been charged me with 35 Euros!!! for the dispatch cost.
Fotios
Today it is 3/3/2008 and when i asked Nuova Elettronica when they dispatched the kit LX.1690, they answeered me in 26/02/2008. Yet i've not received the kit. Sorry for the delayed report about the performance of it, but me i am not responsible for this. Be patient please. The worse it is that Nuova Elettronica has been charged me with 35 Euros!!! for the dispatch cost.
Fotios
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.