| Products > Test Equipment |
| Siglent SDS2000X Plus Bandwidth & Aliasing Application Note |
| << < (5/9) > >> |
| knudch:
@Performa01 Does your scope have the 500ps/div setting ? It might be you "only" have the 350MHz ;) If that the case...then 500MHz BW option only gives the 500ps/div setting ::) According to your results |
| Performa01:
Thanks for all the hints regarding the model name miracle. Given the fact, that I certainly know what I have and in this very thread of all things, where I clearly demonstrated the 580 MHz bandwidth in half channel mode, any speculations that I might "only have 350 Mhz" sound a bit strange. To sum it up one last time: * My instrument is an SDS2354X Plus. It says so on the front cover, it says so in the System Info box and it puts this model name in the file names of the screenshots. * My instrument has the 500 MHz option. It has a measured 580 MHz bandwidth, it has the 500 ps/div timebase setting and it shows no bandwidth option in the Options dialog.I have a very early version of this instrument and I got it long before its official release date. My early hardware is version 2.00 and I still have the blinking power button, yet I wouldn't expect this to have any impact on the model name, as this should be a software thing. Now I would like to leave it at that and get back on topic, which is bandwidth and aliasing. I have updated my initial posting yesterday to provide a corrected version of the document. |
| bdunham7:
--- Quote from: Performa01 on January 06, 2022, 04:31:29 am ---Given the fact, that I certainly know what I have and in this very thread of all things, where I clearly demonstrated the 580 MHz bandwidth in half channel mode, any speculations that I might "only have 350 Mhz" sound a bit strange. --- End quote --- It might seem a strange question until you consider that my scope also says SDS2354X+, does show a 500M optional upgrade and does not have the 500ps/div setting, yet still has 540MHz + BW. We didn't know you had the 500ps/div. --- Quote ---Now I would like to leave it at that and get back on topic, which is bandwidth and aliasing. --- End quote --- OK! On that topic (I think) can you tell me how the 20M, 200M and to the extent it does or doesn't exist, the 350M BW limiters are implemented? The 200M BW limiter only appears on the 2354X+ and 2504X+ models, the 350MHz BW is not manually selectable. Are these done in software or hardware and are they the exact same BW limiter that would be used to define a lower model? IOW, does a 2204X+ behave exactly like a 2354X+ with the 200M BW limit on? |
| Performa01:
--- Quote from: bdunham7 on January 06, 2022, 04:46:22 am ---It might seem a strange question until you consider that my scope also says SDS2354X+, does show a 500M optional upgrade and does not have the 500ps/div setting, yet still has 540MHz + BW. We didn't know you had the 500ps/div. --- End quote --- Ah okay, 540 MHz is not that much different from 580 MHz. Well, it all comes down to the fact that manufacturers cannot define a precise bandwidth limit in hardware, because: * The low order gaussian or Bessel filters don’t have a flat passband, so the corner frequency has to be placed high in order to get e.g. less than 2 dB amplitude drop at the specified bandwidth. * It doesn’t pay off to implement a discrete filter (and maybe an additional buffer) in the high speed signal path right before the ADC. In practice, the integrated bandwidth limiter of the PGA gets normally used for this - and this has high tolerances which must be allowed for. Consequently, the actual bandwidth limit might be at nominal 650 MHz +/-25 %. --- Quote from: bdunham7 on January 06, 2022, 04:46:22 am ---OK! On that topic (I think) can you tell me how the 20M, 200M and to the extent it does or doesn't exist, the 350M BW limiters are implemented? The 200M BW limiter only appears on the 2354X+ and 2504X+ models, the 350MHz BW is not manually selectable. Are these done in software or hardware and are they the exact same BW limiter that would be used to define a lower model? IOW, does a 2204X+ behave exactly like a 2354X+ with the 200M BW limit on? --- End quote --- Well, at least a part of this question has been dealt with in my document already, e.g. page 8: --- Quote ---We can still use the 200 MHz bandwidth limit to reduce aliasing, as shown in Figure 5 – but this cripples our DSO to a lower bandwidth instrument, increasingly unsuitable for looking at fast signals. Quite obviously we cannot have it all at the same time. Actual -3 dB corner frequency is only 185 MHz, attenuation at 500 MHz is little more than 10 dB, so the AA-effect is limited anyway. Please note that this would probably not be a suitable emulation for the genuine 200 MHz model, even when the upper bandwidth limit was a little higher, because in the 200 MHz model there might be an additional digital filter to get a uniform bandwidth limit at e.g. 240 MHz. In this case aliasing artifacts near Nyquist will see some additional suppression. --- End quote --- First of all, Siglent have indeed considered a more effective AA-filter already for the SDS2000X-E. The idea had been dropped as there just is no good compromise between bearable effort, effectiveness, and still good pulse response. I do not know about all the details. Many of them are obvious anyway. What do we have: * The manually selectable bandwidth limits are the ones in the PGA. They have a typical tolerance of 20 to 25 % (mind you, “typical” does not mean guaranteed), hence also the 185 MHz for my 200 MHz bandwidth limit. * The 350 MHz bandwidth limit should be there, but we cannot be sure because 460 MHz is so close to Nyquist and this “bandwidth limit” might just be the natural aperture error. a typical PGA might have 20 MHz, 100 MHz, 200 MHz, 350 MHz, 650 MHz and 750 MHz bandwidth limits with typically(!) 20-25 % tolerance. What to choose for 350 MHz? 350 MHz might fall into the -25 % category and users will complain if their individual scope has a measured bandwidth of only 270 MHz. So, 650 MHz it has to be. Yes, this is just an example, maybe the PGA used in the SDS2000X Plus has some different bandwidth options, but at 25 % tolerance, it would have to be at the very least 470 MHz. So, you can just as well forget about it. * An SDS1104X-E is different from the SDS1204X-E in that it has a digital 18 dB/octave filter, limiting its bandwidth to precisely 110 MHz. I do not know if lower bandwidth versions of the SDS2000X Plus have something comparable, but it doesn’t seem to be the case. |
| hpw:
Nice measurements... Well, may show the HW setup as may you simple measure a 50E chain using a given cable... As looking for overshoots in real gear (as LVHC logic) things fails soon by the used probe.. :phew: This means using a 300Mhz probe witch has a to high cap as ~20pF will draw more overshoots as using a FET/1GHz as ~1pF. Not only the BW is important, the phase shifts will cause or show the overshoots. So phase is currently neglected. Attached a Spectrum of a 10MHz clock up to 1.5GHz using as SSA juts to see that we get or have a large fourier content. |
| Navigation |
| Message Index |
| Next page |
| Previous page |