| Products > Test Equipment |
| SIGLENT SDS1102CML - First Impressions |
| << < (12/27) > >> |
| ChildOfVision:
--- Quote from: Pippy ---You have to be in 'normal' memory mode and not 'long' memory mode to get the full 1GSamples/s. Which is a shame really. --- End quote --- --- Quote from: rf-loop ---In long memory mode it can not use 1GSa/s at all. --- End quote --- You both are right! I knew this. But I'm pretty sure scope was in short mem. mode! Look what I did: 1) Turn on the scope, check all stuff (I believe I've checked memory too) - isn't memory set to short by default btw? 2) I connected probe (not probe that comes with oscilloscope - it is really bad quality, I can't do probe compensation at all, but ok) to the BNC-A only. 3) I either press "Auto" button or turn chanel A manually - only "A" (or "I" if you want) button is illuminated! 4) When I rotate "S" dial, sample rate increases, but only up to 500MSa/a! Then I turn off scope, and repeat all this few times - all the same! After that, I "accidentally", by feel, how it usually goes in life :), decided to try ch.B ONLY (manually turned off ch.A), repetated steps 1-4 and I got 1GSa/s! Than, I can (without restart) change ch.B to ch.A and 1GSa is still there! I must note: I turned scope on/off only few times after purchase, so maybe it was long memory somehow stayed switched on (remains from factory testing)? I forgot to mention: I also have/had problems with switching from Real-time to Equ-time mode - it was impossible to do that - I got message: "Function isn't useable!" - I didn't use any "special" things, just normal, fundamental measurement of some sine waves! But, it seems to me, after, say, 10 switching-on's that both problems disappeared - I have to check this again! --- Quote from: rf-loop" ---Afaik, SDS1000CFL models (2 and 4 channel) have real time clock. SDS1000DL/CNL/CML do not have real time clock circuit (it is not FW it is hardware). --- End quote --- Well, I'm sure you know about this better than me ;)! But, I think I saw somewhere in Siglent's leaflet or datasheet something like "SDS1072/1102CML have precise 6-digit frequency measurements based on "real-time clock" (or something like that), not by counting pixels from screen" - but I'm really not sure about that! Just look at this picture (it has zoom) - SDS1102CML with clock - if there is no way (simply because there is no RTC onboard) that it has clock - than they just photoshoped image from CFL series, as you said! "False advertising" -hehe :-BROKE. http://www.siglent.com/en/product/detail.aspx?id=100000005058775&nodecode=119008001 That "clock-issue" isn't really a big thing, but I want to know do I have correct FW (or FW/HW combination). --- Quote from: rf-loop ---FW:5.01.02.22 (not just yet available from Siglent download center, I hope they update download center asap so that peoples can update new FW) --- End quote --- Where you find this version (no need to answer!!!)? Industrial espionage 8)? Just kidding! What did you get with this new FW? Do you guys know is it possible to "drag" automated-measurements "block" across the screen, I can't find that in Manual? It is especially irritatingly during FFT measurements: that "info-block" anchors itself just far left covering the FFT spectrum in really important part!? Cheers! Child |
| rf-loop:
Where there read Ch A. There read on the buttons and on the front panel CH1 and CH2 exactly and also in menu system. Real time clock (perhaps chinglish) This is internal reference clock (oscillator) what is used for two frequency counters. They are different as this what is measured from captured signal using Measure menu selections. This counter counts trigger events and it is internal reference and works like normal frequency counters and result(s) is displayed on the screen. (separate counter for both channels (of course if both channels run with separate trig!) In measurement menu, freq measurement is totally different method. Real time clock is clock what show time and date. This function is in CFL models. If you have get broken probe (manufacturing error), there is 1 year warranty. Btw, of course you use 10x setting for compensation adjustment. (1x do not have adjustment at all) Where I get scope(s) with FW 5.01.02.22 = from factory as normal for distributors.. Can you repeat this what you say about 500MSa/s limit with really known settings? If you can repeat it so that you sure know, please tell exactly how I can setup and use scope so that I can find this situation. I have now tried with 5.01.02.13 and 5.01.02.22 many many times and I have not finded this. |
| ChildOfVision:
--- Quote from: rf-loop ---Where there read Ch A. There read on the buttons and on the front panel CH1 and CH2 exactly and also in menu system. --- End quote --- I will double-check this! --- Quote from: rf-loop ---Real time clock (perhaps chinglish) /YES, it IS chinglish :D/ This is internal reference clock (oscillator) what is used for two frequency counters. They are different as this what is measured from captured signal using Measure menu selections. --- End quote --- Understood! Am I thinking right: f-measurements on the main screen is more accurate than readings from far-right menu? --- Quote from: rf-loop ---Real time clock is clock what show time and date. /I knew that/ This function is in CFL models. /Didn't know that :-[/[/color] --- End quote --- Why they didn't put RTC chip on ALL models, it is few cents costs... |O First of two probes I can't compensate, second is OK, but pretty flimsy design, It is almost twice thinner (probe, not cable) than my Testek LF-312 (official replacement for older Hameg's HM probe, afaik, 1:1/1:10, 150MHz)! I don't know is it worthwile to ask replacement... Interestingly: they have a whole spectra of probes; 60MHz AND 70MHz - if there is some difference - kill me! Maybe 60MHz and 200MHz have some difference... (I got 2x 70MHz). --- Quote from: rf-loop ---Btw, of course you use 10x setting for compensation adjustment. (1x do not have adjustment at all) --- End quote --- Note: I worked hard with analog oscilloscopes at my university (physics) for many years, I own Hameg HM-303, so I'm almost TOTAL beginner in digital "oscilloscoping", thus I know (basics, at least) about compensation (only applicable and necessary for "non 1:1" probes)... --- Quote from: rf-loop ---Can you repeat this what you say about 500MSa/s limit with really known settings? --- End quote --- YES, I will! Just, please, be patient, I'm struggling now with my PC (perhaps due to high temps here (over 40°C) :phew:)! Thanks again! Child! |
| rf-loop:
--- Quote from: Deckert on July 09, 2013, 09:34:19 pm --- --- Quote from: illyesgeza on July 09, 2013, 10:08:12 am ---Look at this two signals. The difference is longmem on and longmem off. 800khz / 20khz = 40, but the samplingrate was 500KS/s and 25MS/s 25 / .5 = 50 There is a bug in the FFT calc. --- End quote --- Agreed - I noted the same discrepancy in my review of the Siglent/Atten 'scope using the 3.x firmware. So this is an issue they did not fix in the 5.x firmware? When I tested this on the Rigol DS1102E, I noted that the Rigol simply switches back to short-mem "Normal" mode when you enter FFT mode. --deckert --- End quote --- I hope Siglent also do this: If FFT then LongMem ---> Normal It is simple and easy and does not require much. Other even better solution is if user have selected LongMemory. "Function is not useable in LongMemory mode". There is no any reason to run FFT in LongMem mode. Nothing. FFT use just 1024 sample points and thats all. |
| illyesgeza:
because the FFT calculation is too time consuming for more than 1024 points, it would be better to set a window of 1024 points on the wave in LONGMEM mode and compute the FFT on this window. in this way the user can obtain a much better rezolution in frequency domain. the only thing what Siglent has to do is to compute FFT on 1024 points from an user defined startingpoint on time domain. so no need to inhibit LONGMEM. I bet on what you want that Siglent newer will do so |
| Navigation |
| Message Index |
| Next page |
| Previous page |