EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: rhb on June 23, 2020, 11:19:13 pm
-
These seem to me the best options. Others are more money than I want to spend or rather old e.g. CML, etc. So I'm reluctant to spend a lot of time testing a scope that is about to hit EOL.
In the interest of getting a scope ordered soon, I've set a 10 day limit on the poll. I'll buy whichever gets the most votes.
Have Fun!
Reg
-
I put my vote in for the 1104 on the basis that it can be unlocked with a 200MHz code and is suppose to tap out at 150s or so i think it was on its front end so interested to see the end result
the 1202 i would expect to stay no better than 200 as there is no further code for them that i recall.. tv84 will correct me if im wrong
-
I voted for sds1202 (1104 is out for my budget)
if the war is for entry level oscilloscopes, 1202 is more suitable than 1104...my two cents
^-^
-
I put my vote in for the 1104 on the basis that it can be unlocked with a 200MHz code and is suppose to tap out at 150s or so i think it was on its front end so interested to see the end result
the 1202 i would expect to stay no better than 200 as there is no further code for them that i recall.. tv84 will correct me if im wrong
There was to be at least 100 and 200 MHz versions of SDS1*02X-E however when released to the west markets only the 200 MHz SDS1202X-E is offered and has a proven BW of some 225 MHz.
SDS1202X-E release heralded free decode of the common protocols since offered in all subsequent Siglent DSO's.
It might seem the 4ch X-E's just have another 2 channels tacked on but this is not the case as there's a 2nd 1 GSa/s ADC and another 14 Mpts memory support. However it doesn't just stop there, 4ch X-E's have the webserver and Bode plot features
-
I put my vote in for the 1104 on the basis that it can be unlocked with a 200MHz code and is suppose to tap out at 150s or so i think it was on its front end so interested to see the end result
the 1202 i would expect to stay no better than 200 as there is no further code for them that i recall.. tv84 will correct me if im wrong
There was to be at least 100 and 200 MHz versions of SDS1*02X-E however when released to the west markets only the 200 MHz SDS1202X-E is offered and has a proven BW of some 225 MHz.
SDS1202X-E release heralded free decode of the common protocols since offered in all subsequent Siglent DSO's.
It might seem the 4ch X-E's just have another 2 channels tacked on but this is not the case as there's a 2nd 1 GSa/s ADC and another 14 Mpts memory support. However it doesn't just stop there, 4ch X-E's have the webserver and Bode plot features
Really? thats surprising about the web server and the plotter.. i thought that was a standard feature to all of their lineup
-
I put my vote in for the 1104 on the basis that it can be unlocked with a 200MHz code and is suppose to tap out at 150s or so i think it was on its front end so interested to see the end result
the 1202 i would expect to stay no better than 200 as there is no further code for them that i recall.. tv84 will correct me if im wrong
There was to be at least 100 and 200 MHz versions of SDS1*02X-E however when released to the west markets only the 200 MHz SDS1202X-E is offered and has a proven BW of some 225 MHz.
SDS1202X-E release heralded free decode of the common protocols since offered in all subsequent Siglent DSO's.
It might seem the 4ch X-E's just have another 2 channels tacked on but this is not the case as there's a 2nd 1 GSa/s ADC and another 14 Mpts memory support. However it doesn't just stop there, 4ch X-E's have the webserver and Bode plot features
Really? thats surprising about the web server and the plotter.. i thought that was a standard feature to all of their lineup
SDS1000X-E Webpage:
Browser control/onboard webpage for software free monitoring (4 channel models only)
Bode plotting function controls external generator (SIGLENT SDG or SAG generators, 4 channel models only)
Fear not, I've been nagging them about fixing this ^ ! :horse:
The more expensive SDS2202X-E and it's 350 MHz brother has this and also the Chinese only SDS1002X-C that they should really get onto releasing to western markets.
-
I put my vote in for the 1104 on the basis that it can be unlocked with a 200MHz code and is suppose to tap out at 150s or so i think it was on its front end so interested to see the end result
the 1202 i would expect to stay no better than 200 as there is no further code for them that i recall.. tv84 will correct me if im wrong
There was to be at least 100 and 200 MHz versions of SDS1*02X-E however when released to the west markets only the 200 MHz SDS1202X-E is offered and has a proven BW of some 225 MHz.
[snip]
A 500 MSa/s channel has a 250 MHz Nyquist. So in order to provide a decent step response it should have a 150-175 MHz -3 dB corner and sold as such. If an SDS-1202X-E has a -3 dB point at 225 MHz the step response is complete crap! I don't even need to see the step response to be certain of that. I know what the Fourier transform looks like if you drop from - 3 dB at 225 MHz to - 48 dB at 250 MHz. The step response has lots of overshoot and rings for a long time.
A 5 MHz analog scope will show signals out well past 10 MHz. The aliasing issue lead to marketing guys promoting totally stupid filter profiles with time domain responses ranging from bad to utterly ridiculous.
I am thoroughly fed up with claims that a 100 MHz Rigol has a BW to 135 MHz or that a 200 MHz Siglent has a BW to 225 MHz. It is *pure* BS and demonstrates complete ignorance of the Fourier transform.
TANSTAFL!
There are *two* things that matter:
What is the time domain response
What is the sample rate and -3 dB BW needed to produce that time domain response without aliasing
*All* the rest is pure BS. Fourier proved all this 210 years ago.
Have Fun!
Reg
-
I put my vote in for the 1104 on the basis that it can be unlocked with a 200MHz code and is suppose to tap out at 150s or so i think it was on its front end so interested to see the end result
the 1202 i would expect to stay no better than 200 as there is no further code for them that i recall.. tv84 will correct me if im wrong
There was to be at least 100 and 200 MHz versions of SDS1*02X-E however when released to the west markets only the 200 MHz SDS1202X-E is offered and has a proven BW of some 225 MHz.
[snip]
A 500 MSa/s channel has a 250 MHz Nyquist. So in order to provide a decent step response it should have a 150-175 MHz -3 dB corner and sold as such. If an SDS-1202X-E has a -3 dB point at 225 MHz the step response is complete crap! I don't even need to see the step response to be certain of that. I know what the Fourier transform looks like if you drop from - 3 dB at 225 MHz to - 48 dB at 250 MHz.
Posts in this thread might prove otherwise:
https://www.eevblog.com/forum/testgear/dso-bandwidth-test-sds1104x-e-dsox1102g-to1104-gds1054b/ (https://www.eevblog.com/forum/testgear/dso-bandwidth-test-sds1104x-e-dsox1102g-to1104-gds1054b/)
-
the 1202 i would expect to stay no better than 200 as there is no further code for them that i recall.. tv84 will correct me if im wrong
For comparison sake we can always create a 100MHz license and downgrade the official 1202 to 1102.
I think Reg shouldn't be spending more money. Of course I would like to see a Siglent in the comparison (1104 is a more universal version in the forum).
-
if you put a 4 channel out there against a 2 channel, it is obvious that the choice for the 4 channel wins ... I will be the only one to have recommended the 2 channel :-[
-
if you put a 4 channel out there against a 2 channel, it is obvious that the choice for the 4 channel wins ... I will be the only one to have recommended the 2 channel :-[
It's not a comparison of the usability of 4 channels. It's the technical aspects of the HW that should be similar to the 2 channel.
-
siglent typically has one full adc conversion rail per two channels and its the front ends being compared so if dual channel you just stick to 1+2 or 3+4 and its the same as doing a two channels with only any differences in the front end hardware of the two channels and the backing adc
tv.. a 200 downgraded to a 100? :-DD i have to say that would be original.. i dont think anyone has done that one on purpose anyway lol.. i'd be curious to see the result
-
A 500 MSa/s channel has a 250 MHz Nyquist. So in order to provide a decent step response it should have a 150-175 MHz -3 dB corner and sold as such.
The two numbers you need to look at when you buy a 'scope are analog bandwidth (-3dB point on the input circuitry) and sample rate.
It's up to the user to figure out the consequences of the combination of the two.
I am thoroughly fed up with claims that a 100 MHz Rigol has a BW to 135 MHz or that a 200 MHz Siglent has a BW to 225 MHz. It is *pure* BS and demonstrates complete ignorance of the Fourier transform.
A 100/135Mhz Rigol has a sample rate of 1GHz when there's only one channel on. That's plenty to make the theory work. Same goes for two channels (one trigger and one for important signal).
Does it fall apart when you turn on all four channels? Yes, but you're allowed to turn channels off for critical measurements.
Scopes which are higher up the food chain usually have better ratios, eg. The Rigol MSO5000 has 8GHz sample rate and four channels at 350MHz. It never dips below 2GHz sample rate no matter how many channels you switch on.
There's also the trick of using dots mode and turning up the screen persistence to "reconstruct" the signal. That's what was going on here:
https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg1818215/#msg1818215 (https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg1818215/#msg1818215)
That's a whole other can of worms though. :popcorn:
-
A 500 MSa/s channel has a 250 MHz Nyquist. So in order to provide a decent step response it should have a 150-175 MHz -3 dB corner and sold as such. If an SDS-1202X-E has a -3 dB point at 225 MHz the step response is complete crap! I don't even need to see the step response to be certain of that. I know what the Fourier transform looks like if you drop from - 3 dB at 225 MHz to - 48 dB at 250 MHz. The step response has lots of overshoot and rings for a long time.
A 5 MHz analog scope will show signals out well past 10 MHz. The aliasing issue lead to marketing guys promoting totally stupid filter profiles with time domain responses ranging from bad to utterly ridiculous.
I am thoroughly fed up with claims that a 100 MHz Rigol has a BW to 135 MHz or that a 200 MHz Siglent has a BW to 225 MHz. It is *pure* BS and demonstrates complete ignorance of the Fourier transform.
Why I totally agree with you that more space should be left for filter rolling off between -3dB point and Nyquist this may not be the entire picture.
For example you seem to assume that with 500 MSa/s the filter will be the same as with twice the sampling rate. This may not be the case and when the sampling rate is dropped (because additional channel is switched on) maybe the analogue BW is reduced as well. The Siglent's single channel BW maybe just shy of 240MHz and maybe even below 200MHz when the connected other channel activated.
Note: Sampling rate reduction may happen not only just because of the connected channel activation but because such a long time base is selected that acquisition memory would not last the entire display to be filled with full sampling rate (so it's better not to say never).
Another point is that the amount of overshoot and other artifacts depend on how many harmonics fit to the rolling off side of the filter.
As we're probably not talking about a real step response but a fast rise time periodic square like signal, the base frequency matters a lot here. For example with a 10MHz base, the harmonics are spaced close enough that multiple of them will fit to the rolling off side of the filter. The consequence is that the signal does not look very nice. However what if a 100MHz base signal is used (with 1 GSa/s)? Not enough number of harmonics fit to the pass band and the signal may look nicer (but actually very sine like) than the same rise time but lower base signal (none of them show the real shape though).
Long story short: If you need only BW an older boat anchor may be enough but if you need both BW and the capability of capturing non repetitive signals you need to move up in the food chain. It's better to forget that a cheap entry level DSO will do both. No matter which of them.
-
There's also the trick of using dots mode and turning up the screen persistence to "reconstruct" the signal. That's what was going on here:
Beware though. I didn't want to get into details, but at some point on this thread I mentioned some "signal processing differences". The DS1000Z scopes don't have a true dots mode. They still interpolate.
The Principle Of Least Astonishment would dictate that in dots mode a dot should represent a sample as faithfully as possible. That seems to be the case for other manufacturers, not for Rigol or at least not for the DS1000Z.
So the dots mode is a purely cosmetic thing without any value. Quite the contrary, it can be misleading.
-
the dots mode is a purely cosmetic thing without any value.
No arguments there.
I'd vote to remove the 'feature' but it seems quite popular. There's lot's of people who claim to "see" extra information in the dots by squinting their eyes or something. I'm not one of them.
-
if you put a 4 channel out there against a 2 channel, it is obvious that the choice for the 4 channel wins ... I will be the only one to have recommended the 2 channel :-[
Old Timer here with another vote for SDS1202X-E. It's my first foray into the digital world from the analog cathode ray scope, (Hitachi v-1050F). Not too difficult to learn the ropes,
but the on-screen readings are so small that it is difficult for these old eyes to read. Two channels are all I've ever needed.
-
if you put a 4 channel out there against a 2 channel, it is obvious that the choice for the 4 channel wins ... I will be the only one to have recommended the 2 channel :-[
It depends on what you need it for. Generally speaking (which implies there are exceptions) a 2 channel oscilloscope is enough for doing repair work. Ofcourse having 4 channels is nicer.
-
I'm clearly going to have my work cut out for me dispelling a lot of mythology about time series acquired with a scope. I should like to suggest getting clear on what I posted this morning to the Scope Wars thread with regard to impulse response and passband shape. And on aliasing.
It's quite clear that several people don't understand the relationship between the impulse response and the passband. Or sample dithering. It is not magic; it's mathematics.
Have Fun!
Reg
-
I'm clearly going to have my work cut out for me dispelling a lot of mythology about time series acquired with a scope. I should like to suggest getting clear on what I posted this morning to the Scope Wars thread with regard to impulse response and passband shape. And on aliasing.
It's quite clear that several people don't understand the relationship between the impulse response and the passband. Or sample dithering. It is not magic; it's mathematics.
Have Fun!
Reg
dspics are my bread and butter... live and die by iir/fir :P math is big bucks... you are doing the lords work :clap:
-
Except for the fact that there is no money in it these days, I'd suggest producing 3D images from seismic. That is *very* difficult stuff. Even the easy codes are typically wrong.
Please look at the "Scope Wars" posts. This was just supposed to be a poll.
Have Fun!
Reg
-
Gotta work in the niche markets.. Its still out there.. especially in iot telecom for heavy industry mobile machines, CAT, John Deer, etc
-
the dots mode is a purely cosmetic thing without any value.
No arguments there.
I'd vote to remove the 'feature' but it seems quite popular. There's lot's of people who claim to "see" extra information in the dots by squinting their eyes or something. I'm not one of them.
I mean it's poorly implemented in the Rigol DS1000Z.
-
Dot mode and persistence are *very* valuable done correctly. I have scopes that do that and I have scopes that don't. I'll be showing examples of both including a LeCroy DDA-120 sampling at 8 GSa/s with a GPSDO time base reference input.
Have Fun!
Reg