Siglent scope comparison

[garymck]

Had a look at Trio's web site and see the new 1202X-E for $562+GST, but then see the SDS1104CFL  for $495+GST

so 1202X-E  2 channels 200mhz 1 GSa/s
vs SDS1104CFL 4 channels 100mhz  2 GSa/s

My understanding (as a newbie) is that the 200mhz is not really useable due to the sample rate, whereas the 100mz with higher sample rate - even if split amongst 4 channels would be no worse, and probaby better than the Rigol 1054z?

Wouldn't, at that price, make the SDS1104CFL a great competitor for the Rigol 1054z? Are there any features that the Rigol at $579+GSThas that makes it a better buy than the SDS1104CFL
cheers
Gary

[ataradov]

It is hard to say for all locations. In the US, I see the cheapest price for SDS1104CFL is $800-$880, and the cheaper places don't have it in stock. And general availability seems to be sparse.

1054z is readily available from a ton of places for $400. And if you want to shop around, you can get it as cheap as $350.

[tautech]

Had a look at Trio's web site and see the new 1202X-E for $562+GST, but then see the SDS1104CFL  for $495+GST

Bit cheaper that I can do X-E due to AU-US$ has a better x rate.
The CFL might be old stock or clearance. 

so 1202X-E  2 channels 200mhz 1 GSa/s
vs SDS1104CFL 4 channels 100mhz  2 GSa/s

My understanding (as a newbie) is that the 200mhz is not really useable due to the sample rate, whereas the 100mz with higher sample rate - even if split amongst 4 channels would be no worse, and probaby better than the Rigol 1054z?

Only with a signal at max BW will Nyquist be challenged and unless you have a special need case for 200 MHz you're very unlikely to have any problems.

Wouldn't, at that price, make the SDS1104CFL a great competitor for the Rigol 1054z? Are there any features that the Rigol at $579+GSThas that makes it a better buy than the SDS1104CFL

Plenty except the BW which doesn't need to be hacked.

The X-E is a good little scope if 2 channels can do it for you, no need to hack anything.
If you must have 4 channels and don't want to hack there's not any other options close to this price level......yet. 

[garymck]

[/quote]

The X-E is a good little scope if 2 channels can do it for you, no need to hack anything.
If you must have 4 channels and don't want to hack there's not any other options close to this price level......yet. 
[/quote]

I can wait :-)

Still just new at electronics and trying to find out as much as ai can whilst I save for the gear I want...

cheers
Gary

[DaJMasta]

1GS/s is plenty for 200MHz bandwidth, you need over 2x the frequency of the signal you're measuring samples per second.  There is some variation where a 4 channel scope will switch to half the sample rate when using 3 or 4 channels, and most will, but any of these should be fine for the bandwidth you're looking at.  You can even oversample for high resolution modes with this kind of sampling rate.


The big disadvantage of the SDS1104CFL over the 1054Z is the sample memory - 24k points vs. the 1054Z's 12M points.  Big difference.  I also believe the waveform updates per second on the 1054Z are much faster - though they are unspecified on the SDS1104CFL.

In terms of general performance and usage, the 1054Z is much closer in featureset to the 1102X or 1202X-E, and the SDS1104CFL is more comparable to previous generation entry level digital scopes.  Some of the more comprehensive math features, intensity grading, deep memory, and faster waveform update rate don't always translate well to on-paper specs, but go a long way towards the responsive feel and overall useful featureset of the newer scopes.

[abraxa]

1GS/s is plenty for 200MHz bandwidth, you need over 2x the frequency of the signal you're measuring samples per second.

Sampling at 5x the maximum bandwidth may seem excessive but that's actually necessary. Imagine measuring a 100MHz sine wave on a 100MHz scope with a 200MS/s capture rate. Leaving the system bandwidth issue aside (-3db and such), you capture two equidistant points of the sine wave at every cycle. Now how do you know whether you're measuring at the peaks, at the zero crossings or somewhere in between? Could be anywhere, so the amplitude of the sine wave the scope shows could be completely off from what it actually is. But as you always capture at the same point of the cycle, you'll never know. For that reason, having a higher sample rate makes sense, even if Nyquist says it's not strictly required.

[rstofer]

The SDS1202X-E has 1 GHz rate and two channels so 500 MSPS each and at 200 MHz that is 2.5 sample rate.

The DS1054Z has 1 GHz rate and 4 channels so 250 MSPS each and at 100 MHz that is 2.5 sample rate.

They're the same!  Both sample faster than the Nyquist rate and the 2.5x has been an industry standard for a very long time.  Of course, if you aren't using all the channels, the sample rate could be higher.

Furthermore, when invoking the name of Nyquist (or Shannon), it is important to include the other fact:  The input signal must be bandwidth limited.  If you don't want aliasing, your signal can not contain components at higher than 2x the sample frequency.

Yes, theoretically, the sample rate and waveform frequency could get locked such that sampling doesn't happen quite right.  In theory...  In practice, the 2.5x rate has been working for a very long time.

All sampling scopes share the same theoretical issue.  What if?  If you want to avoid this, buy an analog scope.

[Fungus]

The big disadvantage of the SDS1104CFL over the 1054Z is the sample memory - 24k points vs. the 1054Z's 12M points.  Big difference.

??

'24M' is a lot more than '24k'.

(plus it's trivial to hack a 1054Z to get 24M)

[Fungus]

1GS/s is plenty for 200MHz bandwidth, you need over 2x the frequency of the signal you're measuring samples per second.

Sampling at 5x the maximum bandwidth may seem excessive but that's actually necessary. Imagine measuring a 100MHz sine wave on a 100MHz scope with a 200MS/s capture rate. Leaving the system bandwidth issue aside (-3db and such), you capture two equidistant points of the sine wave at every cycle. Now how do you know whether you're measuring at the peaks, at the zero crossings or somewhere in between?

Mathematically: You only need "over 2x" to capture a sine wave.

(and many people will argue that you need more in real life because signals and 'scopes are imperfect with their harmonic distortions, 8-bit ADCs, etc - 10x is the figure usually given).

The SDS1202X-E has 1 GHz rate and two channels so 500 MSPS each and at 200 MHz that is 2.5 sample rate.

The DS1054Z has 1 GHz rate and 4 channels so 250 MSPS each and at 100 MHz that is 2.5 sample rate.

That's if you turn on all channels. If you only use one channel then you get the full 1GHz sample rate on that channel.

[nctnico]

1GS/s is plenty for 200MHz bandwidth, you need over 2x the frequency of the signal you're measuring samples per second.

Sampling at 5x the maximum bandwidth may seem excessive but that's actually necessary. Imagine measuring a 100MHz sine wave on a 100MHz scope with a 200MS/s capture rate. Leaving the system bandwidth issue aside (-3db and such), you capture two equidistant points of the sine wave at every cycle. Now how do you know whether you're measuring at the peaks, at the zero crossings or somewhere in between? Could be anywhere, so the amplitude of the sine wave the scope shows could be completely off from what it actually is. But as you always capture at the same point of the cycle, you'll never know. For that reason, having a higher sample rate makes sense, even if Nyquist says it's not strictly required.

Wrong! You are forgetting that oscilloscopes use signal reconstruction which works with samplerates which are 2.5 times (or more) the frequency you are measuring. sin x/x reconstruction is using the actual sample points so it can't make up things which aren't there. Unfortunately Rigol seems to have messed up sin x/x reconstruction on their DS1000Z series so that is an exception to the rule.
When it comes to Siglent scopes: there are other choices out there with less bugs for a similar amount of money. GW Instek GDS1000B series or the MicSig TO1000 series for example.

[Electro Detective]

SDS1104CFL is good value at that price (in Australia), but there seems to be an ongoing FIRMWARE FIASCO that isn't talked about much,
or quickly swept under the bench.

And from what I've seen and read, is the Rigol any better with their firmware?

If they're crippling        these cheaper scopes with half @$$ed firmware to frustrate cash strapped students, hobbyists and working to the bone repair techs to buy better models or go the big brands, it's not going to happen.
The internet will quickly crush that money grubbing concept. 

[boggis the cat]

GW Instek are also a good value brand.  I was favouring their low-end four channel 'scope over Rigol's -- not 'hackable', but works properly (to put it bluntly) -- but decided to go with the new Siglent as I'm not likely to need more than two channels often (if at all).

If you are just starting out then you may be able to pick up a 'scope second hand for a lot less than buying one new, so you should consider that.

[StillTrying]

In practice I find the 2Ch CML+'s 40k/20k sample memory enough, - which is lucky because there's no control over the sample rate or memory depth.

At mid TB speeds it chooses 500 or 625 memory samples per div. per waveform, which gives about X0.5 X compression and about X50 X expansion on stopped or single capture waveforms.

YMMV.

I'm totally confused by how the SDS1104CFL uses its '24k memory' from here:
https://www.siglent.eu/sds1104cfl-100mhz-4-channel.html

[Fungus]

And from what I've seen and read, is the Rigol any better with their firmware?

Rigol has had plenty of firmware updates, everything is pretty much fixed now.

If you need 4 channels then an unlocked Rigol is still King at the low end.

If you need serial decoders, etc., then the Instek is out of the picture. It doesn't do that. It also comes off worse in many respects compered to the Rigol. I'd only go for it if you need better FFTs and aren't a microcontroller sort of person.

Then again, if you're not a microcontroller sort of person the 2-channel Siglent soundly beats the Instek for FFTs, bandwidth, etc. It's still a bit new on the market though, let's hope they're responsive with updates.

[Electro Detective]

The deal breaker for me is the shared channel controls on the Rigol 1054z and no independent trigger input, 
so I went the Siglent CFL which has vertical and gain above each channel etc and independent trigger input. 
You get full use of triggered 4 channels in any combination instead of wishing it could be done during a fiddly set up. 

So far the Siglent has performed ok for the simple 'noob DSO owner' tests I performed, and because of the independent channel controls like on my analogue oscilloscopes, the learning curve was a matter of minutes.. tbh figuring out how the Run and One Shot capture set up worked.     

I hope Siglent sort out the ongoing negative vibe FIRMWARE FIASCO real soon, for everyone's sake and to concrete (more) confidence in the product.
Perhaps they can ask their PC motherboard manufacturer neighbours for a few tips, who have been releasing useful, working and properly documented Firmware/Bios updates for years without fail.   

My 6.2   

[nctnico]

I hope Siglent sort out the ongoing negative vibe FIRMWARE FIASCO real soon, for everyone's sake and to concrete (more) confidence in the product.

Did you search this forum for other Siglent equipment? If you did then you'd know it takes at least years before firmware gets to a relatively bug free level if the piece of equipment is popular.

[garymck]

Hi,
thanks so much for all your comments. I'm actually wondering whether the low end scopes will do what I need. I have basically eliminated the Rigol because of its poor fft capability-saw this in a review comparo. Does the hacked Rigol perform any better in this area?

So in Australia the Rigol is around Aus$600, if I increased my budget to Aus$1000, are there any better options? To expand, my interests are in 3 areas: Audio, Micro controllers, and Radio Astronomy (the latter a new interest - my hobby has been CCD imaging in my home observatory for years...). I've been watching local secondhand sources, but nothing decent has come up.
cheers
Gary

[StillTrying]

The deal breaker for me is the shared channel controls on the Rigol 1054z

I've always thought the Rigol should have had 2 Y controls, Channels 1 & 2 switchable to 3 & 4.

[Electro Detective]

I hope Siglent sort out the ongoing negative vibe FIRMWARE FIASCO real soon, for everyone's sake and to concrete (more) confidence in the product.

Did you search this forum for other Siglent equipment? If you did then you'd know it takes at least years before firmware gets to a relatively bug free level if the piece of equipment is popular.

Is that implying most of their product will take years before it's bug free?   

Why would I waste time searching the forum for other Siglent equipment? My only concern is what I own now, and guess what, the CFL has been out for a few years already,

and so has EEVblog, with members reporting those bugs giving the manufacturers a free lunch,

just like I'm reporting the obvious ongoing  FIRMWARE FIASCO going on that needs to be resolved first,
so that frustrated customer$ have half a prayer of finding the correct firmware file, that might make the DSO bug free,
or less buggy (useable and tolerable) which is the reality in most cases. 

[Electro Detective]

The deal breaker for me is the shared channel controls on the Rigol 1054z

I've always thought the Rigol should have had 2 Y controls, Channels 1 & 2 switchable to 3 & 4.

Agreed, there's a few end user inconveniences and a couple of knobs thrown in they could have sorted out easily and at no extra cost (well maybe US $1.50 extra) 

I don't favour it's layout, it's counter intuitive for me if doing a 3 channel hook up (sacrificing ch 4 as a trigger  )  the bench a multiple probe circus, and having to stuff about with Rigol ergonomix,
and don't get me started on the extra features 'time trial' which is a total TEASE as it tick-tock times out whether you have trialled that feature or not, because it's based on the scopes overall ON time.
How hard is it to program individual trial periods into DSO firmware? it's old school trialware programming that PC and Mac software author vendors have had for like ...2 decades?   

So it's a no thanks on the crackable freebies Rigol for my humble purposes as a pseudo 4 channel,
the Siglent CFL is what it is, but gives you 4 proper wysiwyDial channels AND separate BNC trigger input,
so a real 4 channel show can go on..

 

[nctnico]

I hope Siglent sort out the ongoing negative vibe FIRMWARE FIASCO real soon, for everyone's sake and to concrete (more) confidence in the product.

Did you search this forum for other Siglent equipment? If you did then you'd know it takes at least years before firmware gets to a relatively bug free level if the piece of equipment is popular.

Is that implying most of their product will take years before it's bug free?   

Why would I waste time searching the forum for other Siglent equipment? My only concern is what I own now, and guess what, the CFL has been out for a few years already, and so has EEVblog, with members reporting those bugs giving the manufacturers a free lunch,

You would not have wasted your time because you would have learned Siglent uses a hit&run strategy. First luring in people buying cheap crap and then delaying (or even abandoning) firmware development leaving the customers with half baked products. I have given up on Siglent (and Rigol too BTW). I have learned the hard way it is better to spend a bit more money and get a product which actually works as it should. If you can return your scope then do so because it is utterly foolish to wait and see if Siglent ever fixes it. In the long run it is better to return half baked products because that is the most clear signal you can give to a manufacturer.

[ebclr]

If you wanna have a decent instrument, open source all your firmware, I'm sure community will solve all the bugs and make your product better and more marketable, let's the people who know how to do software do that, and you will profit for your big line of cheap workmanship producing the hardware.

Let's the pro's do that.

It's a nice idea, don't ignore this idea, at least evaluate the pro and cons

[ebastler]

I don't favour it's layout, it's counter intuitive for me if doing a 3 channel hook up (sacrificing ch 4 as a trigger  )  the bench a multiple probe circus, and having to stuff about with Rigol ergonomix,
and don't get me started on the extra features 'time trial' which is a total TEASE as it tick-tock times out whether you have trialled that feature or not, because it's based on the scopes overall ON time.
How hard is it to program individual trial periods into DSO firmware? it's old school trialware programming that PC and Mac software author vendors have had for like ...2 decades?   

So it's a no thanks on the crackable freebies Rigol for my humble purposes as a pseudo 4 channel,
the Siglent CFL is what it is, but gives you 4 proper wysiwyDial channels AND separate BNC trigger input,
so a real 4 channel show can go on..

No offense intended, but you seem to hold rather strong opinions based on rather limited experience...

Have you actually used the Rigol, and tried out the shared Y control? You might find that it's a very acceptable compromise, in return for the compact form factor and the additional features which the Siglent CFL lacks. Have you done any actual four-channel measurements? I have found that I am more likely to benefit from the advanced trigger options offered by the Rigol, than to be in desperate need for a separate trigger input on top of the four channels.  And I'm afraid that, during a four-channel measurement, you will find your bench to be a "multiple probe circus" with the Siglent too... 

Anyway, I guess my point is: Relax! These dog fights over oscilloscope preferences get old really quick. Enjoy your new scope!

[DaJMasta]

Hi,
thanks so much for all your comments. I'm actually wondering whether the low end scopes will do what I need. I have basically eliminated the Rigol because of its poor fft capability-saw this in a review comparo. Does the hacked Rigol perform any better in this area?

So in Australia the Rigol is around Aus$600, if I increased my budget to Aus$1000, are there any better options? To expand, my interests are in 3 areas: Audio, Micro controllers, and Radio Astronomy (the latter a new interest - my hobby has been CCD imaging in my home observatory for years...). I've been watching local secondhand sources, but nothing decent has come up.
cheers
Gary

It's ultimately your choice in the end, but I'm not sure poor FFT performance on a scope should necessarily be a deal breaker, at least for the applications you're mentioning.  If you want an audio FFT, then the 8 bit converters (or swanky 12 bit high res converters) in scopes are never going to give you as much detail as a cheap USB sound card with a 16 or 24 bit ADC on its input channel, and there is free software available that will blow away almost any scope FFT performance because it has access to the main CPU and RAM of a PC.  The limitation on using an audio ADC is bandwidth, of course, but if you can't hear beyond 20kHz or so, you don't need megasamples per second to see what's going in with great detail.

On the radio astronomy side, the problem is actually the same, but the solution is different.  Not so much to get resolution in your bins over a wide span, but to get the extremely low noise performance that you need to pick up things very far away, an 8-12 bit converter simply doesn't have the dynamic range to see quiet signals over its noise floor.  If you're looking at something transmitting in the scope's bandwidth limit and at a very high power (terrestrial sources, basically), then you could theoretically see it... but the low noise and high dynamic range of a proper spectrum analyzer is really what you'd be after.

As for the micro controller part, getting SPI or I2C decoding for a free is a very nice bonus, and it could be useful to have more than 2 channels (like with SPI), though two channels can be good for some serial decoding work and are fine for most basic analog bits interfacing with the micro or for monitoring noise on power rails or what have you.


And from my experience, owning two rigol scopes and a siglent and currently using a rigol scope and a siglent function gen..... the prevalence of software bugs are being very, very exaggerated.  I don't mean to dismiss them entirely, as I have run into a couple of minor ones over the years, they are usually relatively minor issues with parts of the instrument that aren't as commonly used... and they do get updated.  Even budget scopes will be fairly current for several years (take a look at the 1054Z, for example, which has been recommended for years and still is), and they will get half a dozen or a dozen updates over that lifetime.  Sure, if you buy the month it comes out and the first update to fix a glitch takes a few months you won't be happy, but if they've been on the market for a bit, any huge, usage-breaking bugs will not be present from any major manufacturer's firmware... and just reading up on them should give you a good idea of what the problem actually amounts to.  Especially for cheaper scopes from major manufacturers, there's no shortage of people with experience with them available online.

Bottom line is that probably many of the scopes mentioned will probably be great tools to add to your workspace.  Maybe find a couple make-or-break features to stick with, or play around with the UIs to find your preference (I actually liked my Siglent 1102X's UI a bit better than my current Rigol DS4024, but the hardware is hardly comparable), or just try to find a good deal.

[garymck]

Hi,

thank you for a really useful and thoughtful post.
much appreciated, as are all earlier posts from people who have replied to my query.

I also just came across this FFT program for Rigols:

http://hackaday.com/2015/09/22/a-better-spectrum-analyzer-for-your-rigol-scope/

which may be helpful.

cheers
Gary

Hi,
thanks so much for all your comments. I'm actually wondering whether the low end scopes will do what I need. I have basically eliminated the Rigol because of its poor fft capability-saw this in a review comparo. Does the hacked Rigol perform any better in this area?

So in Australia the Rigol is around Aus$600, if I increased my budget to Aus$1000, are there any better options? To expand, my interests are in 3 areas: Audio, Micro controllers, and Radio Astronomy (the latter a new interest - my hobby has been CCD imaging in my home observatory for years...). I've been watching local secondhand sources, but nothing decent has come up.
cheers
Gary

It's ultimately your choice in the end, but I'm not sure poor FFT performance on a scope should necessarily be a deal breaker, at least for the applications you're mentioning.  If you want an audio FFT, then the 8 bit converters (or swanky 12 bit high res converters) in scopes are never going to give you as much detail as a cheap USB sound card with a 16 or 24 bit ADC on its input channel, and there is free software available that will blow away almost any scope FFT performance because it has access to the main CPU and RAM of a PC.  The limitation on using an audio ADC is bandwidth, of course, but if you can't hear beyond 20kHz or so, you don't need megasamples per second to see what's going in with great detail.

On the radio astronomy side, the problem is actually the same, but the solution is different.  Not so much to get resolution in your bins over a wide span, but to get the extremely low noise performance that you need to pick up things very far away, an 8-12 bit converter simply doesn't have the dynamic range to see quiet signals over its noise floor.  If you're looking at something transmitting in the scope's bandwidth limit and at a very high power (terrestrial sources, basically), then you could theoretically see it... but the low noise and high dynamic range of a proper spectrum analyzer is really what you'd be after.

As for the micro controller part, getting SPI or I2C decoding for a free is a very nice bonus, and it could be useful to have more than 2 channels (like with SPI), though two channels can be good for some serial decoding work and are fine for most basic analog bits interfacing with the micro or for monitoring noise on power rails or what have you.


And from my experience, owning two rigol scopes and a siglent and currently using a rigol scope and a siglent function gen..... the prevalence of software bugs are being very, very exaggerated.  I don't mean to dismiss them entirely, as I have run into a couple of minor ones over the years, they are usually relatively minor issues with parts of the instrument that aren't as commonly used... and they do get updated.  Even budget scopes will be fairly current for several years (take a look at the 1054Z, for example, which has been recommended for years and still is), and they will get half a dozen or a dozen updates over that lifetime.  Sure, if you buy the month it comes out and the first update to fix a glitch takes a few months you won't be happy, but if they've been on the market for a bit, any huge, usage-breaking bugs will not be present from any major manufacturer's firmware... and just reading up on them should give you a good idea of what the problem actually amounts to.  Especially for cheaper scopes from major manufacturers, there's no shortage of people with experience with them available online.

Bottom line is that probably many of the scopes mentioned will probably be great tools to add to your workspace.  Maybe find a couple make-or-break features to stick with, or play around with the UIs to find your preference (I actually liked my Siglent 1102X's UI a bit better than my current Rigol DS4024, but the hardware is hardly comparable), or just try to find a good deal.