Choosing between entry-level 12-bit DSOs

[awakephd]

6% off at Saelig with their discount code would be a good start.

I have not forgotten that! That doesn't get it all the way down to $390, but it gets part of the way there. Still, I won't have the money until early June, so I will wait a bit.

[mianos]

One possible example would be SPI decoding. I haven't ever "needed" that before, but then again, I haven't used SPI that much until recently, and of course haven't ever had any way to check it before.

I have done lots of SPI, Serial and I2C tracing. The logic analysers with normal probes on oscilloscopes are very very weak compared to even the minimal dedicated logic analyser.

You can buy a Salea/SigGrok/pulseview 8 channel simple 24Mhz logic analyser for 10 bucks that completely blows away the DHO800 oscilloscope without the dedicated logic probe. I have never used an oscilloscope with a dedicated logic probe but that's not related to the topic here. I have used professional dedicated logic analysers, but that's many years ago. They were great.

I have the Rigol DHO802 myself as well.

At $100 AUD bucks you can buy the old model 16 channel Dreamsource lab non Plus versions, for $200 you can buy the newer Plus versions. These things are fine for pretty much anything I have done with the esp32 and ARM CPUs. SPI buses are nothing compared to  the ability of a Plus. The Dreamsource lab even have a dedicated click pin so 50Mhz is no worries.

TBH, I used the 8 channel one for years, it saved my heaps of time over and over. I'd still reach for it instead of the toy SPI decoder on my Rigol.

[tggzzz]

One possible example would be SPI decoding. I haven't ever "needed" that before, but then again, I haven't used SPI that much until recently, and of course haven't ever had any way to check it before.

I have done lots of SPI, Serial and I2C tracing. The logic analysers with normal probes on oscilloscopes are very very weak compared to even the minimal dedicated logic analyser.

You can buy a Salea/SigGrok/pulseview 8 channel simple 24Mhz logic analyser for 10 bucks that completely blows away the DHO800 oscilloscope without the dedicated logic probe. I have never used an oscilloscope with a dedicated logic probe but that's not related to the topic here. I have used professional dedicated logic analysers, but that's many years ago. They were great.

Just to amplify that, the reason LAs are useful is that they allow you to ignore most of the data, so you can concenrate on the information.

The simplest (but very important) example is of only capturing information when a clock changes, so you see the sequence of states rather than a sequence of samples.

More interesting examples are not starting to capture information until a specific sequence of events/states occurs. That might not be possible with very cheap devices which rely on post-processing captured data.

The same can apply to protocol analysers, at the message level rather than the bit/state level.

One possible advantage of MSOs is easier alignment of the waveforms and digital data, but with a little imagination that can be done with separate instruments.

[mianos]

One possible advantage of MSOs is easier alignment of the waveforms and digital data, but with a little imagination that can be done with separate instruments.

This may well be true but the use case is getting pretty slim here and yes I agree, it's no issue at all hooking a line to both the oscilloscope and the LA.
It's also a normal thing to use a GPIO as a trigger for either. Then you have the best of all worlds. You can decide when you want to trigger without using your brain and working out some rando sequence to get the LA to trigger on.

Something to add, I do often use an oscilloscope on a SPI or serial bus. It's often actually the first thing I do, before I reach for the logic analyser.
It's a better instrument to look at the signals, check their existence and levels. Specially with I2C and pull-ups. But once I know they look OK, I don't look at it again.

[tggzzz]

One possible advantage of MSOs is easier alignment of the waveforms and digital data, but with a little imagination that can be done with separate instruments.

This may well be true but the use case is getting pretty slim here and yes I agree, it's no issue at all hooking a line to both the oscilloscope and the LA.
It's also a normal thing to use a GPIO as a trigger for either. Then you have the best of all worlds. You can decide when you want to trigger without using your brain and working out some rando sequence to get the LA to trigger on.

Something to add, I do often use an oscilloscope on a SPI or serial bus. It's often actually the first thing I do, before I reach for the logic analyser.
It's a better instrument to look at the signals, check their existence and levels. Specially with I2C and pull-ups. But once I know they look OK, I don't look at it again.

Agreed.

I wish more people realised the sound technical benefits of (1) using an analogue time domain instrument (a scope) to ensure signal integrity, then (2) flipping to the digital time domain for digital signals (an LA, protocol analysers, printf() etc).

Yes, you can insert screws with a hammer, but it is better to use a screwdriver

[shapirus]

Yes, you can insert screws with a hammer, but it is better to use a screwdriver

At the same time it's always worth remembering that there are certain types of screws that are actually designed to be inserted with a hammer.

[Mechatrommer]

i dont know you people, but for me SPI I2C UART etc is more of software issue. when i looked quickly at oscilloscope for say SPI signal, i can know what issue in software, when software is fixed, there is no more need to look SPI on dso, so i found myself only spent very little time on such mode. different with reverse engineering other product's SPI, we need longer time with dso and probably need to save data in PC too. parallel logic otoh, MSO has its place, things like alignment, delay etc and looking at total 16CH at real time. but DHO900 only provides very basic LA function, i dont think i can trust to look at ns range delay (member reporting only 156.25MSps) and no parallel logic decoding i can see, which is very useful as well, i hope there will be upgrade for that in next FW version. but since i treat the feature as bell and whistle only (i hack dho800 to dho900) i cant hope too much. since i dont have 16CH LA (i only have cloned 8CH salae logic but i barely use) so it encourages me to clone the rigol LA probe to be used on DHO800, why not? major parts already there in FPGA and on PCB! same thing with AFG. legit salae 16CH costs a fortune, maybe like 5-10X the cost of DHO800, i saw some cloned 16CH in aliexpress, but since i dont use LA often i'm not going to buy, and how software is going to looks and functions is also suspect for northen hemisphere cloned products, if it can use salae hacked SW, then thats fine. ymmv.

[Fungus]

Yep. DSO is used to see if the pullups on the I2C bus are adequate and if the devices are responding to messages.

After that it's printf().

I've got one of those cheapo USB devices and I never use it. It's such a pain to press 'record' and wait then stop it and go looking for something useful.

I guess it might be useful if I was reverse engineering a protocol or something where you need to capture data for a long time then go into it. My Rigol has quite a lot of memory at I2C speeds, too, though and can do a full memory decode.

[Fungus]

FWIW here's my trick for easily finding the frequency of a peak in the Rigol's FFT.

The trick is that the frequency of the FFT at the center of the screen is always shown at the bottom.



Once you know that you just use your finger to move the FFT left/right and put peaks at the center. You can also pinch-zoom to expand/reduce the FFT. It's really easy.

If you want a table with numbers for printing or whatever, it has it, but... really not needed most of the time.

Just more proof that sometimes you just need to throw the old mentality out when you have a 'scope that's interactive in this way. This sort of thing isn't obvious to people who are "Choosing between entry-level 12-bit DSOs" and only looking at screenshots.

Here's me just swiping left/right on an FFT of the calibration signal, putting different peaks at the center.

[Harrow]

@awakephd  So, I promised to give you a description of my experience with the Rigol DHO800 once I received it. It's been amusing to watch the Rigol and Siglent factions battle it out here, but I come from a position of total neutrality. I've had it for about a week now and I'm really enjoying it. The last time I'd regularly used a CRO was a 2-tonne Tektronix at university in the 80's. Since then, I've only worked with voltages of 132 kV and above. I've recently set up a modest home lab and I'll mostly be using it for switch mode power supply design and some audio amplifier work (I've already managed to repair two old hi-fi amps for friends). I also wanted a scope that I could easily run off a battery to float. I have a small sealed AGM deep-cycle lead acid battery that will give me around 3 hours of runtime, so with wi-fi remote control, (which works quite well), it will let me safely do a couple of odd things that I have in mind. That was one of the deciding factors that tipped me toward the Rigol, plus the unavailability of the Siglent 800X in Australia when I placed my order.

I know this is something that all modern DSOs would do, but I didn't realise just how useful the measuring functions would be. This week I've been tuning the gate drivers for an SMPS project I'm working on and it has made life so much easier. Having Vmax, rise & fall times, duty cycle, overshoot %, etc, listed to the side of the screen is fantastic. My only gripes so far are some relatively minor things with the screen design. I did a little mock-up that I'll attach below. I'm sure that Rigol will see it and address all my points in the next firmware update.

1. Why do they waste the vertical space of such a small screen with all those grey borders (that I've highlighted in RED), I don't know? I guess it's from carrying the design over from their larger screen models. Anyway, it's not an issue in practice, but sure, it would be nice for the waveform window to have a little extra height.

2. I've edited the image to add the probe ratios. Why on earth haven't they put them there? Instead, the space is used to show the text "20 MHz" when you enable Bandwidth Limiting. I don't know how often people use that, but since it's an ON/OFF thing, it could easily be indicated by a symbol next to the channel number to make room for the probe ratios to be shown.

3. The 'Results' list on the right of the screen could show 8 results instead of 5 if they didn't waste all that vertical space. They could still keep the little graphic by moving it to the right of the text. It's nice that it lets you have items from different channels at the same time, so you could easily make good use of 8 quantities instead of 5. (Does the Siglent allow measurements from multiple channels at the same time? Defpoms review seemed to suggest not, I could be wrong, although it seems to allow you to display more items at once.)

4. I might like the waveform window to contract when the 'Result' list is displayed on the right since I have that displaying pretty much all the time, then again it is transparent and you can see through it. (I believe there is an option to toggle that on and off on the SDS800X).

5. I'm not sold on the idea of 'maths' traces being displayed in individual separate windows. It hasn't been a problem for me yet, but I'm guessing a scenario will come up soon where I'd like to overlay a 'maths' trace on the regular waveform traces, although you can stack the windows vertically with aligned time scales, so that's something at least.

Anyway, that's it. If I got to choose one item on the list to be implemented, it would be number 3, followed by number 5. My needs are pretty basic. I'm doing simple circuit design and just want to see what is happening. There are probably all sorts of nuanced differences in complex functionality between the SDS800X and DHO800 that I have no idea about. I'll need to do some bode plots soon, which I won't be able to automate, but having the ability to display calculated waveform peaks and phase differences between channel traces will make that a piece of cake anyway. The FFT looks good to me, at least at the frequencies I'm interested in. It seems fast enough and the accompanying table you can turn on is nice.

If the Siglent had been available, I might have got it instead. It seems to have been designed to make slightly better use of the screen real estate, whereas the Rigol looks designed to give 'wow' factor at first glance....engineers vs marketers and graphic designers, but I'm sure either would do the job for me just fine...we are really are spoiled for choice and value.

[electronics hobbyist]

3. The 'Results' list on the right of the screen could show 8 results instead of 5 if they didn't waste all that vertical space. They could still keep the little graphic by moving it to the right of the text. It's nice that it lets you have items from different channels at the same time, so you could easily make good use of 8 quantities instead of 5. (Does the Siglent allow measurements from multiple channels at the same time? Defpoms review seemed to suggest not, I could be wrong, although it seems to allow you to display more items at once.)

Yes, SDS800X HD supports different channel measurements, up to 5, simple measurement mode can support dozens of measurement items.

measurement review can refer to:
https://www.eevblog.com/forum/testgear/sds800x-hd-review-demonstration-thread/msg5293750/#msg5293750

[2N3055]

You don't  have experience with the device yet, just first impressions..
You call it a draw because you have low expectations and then you immediately make a list of things you don't like..
I would let some time go by before making that conclusion.

[Harrow]

You don't  have experience with the device yet, just first impressions..
You call it a draw because you have low expectations and then you immediately make a list of things you don't like..
I would let some time go by before making that conclusion.

Yep, as I said, I have simple needs and either would be just fine for me. Maybe I'll find myself constrained in the near future, or I might never have the opportunity to find out what I'm missing out on and be happy living in blissful ignorance. (And I do prefer the beige case )

[Fungus]

1. Why do they waste the vertical space of such a small screen with all those grey borders (that I've highlighted in RED), I don't know? I guess it's from carrying the design over from their larger screen models. Anyway, it's not an issue in practice, but sure, it would be nice for the waveform window to have a little extra height.

Yep. It's easy to imagine it will make a difference but it really isn't an issue.

3. The 'Results' list on the right of the screen could show 8 results instead of 5 if they didn't waste all that vertical space.

Somebody already did that in the hacking thread. It turns out a lot of the UI is controlled by XML files.

I think the hacking/extending of the DHO800 is just getting started and we can expect big things in the future.

@awakephd  So, I promised to give you a description of my experience with the Rigol DHO800 once I received it. It's been amusing to watch the Rigol and Siglent factions battle it out here, but I come from a position of total neutrality. I've had it for about a week now and I'm really enjoying it.

 

[Harrow]

3. The 'Results' list on the right of the screen could show 8 results instead of 5 if they didn't waste all that vertical space.

Somebody already did that in the hacking thread. It turns out a lot of the UI is controlled by XML files.

I think the hacking/extending of the DHO800 is just getting started and we can expect big things in the future.

Oh, wow, thanks for the heads up. If I can manage that it would really make my day. I'll have to dedicate some time to reading the hacking thread and learn XML (and Android??)...it sounds like there's a lot of fun to be had!

[Fungus]

3. The 'Results' list on the right of the screen could show 8 results instead of 5 if they didn't waste all that vertical space.

Somebody already did that in the hacking thread. It turns out a lot of the UI is controlled by XML files.

I think the hacking/extending of the DHO800 is just getting started and we can expect big things in the future.

Oh, wow, thanks for the heads up. If I can manage that it would really make my day. I'll have to dedicate some time to reading the hacking thread and learn XML (and Android??)...it sounds like there's a lot of fun to be had!

https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho800900-scope/msg5365715/#msg5365715

2. I've edited the image to add the probe ratios. Why on earth haven't they put them there?shown.

That's been done, too...

https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho800900-scope/msg5366141/#msg5366141

[Mechatrommer]

That's been done, too...

thats unofficial hack i dont think its proper to bring it up here, ymmv. the fact still stands that GUI developer is an artist, so much wasted space that can be used for technical infomations. but its minor thing for me, i will concentrate on my subject matter ignoring artistic nature of it.

[awakephd]

@awakephd  So, I promised to give you a description of my experience with the Rigol DHO800 once I received it. It's been amusing to watch the Rigol and Siglent factions battle it out here, but I come from a position of total neutrality. I've had it for about a week now and I'm really enjoying it. The last time I'd regularly used a CRO was a 2-tonne Tektronix at university in the 80's. Since then, I've only worked with voltages of 132 kV and above. I've recently set up a modest home lab and I'll mostly be using it for switch mode power supply design and some audio amplifier work (I've already managed to repair two old hi-fi amps for friends). I also wanted a scope that I could easily run off a battery to float. I have a small sealed AGM deep-cycle lead acid battery that will give me around 3 hours of runtime, so with wi-fi remote control, (which works quite well), it will let me safely do a couple of odd things that I have in mind. That was one of the deciding factors that tipped me toward the Rigol, plus the unavailability of the Siglent 800X in Australia when I placed my order.

I know this is something that all modern DSOs would do, but I didn't realise just how useful the measuring functions would be. This week I've been tuning the gate drivers for an SMPS project I'm working on and it has made life so much easier. Having Vmax, rise & fall times, duty cycle, overshoot %, etc, listed to the side of the screen is fantastic. My only gripes so far are some relatively minor things with the screen design. I did a little mock-up that I'll attach below. I'm sure that Rigol will see it and address all my points in the next firmware update.

1. Why do they waste the vertical space of such a small screen with all those grey borders (that I've highlighted in RED), I don't know? I guess it's from carrying the design over from their larger screen models. Anyway, it's not an issue in practice, but sure, it would be nice for the waveform window to have a little extra height.

2. I've edited the image to add the probe ratios. Why on earth haven't they put them there? Instead, the space is used to show the text "20 MHz" when you enable Bandwidth Limiting. I don't know how often people use that, but since it's an ON/OFF thing, it could easily be indicated by a symbol next to the channel number to make room for the probe ratios to be shown.

3. The 'Results' list on the right of the screen could show 8 results instead of 5 if they didn't waste all that vertical space. They could still keep the little graphic by moving it to the right of the text. It's nice that it lets you have items from different channels at the same time, so you could easily make good use of 8 quantities instead of 5. (Does the Siglent allow measurements from multiple channels at the same time? Defpoms review seemed to suggest not, I could be wrong, although it seems to allow you to display more items at once.)

4. I might like the waveform window to contract when the 'Result' list is displayed on the right since I have that displaying pretty much all the time, then again it is transparent and you can see through it. (I believe there is an option to toggle that on and off on the SDS800X).

5. I'm not sold on the idea of 'maths' traces being displayed in individual separate windows. It hasn't been a problem for me yet, but I'm guessing a scenario will come up soon where I'd like to overlay a 'maths' trace on the regular waveform traces, although you can stack the windows vertically with aligned time scales, so that's something at least.

Anyway, that's it. If I got to choose one item on the list to be implemented, it would be number 3, followed by number 5. My needs are pretty basic. I'm doing simple circuit design and just want to see what is happening. There are probably all sorts of nuanced differences in complex functionality between the SDS800X and DHO800 that I have no idea about. I'll need to do some bode plots soon, which I won't be able to automate, but having the ability to display calculated waveform peaks and phase differences between channel traces will make that a piece of cake anyway. The FFT looks good to me, at least at the frequencies I'm interested in. It seems fast enough and the accompanying table you can turn on is nice.

If the Siglent had been available, I might have got it instead. It seems to have been designed to make slightly better use of the screen real estate, whereas the Rigol looks designed to give 'wow' factor at first glance....engineers vs marketers and graphic designers, but I'm sure either would do the job for me just fine...we are really are spoiled for choice and value.

Harrow, thanks so much for circling back on this. Your first impressions are quite helpful, as it sounds like our interests are not dissimilar. I will admit that I don't typically work with 132kV ... I make it a strict policy to stay under 130kV.

thats unofficial hack i dont think its proper to bring it up here, ymmv. the fact still stands that GUI developer is an artist, so much wasted space that can be used for technical infomations. but its minor thing for me, i will concentrate on my subject matter ignoring artistic nature of it.

I am not sure why it would not be proper to bring up the hacks that are available? As the OP, that is one of the things I asked about early on in this thread. Obviously, different folks will have different priorities and needs, but the degree to which either of the scopes in question is hackable is definitely of interest to me. Perhaps I am not understanding what is meant by "unofficial" hack - I would have thought that any hack was, by definition, unofficial? I plead for patience, as I am no doubt further reinforcing my newbie status ...

[Fungus]

thats unofficial hack i dont think its proper to bring it up here, ymmv. the fact still stands that GUI developer is an artist, so much wasted space that can be used for technical infomations. but its minor thing for me, i will concentrate on my subject matter ignoring artistic nature of it.

I am not sure why it would not be proper to bring up the hacks that are available? As the OP, that is one of the things I asked about early on in this thread. Obviously, different folks will have different priorities and needs, but the degree to which either of the scopes in question is hackable is definitely of interest to me. Perhaps I am not understanding what is meant by "unofficial" hack - I would have thought that any hack was, by definition, unofficial? I plead for patience, as I am no doubt further reinforcing my newbie status ...

I think he means some hacks are "official" in the sense that you enable things by installing license keys (eg. bandwidth upgrade).

I don't see why other hacks are lesser hacks though.

[KungFuJosh]

I don't see why other hacks are lesser hacks though.

As you know, I hate Rigol...but in this case, the hacking sounds fun, and actually made me consider it for a second. 😉

[Mechatrommer]

thats unofficial hack i dont think its proper to bring it up here, ymmv. the fact still stands that GUI developer is an artist, so much wasted space that can be used for technical infomations. but its minor thing for me, i will concentrate on my subject matter ignoring artistic nature of it.

I am not sure why it would not be proper to bring up the hacks that are available? As the OP, that is one of the things I asked about early on in this thread. Obviously, different folks will have different priorities and needs, but the degree to which either of the scopes in question is hackable is definitely of interest to me. Perhaps I am not understanding what is meant by "unofficial" hack - I would have thought that any hack was, by definition, unofficial? I plead for patience, as I am no doubt further reinforcing my newbie status ...

I think he means some hacks are "official" in the sense that you enable things by installing license keys (eg. bandwidth upgrade).

I don't see why other hacks are lesser hacks though.

sorry i sometime mixed up english... first unofficial i meant is its not from official FW in rigol website, and secondly, that specific hack only applies to current FW being hacked/hex edited, once the FW is upgraded, its gone and the original creator of the hack has to "re-edit" the newer FW and redistribute it for others to "re-patch". other type of hack like vendor.bin hack which will still valid on newer FW without you doing anything, FWIW... anyway "unofficial" also can mean, you have to work something after you make a purchase and possibly invalidated the warranty, and some people are not comfortable doing it.

if we can include hackability in decision making, i can vote on rigol DHO800 since upgrading AFG and LA is now possible (at cheap cost), even though with charming 2GSps and faster FFT on siglent SDS800X, i think i can squeeze more functionalities out of DHO800 + downloaded data to PC for analysis. charming measurement in SDS800X such as Trend, i can just eyeball or make some simple rig circuit using arduino or something to get the same functionality, i can also learn in the process, albeit waste more time. but for some, they just want everything to be ready in scope in silver plate to save time or whatever reason, which DHO800 is lacking on few aspects compared to SDS800X... so ymmv.

just as analogy, some people like to buy coat or office dressing and hate jungle trekking dressing, because they work in office for living and afraid if even a little scratch got onto their skin. but for some other people, they swear by jungle or army tactical dressing and will put coat dress in dustbin even if someone give to them for free, because their life is in jungle and have no use to office dressing, so people come here to ask what dress to choose? i question back, what type of person are you? where do you work? if they ask what type oscilloscope to buy? i ask, do you enjoy hacking and tearing your scope apart to do "unofficial" hack? or you are type "i dont want to void warranty sticker" i just want it to work out of the box with plenty of fancy bells and whistles. both situations will give different answer. ymmv cheers.