Rigol MSO5074 or Siglent SDS2104X-Plus ?

[nctnico]

I think the Siglent is better than the Rigol, even with the price premium.

But as usual neither is perfect. Rigol has the better memory handling where Siglent has better features. If you want the best of both worlds you have to look elsewhere. Usually I end up at a different brand (and paying more) to have less compromises.

[Martin72]

Rigol has the better memory handling

This could be egalized in not so far future - Hope, they´ll (siglent) do it right...

I end up at a different brand (and paying more) to have less compromises.

Most here are hobbyists with limited budget.
So we´re talking mostly about brands like siglent, rigol, micsig, gwinstek...
At work, it´s a pure pleasure to work with the lecroy hdo6034a or waverunner 9054 - But the price....

[uski]

I think the Siglent is better than the Rigol, even with the price premium.

Or get a vintage Lecroy DDA-120 for around $300-$400, it is much better than the Siglent, 1GHz, 8GSa/s, 16Mpts, tons of measurements and statistics... it even has a thermal printer!!!

These DDA-120 are all $800 or so on eBay. I've given up trying to find reasonably priced second hand equipment.

[nctnico]

Rigol has the better memory handling

This could be egalized in not so far future - Hope, they´ll (siglent) do it right...

I end up at a different brand (and paying more) to have less compromises.

Most here are hobbyists with limited budget.

Well there are quite a few who are looking around for professional use OR have enough money but can't resist spending less on what is seemingly a device which offers the same features. A while ago I recommended a Siglent signal generator to a customer. This ended up with 2 people chasing a problem for over a week which turned out to be a bug in the generator. So the financial gain turned into a loss.

[Fungus]

This ended up with 2 people chasing a problem for over a week which turned out to be a bug in the generator. So the financial gain turned into a loss.

Yep. The thread about terrible bugs in the Siglent SDS1000X is as long as any other bug thread.

[TK]

I think the Siglent is better than the Rigol, even with the price premium.

Or get a vintage Lecroy DDA-120 for around $300-$400, it is much better than the Siglent, 1GHz, 8GSa/s, 16Mpts, tons of measurements and statistics... it even has a thermal printer!!!

These DDA-120 are all $800 or so on eBay. I've given up trying to find reasonably priced second hand equipment.

Yes, but look for ones that accept best offer... these are really boat anchors and I guess most sellers want to get rid of them... I got mine for $350 shipped.

[nctnico]

I think the Siglent is better than the Rigol, even with the price premium.

Or get a vintage Lecroy DDA-120 for around $300-$400, it is much better than the Siglent, 1GHz, 8GSa/s, 16Mpts, tons of measurements and statistics... it even has a thermal printer!!!

These DDA-120 are all $800 or so on eBay. I've given up trying to find reasonably priced second hand equipment.

Yes, but look for ones that accept best offer... these are really boat anchors and I guess most sellers want to get rid of them... I got mine for $350 shipped.

I agree. Asking price is the highest price you'll have to pay. Just low-ball and see where it goes. It is not like the old Lecroy scopes are rare.

[Elasia]

You could just flip a coin... they are both solid scopes in the same class but each has their own niche to evaluate.. if just a generic scope then you can flip a coin and do just as fine between the two

[Elasia]

Rigol has the better memory handling

This could be egalized in not so far future - Hope, they´ll (siglent) do it right...

I end up at a different brand (and paying more) to have less compromises.

Most here are hobbyists with limited budget.

Well there are quite a few who are looking around for professional use OR have enough money but can't resist spending less on what is seemingly a device which offers the same features. A while ago I recommended a Siglent signal generator to a customer. This ended up with 2 people chasing a problem for over a week which turned out to be a bug in the generator. So the financial gain turned into a loss.

lol.. nothing quite like 2nd guessing yourself wondering wtf then to realize you been had

[CMASupra]

Either scope would probably be fine for you as a hobbyist. Neither is a high-end scope as they are both mid-range scopes from mid-level brands. They're both able to do most everything but aren't necessarily great at anything. I personally bought the Siglent SDS2104X Plus and started using it today. My rationale was that I didn't need 8Gsa/s, but I did want deep memory that lasted a good amount of elapsed time while maintaining a decent sampling rate. Beyond that, both the SDS2104X and the MSO5104 supported everything I needed. (Note that I wasn't considering the MSO5074).

I do wish Dave would do an updated video on the Rigol MSO5000 to show whether they've fixed the issues he encountered with it. Rigol's software update changelog indicates some changes were made that potentially affect bugs Dave found.

It is worth noting that the price to upgrade your oscilloscope to 350MHz would be quite high, especially compared to the price of just buying a higher-end model to start with. Don't count on upgrading unless you're going to hack your oscilloscope. Also, if you need 350MHz bandwidth with 3 or 4 active channels, go for the Rigol. The Siglent is limited to 1GSa/s with all 4 channels on, and 1G/350M=2.86, barely above Nyquist frequency. The 8GSa/s of the Rigol (2GSa/s with all 4 channels on) will be useful at 350MHz.

[Fungus]

Yep. If you need 4 channels with full bandwidth then the Rigol's 8GSamples/sec. give you a lot more headroom.

[uski]

These comments about the number of GS/s vs the number of active channels are interesting and are something I did not consider until now.
The high noise of the analog frontend of the Rigol is still a major issue for me and, in my opinion, is a bigger problem that the reduced bandwidth with all channels due to the GS/s sharing.

In all honesty, in my use case, I do not necessarily need a super high bandwidth. Lower noise matter to me however. That's just my use case, that I share just for discussion purposes, and everyone will be different of course

I have also been sending some offers on eBay but so far all have been denied. I hope they will keep their 20 years old scopes on their shelves for a long while if they are not interested in selling 

[Martin72]

Our older Waverunner LT scopes (1999..2002) are having max. 1GSa/s and you could work with it with no problems.
This was my concerning too before changing to the sds2104x+, leaving the 8GSa/s..
But then I made my mind up, thinking about the scopes on work and yes, the noise were the killer for my rigol 5000.
And the final kill was the answer from rigol, as I´ve asked of what about the frontend of the 7000s and they wrote me, it should be nearly the same.
So that was it, I´ve changed the brand.

[nctnico]

These comments about the number of GS/s vs the number of active channels are interesting and are something I did not consider until now.
The high noise of the analog frontend of the Rigol is still a major issue for me and, in my opinion, is a bigger problem that the reduced bandwidth with all channels due to the GS/s sharing.

In all honesty, in my use case, I do not necessarily need a super high bandwidth. Lower noise matter to me however. That's just my use case, that I share just for discussion purposes, and everyone will be different of course

I can't imagine a case where noise doesn't matter. Another issue with the Rigol MSO5000 is that it seems to have only 6 bits of effective resolution even with the 20MHz bandwidth limit on (according to measurements done by HowardLong whom I trust to be right about it).

[rf-loop]

These comments about the number of GS/s vs the number of active channels are interesting and are something I did not consider until now.
The high noise of the analog frontend of the Rigol is still a major issue for me and, in my opinion, is a bigger problem that the reduced bandwidth with all channels due to the GS/s sharing.

In all honesty, in my use case, I do not necessarily need a super high bandwidth. Lower noise matter to me however. That's just my use case, that I share just for discussion purposes, and everyone will be different of course

I can't imagine a case where noise doesn't matter. Another issue with the Rigol MSO5000 is that it seems to have only 6 bits of effective resolution even with the 20MHz bandwidth limit on (according to measurements done by HowardLong whom I trust to be right about it).

6 bits?

More like barely 5 bits ( 3.65 - 5.24 depending V/div) least if look this Howardlong message and his image "effective bits"
https://www.eevblog.com/forum/testgear/benefits-of-going-with-all-siglent-setup/msg3281580/#msg3281580   there reply #16

[mawyatt]

These comments about the number of GS/s vs the number of active channels are interesting and are something I did not consider until now.
The high noise of the analog frontend of the Rigol is still a major issue for me and, in my opinion, is a bigger problem that the reduced bandwidth with all channels due to the GS/s sharing.

In all honesty, in my use case, I do not necessarily need a super high bandwidth. Lower noise matter to me however. That's just my use case, that I share just for discussion purposes, and everyone will be different of course

I can't imagine a case where noise doesn't matter. Another issue with the Rigol MSO5000 is that it seems to have only 6 bits of effective resolution even with the 20MHz bandwidth limit on (according to measurements done by HowardLong whom I trust to be right about it).

6 bits?

More like barely 5 bits ( 3.65 - 5.24 depending V/div) least if look this Howardlong message and his image "effective bits"
https://www.eevblog.com/forum/testgear/benefits-of-going-with-all-siglent-setup/msg3281580/#msg3281580   there reply #16

I made the decision to go with the Siglent based upon the Rigol's noisy front end and general comments (UI) before seeing the issue with the ENOB! Thank goodness I made the right decision!

The ADC is the most critical component in a sampled data system, which DSOs are. Even with a prefect frontend, if the ADC corrupts the measurement, the end result is compromised. Quality, high speed, good resolution ADCs are quite difficult to design and need to be fabricated in an advanced CMOS process and one could speculate that the design of this 8GSPS ADC has some serious issues. I've seen Keysight's Stingray ADC (5 years ago, actually have some) that's in the latest high speed DSOs, this was a very high performance/speed ADC that required the best engineering talent and lot's of $ to pull off in an advanced CMOS process. Like the superb Griffin DAC, these are masterpieces of integrated circuit chip design expertise and not easy to reproduce!!

Stingray (wafer) Image.

Best,

[nctnico]

These comments about the number of GS/s vs the number of active channels are interesting and are something I did not consider until now.
The high noise of the analog frontend of the Rigol is still a major issue for me and, in my opinion, is a bigger problem that the reduced bandwidth with all channels due to the GS/s sharing.

In all honesty, in my use case, I do not necessarily need a super high bandwidth. Lower noise matter to me however. That's just my use case, that I share just for discussion purposes, and everyone will be different of course

I can't imagine a case where noise doesn't matter. Another issue with the Rigol MSO5000 is that it seems to have only 6 bits of effective resolution even with the 20MHz bandwidth limit on (according to measurements done by HowardLong whom I trust to be right about it).

6 bits?

More like barely 5 bits ( 3.65 - 5.24 depending V/div) least if look this Howardlong message and his image "effective bits"
https://www.eevblog.com/forum/testgear/benefits-of-going-with-all-siglent-setup/msg3281580/#msg3281580   there reply #16

I was trying to be nice for a change..  But you are right. Saying it is 5 bits at most is a more accurate statement.

[nbritton]

I'm asking myself this same question, using the Saelig discount code and my Chase 5% cash back for PayPal transitions the MSO5072 will end up costing $811.74, vs the SDS2104X-Plus costing $1249.31, a difference of $437.57. So as a hobbyist is the Siglent that much better to justify the additional cost?

The 50 Ohm inputs on the Siglent is appealing, so is the 10-bit oversampling, the better UI, and 50 MHz AWG... but are those alone enough to make it worthwhile?

Furthermore, what's the resell value on these when I upgrade to the latest and greatest in a few years... I recall several years ago that Siglent was unlawfully blocking the sale of used equipment on eBay, are they still doing this? The Rigol has a software bundle promo right now, and strictly speaking I would have to reset the scope back to defaults if I were to sell it to someone else, so the extra software bundle license might actually have a little bit of tangible value when I decide to sell it too.

I'm a computer systems engineer, so the ease of hackability of ether scope isn't a deciding factor for me. Can the AWG on ether of these scopes be programmatically controlled, say for instance with a python script? i.g. bit banging?

And technically I already have a Logic Analyzer (Diligent Analog Discovery), and I believe ether scope can do SPI decoding using the four analog channels... so I don't see an obvious need for the expensive LA probe, at least up front. I also figure that if/when I outgrow the LA on the Analog Discovery that it might make more sense at that point to upgrade to a dedicated LA device.

My biggest goal as a hobbyist is building my own 8 or 16 bit computer from scratch so that I can better understand computer engineering at the lowest level. For example, Ben Eater's Youtube channel (https://www.youtube.com/user/eaterbc) is a perfect example of generally what I want to do. However, I also need something for more general purpose use, such as RF, analog, bode plots, FFT, does the Siglent have a spectrum analyzer mode? 

[Fungus]

So as a hobbyist is the Siglent that much better to justify the additional cost?

Only you can answer that, based on your financial situation.

The Rigol has a software bundle promo right now, and strictly speaking I would have to reset the scope back to defaults if I were to sell it to someone else, so the extra software bundle license might actually have a little bit of tangible value when I decide to sell it too.

a) How will a buyer know if you had those options legally or not?
b) The buyer might be a hacker, too.
c) I don't imagine many privately bought Rigols go unhacked so a buyer won't care.

Can the AWG on ether of these scopes be programmatically controlled, say for instance with a python script? i.g. bit banging?

Yes. All functions can be controlled.

My biggest goal as a hobbyist is building my own 8 or 16 bit computer from scratch so that I can better understand computer engineering at the lowest level. For example, Ben Eater's Youtube channel (https://www.youtube.com/user/eaterbc) is a perfect example of generally what I want to do.

The Rigol will do that just fine. Heck, a Rigol DS1054Z will do that just fine if you're looking at CPU clock speeds in the single digit MHz.

However, I also need something for more general purpose use, such as RF, analog, bode plots, FFT, does the Siglent have a spectrum analyzer mode?

You just said you had an Analog Discovery... 

[nbritton]

So as a hobbyist is the Siglent that much better to justify the additional cost?

Only you can answer that, based on your financial situation.

Can the AWG on ether of these scopes be programmatically controlled, say for instance with a python script? i.g. bit banging?

Yes. All functions can be controlled.

My biggest goal as a hobbyist is building my own 8 or 16 bit computer from scratch so that I can better understand computer engineering at the lowest level. For example, Ben Eater's Youtube channel (https://www.youtube.com/user/eaterbc) is a perfect example of generally what I want to do.

The Rigol will do that just fine. Heck, a Rigol DS1054Z will do that just fine if you're looking at CPU clock speeds in the single digit MHz.

Financially speaking, the $437.57 price difference is a drop in the bucket, but at the same time I'm generally frugal unless spending the extra money for more features creates a better value proposition / ROI.

If the AWG can be numerically controlled via software like you said, the Siglent seems to top out at 3 volts, whereas the Rigol can drive 5 volt, and given that I want to work with retro digital circuits I presume I'm going to be working with 5 volt logic... yes? 

[nbritton]

Also, it appears the Rigol is running Linux kernel version 3.12, and the Siglent is running version 3.19. I normally work on the command line via a bash SSH shell, I know the Rigol has an SSH shell but I don't know if the Siglent does. How fleshed out are the user land environments? Do they have a toolchain, can you install applications on them, for instance GCC, Python, Perl, Bash?

I'm more into mathematics and physics then I am electronics, so I like playing with FFT formulas and doing number theory stuff, so being able to program the AWG using arbitrary complex mathematical formulas on the command line would be of great interest to me.

[TK]

So as a hobbyist is the Siglent that much better to justify the additional cost?

Only you can answer that, based on your financial situation.

Can the AWG on ether of these scopes be programmatically controlled, say for instance with a python script? i.g. bit banging?

Yes. All functions can be controlled.

My biggest goal as a hobbyist is building my own 8 or 16 bit computer from scratch so that I can better understand computer engineering at the lowest level. For example, Ben Eater's Youtube channel (https://www.youtube.com/user/eaterbc) is a perfect example of generally what I want to do.

The Rigol will do that just fine. Heck, a Rigol DS1054Z will do that just fine if you're looking at CPU clock speeds in the single digit MHz.

Financially speaking, the $437.57 price difference is a drop in the bucket, but at the same time I'm generally frugal unless spending the extra money for more features creates a better value proposition / ROI.

If the AWG can be numerically controlled via software like you said, the Siglent seems to top out at 3 volts, whereas the Rigol can drive 5 volt, and given that I want to work with retro digital circuits I presume I'm going to be working with 5 volt logic... yes?

I don't think you will use the AWG for retro digital circuits, except to generate a square wave, which you can do with something costing less thatn $10.  You are better with a vintage HP logic analyzer that offers tons of inputs and you can also get one with a pattern generator to simulate vintage logic

[TK]

Also, it appears the Rigol is running Linux kernel version 3.12, and the Siglent is running version 3.19. I normally work on the command line via a bash SSH shell, I know the Rigol has an SSH shell but I don't know if the Siglent does. How fleshed out are the user land environments? Do they have a toolchain, can you install applications on them, for instance GCC, Python, Perl, Bash?

I'm more into mathematics and physics then I am electronics, so I like playing with FFT formulas and doing number theory stuff, so being able to program the AWG using arbitrary complex mathematical formulas on the command line would be of great interest to me.

It doesn't matter as user access is closed, unless hacked... even when you can enable sshd, you don't want to go inside and mess with the pre-installed software (except to do some sort of hacking...).  Remember, it is not a general use PC, it is a test equipment and better to leave the SW as untouched as possible

[tautech]

The 50 Ohm inputs on the Siglent is appealing, so is the 10-bit oversampling, the better UI, and 50 MHz AWG... but are those alone enough to make it worthwhile?

Auto probe sense is very important to some users but a deeper dig into the datasheets and manuals need be required for the full picture of differences.

Furthermore, what's the resell value on these when I upgrade to the latest and greatest in a few years... I recall several years ago that Siglent was unlawfully blocking the sale of used equipment on eBay, are they still doing this?

Oh FFS not this hardy annual again.....NO !

The Rigol has a software bundle promo right now, and strictly speaking I would have to reset the scope back to defaults if I were to sell it to someone else, so the extra software bundle license might actually have a little bit of tangible value when I decide to sell it too.

Current SDS2000X Plus Promo:
https://siglentna.com/news-article/save-up-to-1764-with-a-new-sds2000x-plus-oscilloscope-and-option-bundle/

However, I also need something for more general purpose use, such as RF, analog, bode plots, FFT, does the Siglent have a spectrum analyzer mode? 

No.
2 Mpts FFT only.

If the AWG can be numerically controlled via software like you said, the Siglent seems to top out at 3 volts, whereas the Rigol can drive 5 volt, and given that I want to work with retro digital circuits I presume I'm going to be working with 5 volt logic... yes?

+3V
So with some offset applied 6V is the spec of interest to you. 

[nbritton]

Also, it appears the Rigol is running Linux kernel version 3.12, and the Siglent is running version 3.19. I normally work on the command line via a bash SSH shell, I know the Rigol has an SSH shell but I don't know if the Siglent does. How fleshed out are the user land environments? Do they have a toolchain, can you install applications on them, for instance GCC, Python, Perl, Bash?

I'm more into mathematics and physics then I am electronics, so I like playing with FFT formulas and doing number theory stuff, so being able to program the AWG using arbitrary complex mathematical formulas on the command line would be of great interest to me.

It doesn't matter as user access is closed, unless hacked... even when you can enable sshd, you don't want to go inside and mess with the pre-installed software (except to do some sort of hacking...).  Remember, it is not a general use PC, it is a test equipment and better to leave the SW as untouched as possible

Ah yeah, as a computer systems engineer, I'm usually the guy designing this kind of embedded stuff since my forte is linux, so I have an innate desire to get in there and tinker. I look at scopes that are hackable not completely because of the extra features, but more so I just love to tinker with them and try to do things they were never intended to do. The journey is the sort of the destination for me, I'd be the guy looking to upgrade the kernel version on the darn thing simply for the sake of the challenge.