Keysight DSOX3014T vs DSOX1204G - is it really worth 3.5 times the money? Vote!

[EEVblog]

Bussmann
Can you please explain the forum user account ci11

[tautech]

Everyone has valid points and this noob is no closer to a decision.

Let's add further to the confusion.
Threads like your's come come along from time to time and there's a wealth of info here tucked away in forum archives.
Here's a reasonably recent one comparing the entry level Keysight to other entry level DSO's which in itself is a little unfair to put a 2GSa/s DSO against 1GSa/s DSO's and as you might imagine when getting up in frequency the 2GSa/s DSO's have some advantage.
When we dropped a 2GSa/s $600 2ch Siglent into the mix it allowed for a fairer apple vs apples comparison:
https://www.eevblog.com/forum/testgear/dso-bandwidth-test-sds1104x-e-dsox1102g-to1104-gds1054b/

XY for channel separation waveform display

Providing the channel frequency disparity is not too great DSO's can do well enough here.

Tell us what you need exactly and I'm sure there's enough of us here with 2ch AWG's that can do a videos for a fair comparison.

[2N3055]

To OP.

Sorry for this, I mean no disrespect,  but I'm going to be blunt..


Tautech is right.. This is getting nowhere. I will try to summarize what others and myself think..

You seem to have very strong opinions about things based on little knowledge, and some misconception.

1. USB scopes are plenty fast, it's just they have no buttons. Picoscope 3000D can get up to 100000 WFM's per second. It is more scope than most of people need, with some very advanced features. But some people need (want?) physical scope (with buttons and no PC).

2. You say you want tools to work on audio. Good. So you need a scope that doesn't have much bandwidth, and has low noise and high sensitivity, so you can see signals from microphones and MC pickups and such...

3. I'm still confused about need for X-Y display (apart from oscilloscope art). Do you use it as an audio vectorscope? Could you please explain it a bit more, I would like to know more about it.


- Rigol DS2102A (2GSa/sec, 100MHZ 2ch, 500uV/div)
- Siglent SDS1104X-E (1GSa/sec, 100MHZ 4ch, 500uV/div)
- Siglent SDS2202X-E (2GSa/sec, 200MHZ 2ch, 500uV/div)
- Micsig TO1074 Plus (1GSa/sec, 70MHZ 4ch, 500uV/div)

All of these are up to some 600 USD. If you really need 4ch, then I would chose Siglent SDS1104X-E (1GSa/sec, 100MHZ 4ch, 500uV/div) for audio equipment.
It is waay more capable instrument than DSOX1204G despite not being fancy name.
It is more scope than you need for what you described.
And you get all the digital decodes and very long memory that will be useful at frequencies you're looking at.. And it is fast and OK to use.
DSOX3014T is not a good scope for audio. It is meant for other uses...

For the rest of the money up to DSOX3014T, keep the analog scope you have for X-Y (i agree with Dave), and find a good used Audio Precision audio analyser. Or buy QuantAsylum QA401 Audio Analyzer... Or keep your ARTA setup, it is probably more than enough.

You seem to have decided to treat yourself with "premium product", you know you don't need it, and now you want us to persuade you it's worth it.

In your case it's not.
Nicer knobs, nice logo on the front and actually worse specs for your use case are not worth the money.

Just buy Siglent SDS1104X-E and be amazed how good it is compared to what you had before.
And if you have leftover money buy other specialized audio equipment, because if you want audio measurements, there is no scope in the world that comes even close to even the cheapest audio analysers. You don't like the knobs? You'll get used to it after a week.. EVERY new piece of equipment feels weird to me when I first start using it. MSOX3104T felt weird too at first, and I had to remember where the options are (there is ton of them), had to remember it has touch screen etc etc... After a week or two, it became normal.

This is what I would do in your position. And it seems all the responders mostly. You don't need a poll. People already voted.

If you want to buy "premium" product for the purpose that it will make you feel better that you have treated yourself to something nice, then by all means do it, it's your money.

Hope you don't take this the wrong way.

Best regards,

[Fungus]

You seem to have decided to treat yourself with "premium product"

Seems that way.

, you know you don't need it, and now you want us to persuade you it's worth it.

In your case it's not.
Nicer knobs, nice logo on the front and actually worse specs for your use case are not worth the money.

This guy is a salesman's dream. If he wants clicky knobs, he wants clicky knobs. Let him buy the more expensive device. Specs aren't good enough? Move up the model range to something that is...

[Performa01]

Bussmann
Can you please explain the forum user account ci11

That rings a bell - I already asked myself where from I got that déjà vu...

A request from member ci11 was the whole reason why I've added a couple of posts to my review thread, starting at reply #136:

https://www.eevblog.com/forum/testgear/siglent-sds1104x-e-in-depth-review/msg2021602/#msg2021602

The final challenge was to display a low frequency 150µV signal (reply #138), which of course doesn't yield pretty results even with a 500µV/div frontend. At 10Hz the 1/f noise is substantial on any modern scope with split path input buffers (and no, it's not the CMOS technology but the circuit architecture!) and then there are only some 18 counts (~ 4 bits) resolution left for signals that low. Amazing enough what just 16 times averaging still can squeeze out of this.

The same posting also shows a true 16 bit DSO in comparison, which does a very decent job, even though the highest sensitivity is only 2mV/div.

With regard to X-Y, it's certainly not the lack of speed that makes modern DSOs inferior to an analog scope, but the lack of resolution. With only 8 bits, you get 256 signal levels at best, which means you have 2 vertical pixels per LSB on the screen in Y-t mode. This is not a real problem, because in Y-t mode we still have full resolution on the X-axis.

In X-Y mode, we have the limited resolution on both axes, so each measurement point is a 2x2 pixel square. That doesn't look very pretty. A good intensity graded display can mask this, but the trace gets rather thick and the lack of detail is still there. 10 bits would certainly help with this.

Even better if you have 16 bits like with the PicoScope 4262, see attached video (just rename it to .mp4).
This certainly outperforms analog scopes, if only because we can have the result at full resolution on a full HD screen.

[2N3055]

Also there was the question about low amplitude signals and solid triggering....

My contribution from Picoscope 4262.
No filtering, 16bit.
With filtering and resolution enhancement is much cleaner...

Tool for the job.

P.S. @Performa01, it's your fault that I had to get 4262...  . At least that's what I said to SWIMBO 

[Performa01]

P.S. @Performa01, it's your fault that I had to get 4262...  . At least that's what I said to SWIMBO 

I don't mind to be the fall guy - as long as that purchase made you happy!   
I only hope your SWIMBO isn't cross with me now!

[2N3055]

P.S. @Performa01, it's your fault that I had to get 4262...  . At least that's what I said to SWIMBO 

I don't mind to be the fall guy - as long as that purchase made you happy!   
I only hope your SWIMBO isn't cross with me now!

Just joking...

4262 is nothing short of amazing. With it's 8-9uV noise floor, I can measure power supply noise directly, audio stuff, etc. I use it a lot actually. It was a good purchase, really, thank you for drawing my attention to it...

Best regards,

[egonotto]

Hello,

for that I have a 4262 are the posts of Performa01 not whole unimportant .

As I don't know  was SWIMBO means I search.
Now I know it is "Oberste Heeresleitung" (topmost army command)

@2N3055: you wrote "because if you want audio measurements, there is no scope in the world that comes even close to even the cheapest audio analysers".

I think the 4262 comes very close. So a Picoscope 4262 and a Siglent SDS1104X-E can be a good solution for TO

Best regards
egonotto

[Mr. Scram]

Just joking...

4262 is nothing short of amazing. With it's 8-9uV noise floor, I can measure power supply noise directly, audio stuff, etc. I use it a lot actually. It was a good purchase, really, thank you for drawing my attention to it...

Best regards,

Can't you do power supply noise measurements on the DSOX3014T? I thought it was part of the power analysis suite. Looks like Keysight calls it output analysis.

https://literature.cdn.keysight.com/litweb/pdf/5990-8869EN.pdf?id=2096805

[2N3055]

Hello,

for that I have a 4262 are the posts of Performa01 not whole unimportant .

As I don't know  was SWIMBO means I search.
Now I know it is "Oberste Heeresleitung" (topmost army command)

@2N3055: you wrote "because if you want audio measurements, there is no scope in the world that comes even close to even the cheapest audio analysers".

I think the 4262 comes very close. So a Picoscope 4262 and a Siglent SDS1104X-E can be a good solution for TO

Best regards
egonotto

It also means Minister of Finance  too... 

You are right, I use it for audio, for what I do. Nevertheless, audio analysers stil another level better. But for what I do (and hear) Pico does all I need. But I don't do advanced stuff.

I agree, that is a good pairing. And still much cheaper than DSOX3014T, combined.

Regards,

[2N3055]

Just joking...

4262 is nothing short of amazing. With it's 8-9uV noise floor, I can measure power supply noise directly, audio stuff, etc. I use it a lot actually. It was a good purchase, really, thank you for drawing my attention to it...

Best regards,

Can't you do power supply noise measurements on the DSOX3014T? I thought it was part of the power analysis suite. Looks like Keysight calls it output analysis.

https://literature.cdn.keysight.com/litweb/pdf/5990-8869EN.pdf?id=2096805

Yes, you are correct, that is a part of PWR option. But it is suitable only for ripple on switching power supplies. Lowest real V/div range is 4 mv/div, lower than that is software zoom.

Pico 4262 has noise floor of 8-9uV. With it, you can measure low noise linear supplies, directly, without preamplifier. And to measure noise on LTZ1000, you would need only 100X magnification.

[Mr. Scram]

Yes, you are correct, that is a part of PWR option. But it is suitable only for ripple on switching power supplies. Lowest real V/div range is 4 mv/div, lower than that is software zoom.

Pico 4262 has noise floor of 8-9uV. With it, you can measure low noise linear supplies, directly, without preamplifier. And to measure noise on LTZ1000, you would need only 100X magnification.

Thanks for clarifying, that makes sense. What kind of preamplifier would you use for these kinds of measurements? I've no idea what kind of bandwidth and gain would be suitable.

[2N3055]

Yes, you are correct, that is a part of PWR option. But it is suitable only for ripple on switching power supplies. Lowest real V/div range is 4 mv/div, lower than that is software zoom.

Pico 4262 has noise floor of 8-9uV. With it, you can measure low noise linear supplies, directly, without preamplifier. And to measure noise on LTZ1000, you would need only 100X magnification.

Thanks for clarifying, that makes sense. What kind of preamplifier would you use for these kinds of measurements? I've no idea what kind of bandwidth and gain would be suitable.

It all depends what are you measuring. Is source low or high impedance? Are you interested in 1/f noise, 0.1 Hz to 10 Hz bandwidth is interesting. On general purpose PSUs industry standards are 10 Hz-100 kHz and 100 Hz-100 kHz... So you need front end that should be correct for your source impedance, and you can filter by using analog filters, or you can capture and filter in software, providing you don't have large contributions from out of band that would compromise your dynamic range, in which case you would need analog filtering anyways, or combination of both. I'm looking at making LT1037 based first stage to experiment a bit.. I don't have preamp at the moment. That is why I said 4262 is great because you can direct sample all "normal" PSU in bandwidth from DC to 5MHz.

[Performa01]

Ideally, the preamp would exceed the bandwidth of the DSO, i.e. >5MHz, but that's a tough requirement for low noise / high gain. I think we never need high impedance for these kind of very low noise measurements, hence bipolar audio amplifiers can be used. An input impedance of 100k should still be possible, but DC input coupling would be difficult because of the high input bias current. AC coupling on the other hand has the disadvantage that the source impeance is increased at low frequencies, hence also the noise. A 10µF PP capacitor at the input would result in about 1dB drop at 1Hz, which might be an acceptable compromise, but it is physically big and expensive.

I would recommend the LM4562NA dual op amp (or the very similar LME49720). Cascading the two amplifiers with a gain of 20dB each results in a total gain of 40dB that is flat up to at least 1MHz (-6dB bandwidth ~5MHz in theory). Input noise density would be 2.7nV/sqrt(Hz) typical.

For significantly more money, the OPA2211(A)I provides an input noise density of typical 1.1nV/sqrt(Hz) and similar performance otherwise.

If high input impedance and/or DC coupling is a must, then the AD8676A with typical 2.8nV/sqrt(Hz) can be considered - but the bandwidth is significantly (about 5-6 times) lower.

[EEVblog]

Yes, you are correct, that is a part of PWR option. But it is suitable only for ripple on switching power supplies. Lowest real V/div range is 4 mv/div, lower than that is software zoom.

Pico 4262 has noise floor of 8-9uV. With it, you can measure low noise linear supplies, directly, without preamplifier. And to measure noise on LTZ1000, you would need only 100X magnification.

Thanks for clarifying, that makes sense. What kind of preamplifier would you use for these kinds of measurements? I've no idea what kind of bandwidth and gain would be suitable.

Industry standard PSU noise bandwidth is 20MHz. It's one of the reasons why 20MHz is the standard BW limits on all modern digital scopes.

[2N3055]

Yes, you are correct, that is a part of PWR option. But it is suitable only for ripple on switching power supplies. Lowest real V/div range is 4 mv/div, lower than that is software zoom.

Pico 4262 has noise floor of 8-9uV. With it, you can measure low noise linear supplies, directly, without preamplifier. And to measure noise on LTZ1000, you would need only 100X magnification.

Thanks for clarifying, that makes sense. What kind of preamplifier would you use for these kinds of measurements? I've no idea what kind of bandwidth and gain would be suitable.

Industry standard PSU noise bandwidth is 20MHz. It's one of the reasons why 20MHz is the standard BW limits on all modern digital scopes.

That is correct as per standard measurement protocol for switching PSU-s for desktop computer, monitor, printer, notebook computer, fax machine, photocopy machine and etc. For linear supplies for audio preamp you might measure to 100kHz or for some fast switcher for super fast FPGA  you might go up to 500MHz if that is what is of interest to you.