first time buyers oscilloscope analysis paralysis help!!!!!

[bd139]

Agreed with rstofer on all points.

On point the DS1054Z is slow as hell. The FFT isn't that bad really but it's an absolute bastard to get it to do exactly what you want and you need to know how it works deeply before it's valuable. I am managing to use it as a naive HF spectrum analyser and it works pretty well for that believe it or not. But it takes me 5 minutes to tweak it for each measurement.

[tautech]

I mostly agree. Some comments though.

At this point I sincerely hope Siglent win the market share battle. Rigol do not deserve it due to the crap they are throwing out (I am very very unimpressed by the MSO5000 after actually playing with one and seeing the total lack of progress, plus typically the things like the counter on the DG1022Z being a steamer that is just not going to be fixed) and neither do Keysight (too expensive, very little feature progress). That's not to say any of them are bad and none of them are useful but there are some seriously glaring product holes. The drive to throw new products on the market means we never get maturity these days and it's frustrating as hell.

On performance verification, I mean deep verification rather than superficial front end validation. A DSO is a much more complicated beast and a lot of the features are not well specified or demonstrated and are entirely in the software domain. For example the FFT and decode implementations have no baseline specification other than a feature tick box. If you look at the commercial software market this stuff is automatically regression tested to buggery. This is definitely not evident in Rigol's products at least and at the price point, getting any of that is unlikely. Keysight are on a winner there but you pay for it and there are some horrible issues I've observed in some of the videos on YouTube. I can't speak for Siglent here but they appear to be throwing stuff out quickly and reactive which is again better than both. Velocity towards maturity is a winner.

Yep, that arises from 2 angles, sales to fund it and good quality user feedback.
On the other hand too frequent firmware releases to address issues can be seen by some as the product's still in the infant stage while withholding FW until much has been addressed can frustrate adopters that have found a bug and have to wait for months and months before new FW is released to fix it !
This is hard to get right to everyone's satisfaction.

Plus when good feature requests are accepted that might add additional functionality that was never considered in the design stage it adds more even pressure to the SW engineers and the FW release cycle.
Yet for the patient the rewards can be very satisfying.

I actually nearly bought a keysight the other day to replace my Rigol. After doing a lot more research I skipped this and looked at the Siglent. Then I discovered that some models have had probe compensation issues and reminded myself that I'm going to stay with the known unknowns, which was really the poorly contrived point I was trying to make

HAD with emphasis !
Designers really learnt their lesson on that one and won't overlook that again.

[nctnico]

is there anyone here who doesn't have a financial interest in their advice?

You should turn this question around. Tautech is the only one with a financial interest. If there is anyone else then that person manages to hide it very well.

[tautech]

is there anyone here who doesn't have a financial interest in their advice?

You should turn this question around. Tautech is the only one with a financial interest. If there is anyone else then that person manages to hide it very well.

Nope, there's dozens of resellers and dealers that are members the forum.

[nctnico]

For example the FFT and decode implementations have no baseline specification other than a feature tick box. If you look at the commercial software market this stuff is automatically regression tested to buggery. This is definitely not evident in Rigol's products at least and at the price point, getting any of that is unlikely. Keysight are on a winner there but you pay for it and there are some horrible issues I've observed in some of the videos on YouTube. I can't speak for Siglent here but they appear to be throwing stuff out quickly and reactive which is again better than both. Velocity towards maturity is a winner.

However, not all Chinese brands are throwing untested stuff on the market. MicSig for example only seems to release firmware which is well tested. When I tested the TO1104 (using a thourough test plan) I have not been able to find a bug. The GW Instek GDS-2000E series had a few issues which they fixed within 2 weeks after reporting. Both oscilloscopes where relatively new when I tested them.

[nctnico]

is there anyone here who doesn't have a financial interest in their advice?

You should turn this question around. Tautech is the only one with a financial interest. If there is anyone else then that person manages to hide it very well.

Nope, there's dozens of resellers and dealers that are members the forum.

I think in this thread you are the only dealer actively participating. Overall there are very few equipment dealers which participate and most I've seen are easy to spot and/or make their relationship to a brand very clear.

[tautech]

Today, I would give the Siglent SDS1204X-E serious consideration.  Maybe even try to unlock the SDS1104X-E (details elsewhere). 

Yes that was my take on things too before the $619 SDS2202X-E was released. Apart from having just 2 channels it has all the fruits of the SDS1*04X-E models plus a couple more and 2GSa/s performance.
Unlock them to the SDS2352X-E and you have a really nice little scope.

[Fungus]

The Keysight is hackable for extra bandwidth.

[rstofer]

Today, I would give the Siglent SDS1204X-E serious consideration.  Maybe even try to unlock the SDS1104X-E (details elsewhere). 

Yes that was my take on things too before the $619 SDS2202X-E was released. Apart from having just 2 channels it has all the fruits of the SDS1*04X-E models plus a couple more and 2GSa/s performance.
Unlock them to the SDS2352X-E and you have a really nice little scope.

Nice idea!  I already have a 4 channel scope and since SPI is in the low MHz, there's no reason to buy another 4 channel scope if I can get more bandwidth for the same kind of money in 2 channels.  Features will be the new metric.  How good is the FFT (not that I ever use that feature), how responsive is the GUI, what features does it add over the DS1054Z, those kind of things.

The SDS2352X-E is a 350 MHz scope at a reasonable price without unlocking a lesser model - maybe it's time to get rid of my Tek 485 boat anchor.  But I would want the MSO options (for no really good reason).  It looks to have a sizeable screen.

[tautech]

Today, I would give the Siglent SDS1204X-E serious consideration.  Maybe even try to unlock the SDS1104X-E (details elsewhere). 

Yes that was my take on things too before the $619 SDS2202X-E was released. Apart from having just 2 channels it has all the fruits of the SDS1*04X-E models plus a couple more and 2GSa/s performance.
Unlock them to the SDS2352X-E and you have a really nice little scope.

Nice idea!  I already have a 4 channel scope and since SPI is in the low MHz, there's no reason to buy another 4 channel scope if I can get more bandwidth for the same kind of money in 2 channels.  Features will be the new metric.  How good is the FFT (not that I ever use that feature), how responsive is the GUI, what features does it add over the DS1054Z, those kind of things.

1 Mpts and sharp enough as it is but needs updating with the same features added into SDS1kX-E FFT.
The GUI is fine and no different to Siglent's other X-E DSO's.
28 Mpts mem depth that gives you the same capture length as DSO's with 1 GSa/s and 14 Mpts.
Fastest timebase 500ps.
50 Ohm inputs
Bode plot
Webserver

The SDS2352X-E is a 350 MHz scope at a reasonable price without unlocking a lesser model - maybe it's time to get rid of my Tek 485 boat anchor.  But I would want the MSO options (for no really good reason).

The 2352X-E come with really nice probes, better than those supplied with all 200 MHz models.
They are smaller and shorter with the compensation box at the BNC connector. These are still switchable 10:1, 1:1 too.

It looks to have a sizeable screen.

Just 7" like any other entry level scope but quite bright enough and not glossy at all.

Here's a FFT of the probe Cal and this was like the 1kX-E's before they added the fruits to it. FFT has just been improved.
Just found newly released FW and posted it here:
https://www.eevblog.com/forum/testgear/siglent-sds2000x-e/msg2684748/#msg2684748

I'll come back and post a 2nd shot of the improved FFT.
This is how FFT is now, here with Max Hold that cleans up all the fuzz down on the noise floor and markers and some tables added. Looking at a FFT of the probe Cal output.

[Mr. Scram]

Yes you don't think or for that matter even know !
FFS of course they do !

Good. Where is the page to search and order parts? How long are oscilloscopes supported and repaired?

Should we interpret the lack of a response as the absence of an answer?

[Fungus]

FFT has just been improved. Just found newly released FW and posted it

Keysight also keeps adding stuff:

https://www.keysight.com/main/software.jspx?cc=US&lc=eng&id=2820272

This is how FFT is now, here with Max Hold that cleans up all the fuzz down on the noise floor and markers and some tables added. Looking at a FFT of the probe Cal output.

Yeah, but FFT is just the start. Just look at all the stuff the Keysights can do for signal analysis:



(and how snappy the UI is...)

If it's within budget then the Keysight is easily the best choice for audio work in this price bracket.

[scatterandfocus]

Yeah, but FFT is just the start. Just look at all the stuff the Keysights can do for signal analysis:

I am only seeing a sine being modulated/demodulated.  What's useful about it for audio analysis?  I have very inexpensive software that can do all sorts of useful audio analysis and processing and many other things at a full dynamic range.  Why would I do audio analysis using a scope instead?   

[bd139]

I would use a decent sound card for audio analysis. 24-bit 96KHz M-audio job with some off the shelf software like TrueRTA.

That is quite nice though on the Keysight. I can’t think of a use for it though off the top of my head. I think the most useful thing I’ve seen is the bode plot function as that can be used to evaluate where phase shifts are going to give you a headache. I was going to actually bodge something up with my Rigol stuff and python driving LXI for this but in the end I just poked the circuit in LTspice and worked out the frequency compensation in there and added some headroom and then just swept it by hand to confirm.

Edit: I could have actually solved the above with an analogue scope and an analogue function generator 

[2N3055]

FFT has just been improved. Just found newly released FW and posted it

Keysight also keeps adding stuff:

https://www.keysight.com/main/software.jspx?cc=US&lc=eng&id=2820272

This is how FFT is now, here with Max Hold that cleans up all the fuzz down on the noise floor and markers and some tables added. Looking at a FFT of the probe Cal output.

Yeah, but FFT is just the start. Just look at all the stuff the Keysights can do for signal analysis:



(and how snappy the UI is...)

If it's within budget then the Keysight is easily the best choice for audio work in this price bracket.

What signal analysis? It has 4 basic math operations, low pass filter and 64k FFT. It is really very basic and has none of advanced stuff that 3000 series has.
Rigol DS1000Z literally has better math(and filters), and in this video filter is as slow as it is on the Rigol. Rigol FFT is crap though. But Siglent's FFT is insanelly more capable that what DSOX1102A has.

Also, while Rigol DS1000Z IS kinda slowish on some things, Siglent is really fast and has no worse usability in that regard than Keysight 1000 series.
And Keysight has small memory and only 50 segments in segmented mem.

If I had to choose between DSOX1102A and  SDS2352X-E  I would go with one that has 28x more memory (and up to 80,000 segments), 1 Mpoint FFT, Mixed signal expandability, 350 MHz bandwidth with 50 OHm, and all protocols and triggers included in basic price. And that is not Keysight, sorry...

While I think that Keysight 3000T series is currently still the best overall (all things included) scope in 3000 class, Keysight 1000 and 2000 series are too much money for very basic and very outdated scopes. They have specs of Hameg scopes from 10 years ago. They are very good quality but very basic, and acceptable only if you use only basic things and practically use them as replacement for analog scopes.  If that is your use case, and are willing to pay for brand name and maybe better support (depending on where you are) then they might be good choice. Otherwise, it is like comparing 1965 Porsche 911 (that had 130 HP) with today's Honda Civic Type R (320 HP).. Porsche is a fancy car with great brand but despite all that, Type R is both better and faster.

[nctnico]

Yes you don't think or for that matter even know !
FFS of course they do !

Good. Where is the page to search and order parts? How long are oscilloscopes supported and repaired?

Should we interpret the lack of a response as the absence of an answer?

Well I recall someone being able to order a new main board for his Siglent function generator. But this was a couple of years ago.

[tautech]

Yes you don't think or for that matter even know !
FFS of course they do !

Good. Where is the page to search and order parts? How long are oscilloscopes supported and repaired?

Should we interpret the lack of a response as the absence of an answer?

Well I recall someone being able to order a new main board for his Siglent function generator. But this was a couple of years ago.

I wasn't gunna feed the troll.

[scatterandfocus]

I wasn't gunna feed the troll.

It's a valid question.  I would like to know the answer myself, as I'm sure others would.

[electricMN]

Yes that was my take on things too before the $619 SDS2202X-E was released. Apart from having just 2 channels it has all the fruits of the SDS1*04X-E models plus a couple more and 2GSa/s performance.
Unlock them to the SDS2352X-E and you have a really nice little scope.

Is the SDS2074X hackable to a SDS2304X?

[SilverSolder]

[...]
As someone who fixes a lot of equipment, there are known unknowns and unknown unknowns. The newer kit tends to have a lot of the latter. I spent a good few hours scratching my head over a trigger issue on my DS1054Z the other day which turned out to be the last user cal going wonky. Older kit, you tend to be able to run performance validation against. New stuff, not so much.
[...]

Seems to me you put your finger on something important there:  newer instruments are filled with processors and software that touches the signal before you see it -  you are further away from the underlying reality in some ways; it is only too easy for some easily 'overlookable' setting on some deep menu to subtly change what you are looking at, possibly leading you down the garden path.

[BillB]

I wasn't gunna feed the troll.

It's a valid question.  I would like to know the answer myself, as I'm sure others would.

I have experience sending an in-warranty device for repair to Siglent's US based support (Ohio).  It was a relatively standard process: get an RMA, send the device, get repaired device returned. 

I think the days of having a schematic, bill-of-materials and parts-ordering forms right inside the Users Manual of a product are long-gone.  Most companies these days will fix stuff in warranty, offer to fix it for your for a price out of warranty, and that's about it.  Most aren't going to offer selling spare parts; they'd rather have you buy a new unit.

[tautech]

Most aren't going to offer selling spare parts; they'd rather have you buy a new unit.

Not offer, correct however parts are available.
Equipment has been so reliable that I rarely need to source anything. A folding bail handle that fits all SDM and SDG models IIRC was ~$10 retail.
I had to get one for my best customer and the paperwork was gunna cost more so I just sent it on through to them.

[tautech]

Yes that was my take on things too before the $619 SDS2202X-E was released. Apart from having just 2 channels it has all the fruits of the SDS1*04X-E models plus a couple more and 2GSa/s performance.
Unlock them to the SDS2352X-E and you have a really nice little scope.

Is the SDS2074X hackable to a SDS2304X?

I think so.
Have a look at what Defpom did with his older SDS2102, a pre X model:


However these too are getting dated and don't have the faster processor the more modern X-E's have but do have some features the X-E's don't however many are $$ options.
I keep needling Siglent for a SDS2**4X-E but I'm not sure if they can shoehorn two 2 GSa/s ADC onto a small PCB and still get it into the X-E form factor case and still get acceptable cooling with a somewhat quiet fan.
It would be a neat little DSO if they can.

[bd139]

Looks like Telonic are now distributing Siglent stuff here suddenly according to the booklet that came with Radcom. They may have another customer soon 

[Mr. Scram]

Not offer, correct however parts are available.
Equipment has been so reliable that I rarely need to source anything. A folding bail handle that fits all SDM and SDG models IIRC was ~$10 retail.
I had to get one for my best customer and the paperwork was gunna cost more so I just sent it on through to them.

Can people order say a front casing or a repair out of warranty? How would that work? It's not just about reliability but mishaps and wear do happen. I like being able to order a new casing when I inadvertently ruin the old one.