Which scope manufacturer provides the best user experience?

[Bob Sava]

Definitely had some last minute second thoughts, but am hoping that any sense of buyer's remorse/regret will evaporate once I begin to actually use the scope.

Sounds like your first hand experience with Rigol and Siglent made you look elsewhere. 

[TimNJ]

That is Dave's opinion and he just doesn't like GW Instek for some reason.

Dave junior is a fan though. 

How do you know that?

[TimNJ]

Definitely had some last minute second thoughts, but am hoping that any sense of buyer's remorse/regret will evaporate once I begin to actually use the scope.

Sounds like your first hand experience with Rigol and Siglent made you look elsewhere.

Something like that. That said, while I really don't like Rigol's approach, Siglent isn't half bad. But, Siglent is definitely still a little quirky. I'm not saying that GWInstek will be the answer to all of my problems, but I thought it  was at least worthy of a shot.

[akimmet]

I have a GW Instek GDS-1102B, and the UI is far better than the Rigol DS1054Z I demoed.
The frustrating UI on the Rigol was the main reason for getting the Instek scope over the Rigol.

Since my purchase there have been newer models that are probably better choices than the Instek.
If I needed to get a new scope today, I would probably spend the extra for the newer Keysight DSOX1102G.

[Fungus]

Dave junior is a fan though. 

How do you know that?

Because he said so:


nb. Be sure to watch all the way through.

[TimNJ]

I have a GW Instek GDS-1102B, and the UI is far better than the Rigol DS1054Z I demoed.
The frustrating UI on the Rigol was the main reason for getting the Instek scope over the Rigol.

Since my purchase there have been newer models that are probably better choices than the Instek.
If I needed to get a new scope today, I would probably spend the extra for the newer Keysight DSOX1102G.

That's interesting. The only advantage I see the DSOX1102G would have over the GDS-1102B is optional serial decode (for an extra $150-300) and sampling rate. What do you see in the Keysight that you like, in particular?

GW Instek is like half the price. You do get 2x max sampling rate on Keysight, but only 1Mpt memory, compared to 10Mpt on GW Instek. I think 1Mpt is just too small these days, but maybe that's just my own experience.

[akimmet]

I have a GW Instek GDS-1102B, and the UI is far better than the Rigol DS1054Z I demoed.
The frustrating UI on the Rigol was the main reason for getting the Instek scope over the Rigol.

Since my purchase there have been newer models that are probably better choices than the Instek.
If I needed to get a new scope today, I would probably spend the extra for the newer Keysight DSOX1102G.

That's interesting. The only advantage I see the DSOX1102G would have over the GDS-1102B is optional serial decode (for an extra $150-300) and sampling rate. What do you see in the Keysight that you like, in particular?

GW Instek is like half the price. You do get 2x max sampling rate on Keysight, but only 1Mpt memory, compared to 10Mpt on GW Instek. I think 1Mpt is just too small these days, but maybe that's just my own experience.

Mostly for the built-in signal generator, and segmented memory. I was also impressed by some of the videos that show how fast the UI is.

[TimNJ]

I have a GW Instek GDS-1102B, and the UI is far better than the Rigol DS1054Z I demoed.
The frustrating UI on the Rigol was the main reason for getting the Instek scope over the Rigol.

Since my purchase there have been newer models that are probably better choices than the Instek.
If I needed to get a new scope today, I would probably spend the extra for the newer Keysight DSOX1102G.

That's interesting. The only advantage I see the DSOX1102G would have over the GDS-1102B is optional serial decode (for an extra $150-300) and sampling rate. What do you see in the Keysight that you like, in particular?

GW Instek is like half the price. You do get 2x max sampling rate on Keysight, but only 1Mpt memory, compared to 10Mpt on GW Instek. I think 1Mpt is just too small these days, but maybe that's just my own experience.

Mostly for the built-in signal generator, and segmented memory. I was also impressed by some of the videos that show how fast the UI is.

Oh right I forgot it has a wave gen. Yeah, that does make the Keysight a bit more appealing, for sure.

Do you find the UI of GW Instek slow? (A little late for me now!)

[nctnico]

A build-in wavegen is usually very expensive and limited compared to an external one.

[akimmet]

Do you find the UI of GW Instek slow? (A little late for me now!)

The GW Instek is by no means slow, they are far more responsive than the Rigols I have tried. Only that the Keysight scopes seem to have the most responsive UI in the entry level model range.

[TimNJ]

Just wanted to share a few of the alternatives I was considering.

My main requirements were:
1.) A user interface that doesn't make me want to light myself on fire
3.) Good overall performance + serial decode
2.) Relatively inexpensive (around $1,000USD or less, June 2018)

From a technical perspective, the GW Instek GDS-2000E is the best, but the Siglent isn't too far off. Siglent appears to be a great value, but I wanted a more "finished" product, one with fewer bugs/quirks. Ultimately, I bought the GW Instek in the hope that I would "enjoy" using it more. That's a hard thing to assign a monetary value to, I know. I'll let you all know if I think it was worth it, compared to my experiences with Rigol and Siglent at work.

See below table.

Also, WHO IN THEIR RIGHT MIND is buying a Tektronix DPO2000B ? ? ? ?

Edit: Removed table. Will re-upload with corrections and improvements.

[tautech]

Couple of errors there WRT sampling rate with all channels on. 

[TimNJ]

Couple of errors there WRT sampling rate with all channels on. 

Which ones? I thought I had those all figured out. Thanks.

EDIT: Ah, Siglent SDS1000X-E is 500MSa/s with all 4 channels, isn't it..

[tautech]

Couple of errors there WRT sampling rate with all channels on. 

Which ones? I thought I had those all figured out. Thanks.

EDIT: Ah, Siglent SDS1000X-E is 500MSa/s with all 4 channels, isn't it..

Plus any others that run 1 GSa/s processors for all 4 channels. Do they use one or two ?
DSO architecture has a few traps for the unwary.......another, total memory and how it's divided up with all channels active. What do the data sheets say and what don't they say. 
Another; Trigger path, digital or analogue.

[TimNJ]

Couple of errors there WRT sampling rate with all channels on. 

Which ones? I thought I had those all figured out. Thanks.

EDIT: Ah, Siglent SDS1000X-E is 500MSa/s with all 4 channels, isn't it..

Plus any others that run 1 GSa/s processors for all 4 channels. Do they use one or two ?
DSO architecture has a few traps for the unwary.......another, total memory and how it's divided up with all channels active. What do the data sheets say and what don't they say. 
Another; Trigger path, digital or analogue.

Yeah. As I was sifting through the datasheets for these scopes, I noticed quite a bit of ambiguity. I guess every manufacturer wants to say "We have a 1GSa/s/ch oscilloscope!!" but exactly what that means in each case is kind of unknown. Makes these kinds of comparisons tricky, if not misleading. But it's the best I can do.

[bugi]

Also difficult to compare 2 vs 4 channel scopes in the same sheet; to be fair, the 4 channels should have only the best combination of two channels enabled to compare against 2 channel scopes. Typically the sample rates and memories can then 2x better. But in that case mentioning the 4 channels (of two would not be in use) would be misleading. Yep, almost all such parameters would need own rows in the sheet like "2 channel performance" and "4 channel performance", and the latter row would just have "N/A" for the 2-channel scopes.

[rf-loop]

Just wanted to share a few of the alternatives I was considering.

My main requirements were:
1.) A user interface that doesn't make me want to light myself on fire
3.) Good overall performance + serial decode
2.) Relatively inexpensive (around $1,000USD or less, June 2018)

From a technical perspective, the GW Instek GDS-2000E is the best, but the Siglent isn't too far off. Siglent appears to be a great value, but I wanted a more "finished" product, one with fewer bugs/quirks. Ultimately, I bought the GW Instek in the hope that I would "enjoy" using it more. That's a hard thing to assign a monetary value to, I know. I'll let you all know if I think it was worth it, compared to my experiences with Rigol and Siglent at work.

Also, WHO IN THEIR RIGHT MIND is buying a Tektronix DPO2000B ? ? ? ?

This table have some errors. Why do table without any facts check. This is like trumpth, not truth.

Some have corrected Siglent what also was bullshit in original table. It have 2 ADC and 2 14M memory.
2 pcs 2 channel groups. Both have shared 1GSa/s ADC and 14M memory.

So it 1GSa 14M for one channel but also
1GSa/s  simultaneously for 2 channels with 14M fore both channel.
If more than 2 channels is in use then all channels have max 500MSa/s and 7M

But then GoodWill  GDS1000B (4channel model)
It have ONE Hittite HMCAD1511 ADC and for all channels on simultaneously it really can not 1GSa/s for all channels. Impossible.
So, it can max 1GSa/s for 1 channel in use.
If two channels are in use simultaneously it can max 500MSa/s
If more than 2 chanels in use it can do max 250MSa/s   Why you write 1GSa/s?
I do not know how this 10M memory is arranged.

GDS2000E, do it really have 4pcs 1GSa/s ADC?
How is memory shared if not fully separate ADC's?

Yes I know GoodWill do not tell it clearly in any place what I have read. Why they shame truth - I do not know.

As far as I know GDS2000E use  2 pcs  2x500MSa/s ADC !  (ADC08DL502)
How mem is arraged so that it have 10M for all channels simultaneously, I really do not know is it or not.
But one is sure, it do NOT have 1GSa/s for all channels simultaneously in use. Simply, it is technically impossible. So why you write it have 1GSa/s for all cnhannels on simultaneously? (even GW do not claim it when read carefully)

What other thigs are wrong in table, I have not time for more Fact-checking.

[nctnico]

No need to get your panties in a bunch. GDS2000E has 500Ms/s with 4 channels enabled and 10Mpts per channel with 4 channels enabled. In segmented mode the GDS2000E has 20Mpts of memory per channel available.

[tautech]

GDS2000E has 500Ms/s with 4 channels enabled and 10Mpts per channel with 4 channels enabled. In segmented mode the GDS2000E has 20Mpts of memory per channel available.

Well, yes you will know. Thanks.
So double the price for an 8" display and 3 more Mpts/channel than X-E?
Hmmmm.

[nctnico]

GDS2000E has 500Ms/s with 4 channels enabled and 10Mpts per channel with 4 channels enabled. In segmented mode the GDS2000E has 20Mpts of memory per channel available.

Well, yes you will know. Thanks.
So double the price for an 8" display and 3 more Mpts/channel than X-E?
Hmmmm.

Don't forget firmware which works as advertised  Specs on paper are just that... specs on paper. I originally bought my GDS2204E as an experiment to have a bring-along scope in case I need a scope at a customer. It ended up replacing a rather expensive Agilent scope because it turned out to have less noise and being easier to operate.

[rf-loop]

In SDS1104X-E fast seqmented mode (what GoodWill 2kE do not even have) max memory per channel is 27M. (just cheked one example, 4channel on, each channel 3912 segment, each segments length is 7kpts = 4x 27384000 sample points ) Siglent normal mode waveform history buffer (always backround, works more like GW segmented acquisition using normal acquisition speed, up to measured 118kwfm/s continuous average and 128kwfm/s max peak. These are not data sheet values, these are true measured values)  (1ch in use, 50ns/div)
I have never seen true measured facts about GW speed.

Max segments 4x 80000, max quaranteed acquisition speed is up to measured well over 450000 segment/s (measured) and conservative value in data data sheet 400kwfm/s). (if compare to Keysight... trigger rearm 19us, max 50 segments joke....)
Also Siglent 3 channel 500 data point SFRA useability for serious work is far over Keysight (and GoodWill)

[Bob Sava]

@TimNJ could you compare analog stages between the three scopes, specifically how they handle signal distortion at high gain as explained here:

https://www.eevblog.com/forum/testgear/siglent-technical-support-join-in-eevblog/msg1533101/#msg1533101

[TimNJ]

This table have some errors. Why do table without any facts check. This is like trumpth, not truth.

Some have corrected Siglent what also was bullshit in original table. It have 2 ADC and 2 14M memory.
2 pcs 2 channel groups. Both have shared 1GSa/s ADC and 14M memory.

So it 1GSa 14M for one channel but also
1GSa/s  simultaneously for 2 channels with 14M fore both channel.
If more than 2 channels is in use then all channels have max 500MSa/s and 7M

But then GoodWill  GDS1000B (4channel model)
It have ONE Hittite HMCAD1511 ADC and for all channels on simultaneously it really can not 1GSa/s for all channels. Impossible.
So, it can max 1GSa/s for 1 channel in use.
If two channels are in use simultaneously it can max 500MSa/s
If more than 2 chanels in use it can do max 250MSa/s   Why you write 1GSa/s?
I do not know how this 10M memory is arranged.

GDS2000E, do it really have 4pcs 1GSa/s ADC?
How is memory shared if not fully separate ADC's?

Yes I know GoodWill do not tell it clearly in any place what I have read. Why they shame truth - I do not know.

As far as I know GDS2000E use  2 pcs  2x500MSa/s ADC !  (ADC08DL502)
How mem is arraged so that it have 10M for all channels simultaneously, I really do not know is it or not.
But one is sure, it do NOT have 1GSa/s for all channels simultaneously in use. Simply, it is technically impossible. So why you write it have 1GSa/s for all cnhannels on simultaneously? (even GW do not claim it when read carefully)

What other thigs are wrong in table, I have not time for more Fact-checking.

rf-loop,

Thank you for pointing out these errors. These were the specs I extracted from a "datasheet analysis" of the scopes. I guess I was ignorant to some of the subtleties of scope architecture and the possible differences between scopes which look the same on paper. In my defense, the numbers on the table are what the specifications seem to imply. But alas, not necessarily the whole truth.

I did not intend to mislead, and will of course fix any and all mistakes. We fix each other's mistakes all the time on this forum.

[rf-loop]

This table have some errors. Why do table without any facts check. This is like trumpth, not truth.

Some have corrected Siglent what also was bullshit in original table. It have 2 ADC and 2 14M memory.
2 pcs 2 channel groups. Both have shared 1GSa/s ADC and 14M memory.

So it 1GSa 14M for one channel but also
1GSa/s  simultaneously for 2 channels with 14M fore both channel.
If more than 2 channels is in use then all channels have max 500MSa/s and 7M

But then GoodWill  GDS1000B (4channel model)
It have ONE Hittite HMCAD1511 ADC and for all channels on simultaneously it really can not 1GSa/s for all channels. Impossible.
So, it can max 1GSa/s for 1 channel in use.
If two channels are in use simultaneously it can max 500MSa/s
If more than 2 chanels in use it can do max 250MSa/s   Why you write 1GSa/s?
I do not know how this 10M memory is arranged.

GDS2000E, do it really have 4pcs 1GSa/s ADC?
How is memory shared if not fully separate ADC's?

Yes I know GoodWill do not tell it clearly in any place what I have read. Why they shame truth - I do not know.

As far as I know GDS2000E use  2 pcs  2x500MSa/s ADC !  (ADC08DL502)
How mem is arraged so that it have 10M for all channels simultaneously, I really do not know is it or not.
But one is sure, it do NOT have 1GSa/s for all channels simultaneously in use. Simply, it is technically impossible. So why you write it have 1GSa/s for all cnhannels on simultaneously? (even GW do not claim it when read carefully)

What other thigs are wrong in table, I have not time for more Fact-checking.

rf-loop,

Thank you for pointing out these errors. These were the specs I extracted from a "datasheet analysis" of the scopes. I guess I was ignorant to some of the subtleties of scope architecture and the possible differences between scopes which look the same on paper. In my defense, the numbers on the table are what the specifications seem to imply. But alas, not necessarily the whole truth.

I did not intend to mislead, and will of course fix any and all mistakes. We fix each other's mistakes all the time on this forum.

Yes, there is really many traps if only read manufacturers data sheets and advertisements. Finding truth need some times lot of investigations if can not take equipment on table and test in real life. Some manufactures hide more than others. Sad, we are out from time when documents was nearly perfect if take example old Hewlett-Packard and old time Tektronix where nearly every detailed things was explained with deep details.
 

[TimNJ]

Okay, I am fairly certain this is correct. Please correct me if there are any other glaring errors.

This table probably isn't of much use to anyone, but I'll just leave it here anyway