DSO Bandwidth test SDS1104X-E DSOX1102G TO1104 GDS1054B

[TK]

I had the opportunity to test the Bandwidth on the following Digital Storage Oscilloscopes:

Siglent SDS1104X-E (100MHz 1GSa/s 4 channels)
Siglent SDS1104X-E software modded to SDS1204X-E (200MHz 1GSa/s 4 channels) (Indicated as SDS1104X-E* in the table)
Siglent SDS1202X-E (200MHz 4 channels) (As tested by TAUTECH)
Keysight DSOX1102G (EDUX1002G uphacked to DSOX1102G) (2GSa/s 2+1 channels)
GWInstek GDS1054B (software license hacked to 300MHz) (1GSa/s 4 channels)
Micsig TO1104 (1GSa/s 4 channels)
Agilent 54622D (200MSa/s 2+16 channels) NOTE: Frequency over 200MHz is indicated as ">200MHz" but can see a very well defined sinewave up to 600MHz (Added 5-May-2019)
HP 54645D (2+16 channels) (Added 5-May-2019)

I used an Agilent signal generator to output a 0dBm (632 mVpp) signal.

Agilent Signal Generator --> N to SMA connector --> SMA - SMA cable --> SMA to BNC --> BNC 50ohm passthrough --> DSO

Frequency	SDS1104X-E	SDS1104X-E*	SDS1202X-E	DSOX1102G	GDS1054B	TO1104	54622D	54645D
10 MHz	645 mVpp	652 mVpp	640 mVpp	640 mVpp	630 mVpp	644 mVpp	641 mVpp	643 mVpp
100 MHz	500 mVpp	596 mVpp	608 mVpp	625 mVpp	485 mVpp	516 mVpp	497 mVpp	546 mVpp
200 MHz	70 mVpp	500 mVpp	508 mVpp	215 mVpp	224 mVpp	257 mVpp	250 mVpp	287 mVpp
300 MHz	---	348 mVpp	352 mVpp	28 mVpp	47 mVpp	49 mVpp	106 mVpp	72 mVpp
350 MHz	---	186 mVpp	240 mVpp	---	23 mVpp	---	72 mVpp	---
400 MHz	---	125 mVpp	180 mVpp	---	9 mVpp	---	53 mVpp	---
500 MHz	---	40 mVpp	---	---	---	---	26 mVpp	---
600 MHz	---	---	---	---	---	---	10 mVpp	---

EDIT 5-MAY-2019: Added SDS1104X-E software modded to 200MHz (SDS1104X-E*), Agilent 54622D and HP 54645D
EDIT 7-MAY-2019: Added SDS1202X-E (TAUTECH)

[hgjdwx]

The worst must be SDS1104X-E. The best?

[rf-loop]

I had the opportunity to test the Bandwidth on the following Digital Storage Oscilloscopes:

Siglent SDS1104X-E (1GSa/s 4 channels)
Keysight DSOX1102G (EDUX1002G uphacked to DSOX1102G) (2GSa/s 2+1 channels)
GWInstek GDS1054B (software license hacked to 300MHz) (1GSa/s 4 channels)
Micsig TO1104 (1GSa/s 4 channels)

I used an Agilent signal generator to output a 0dBm (632 mVpp) signal.

Agilent Signal Generator --> N to SMA connector --> SMA - SMA cable --> SMA to BNC --> BNC 50ohm passthrough --> DSO

Frequency	SDS1104X-E	DSOX1102G	GDS1054B	TO1104
10 MHz	645 mVpp	640 mVpp	630 mVpp	644 mVpp
100 MHz	500 mVpp	625 mVpp	485 mVpp	516 mVpp
200 MHz	70 mVpp	215 mVpp	224 mVpp	257 mVpp
300 MHz	---	28 mVpp	47 mVpp	49 mVpp
350 MHz	---	---	23 mVpp	---
400 MHz	---	---	9 mVpp	---

Because there was also modified Good Will 1054B compared to unmodified others so here is modified SDS1104X-E --> SDS1204X-E

SDS1104X-E (sw-mod to 1204X-E)

Generator Hewlett-Packard 8642B  (also checked with HP power meter that levels are well inside specs or much better)
Generator output: N-BNC adapter Suhner.
Cable: M17/084 (RG223) 150cm + Suhner BNC's
Scope input: Terminator Tektronix 011-0049-00


Generator output level set value 0.0dBm

Scope 100mV/div.

10 MHz  636mV
100MHz 561mV
200MHz 494mV
300MHz 342mV
350MHz 256mV
400MHz 171mV

here also member @lzyengineer measurements:SDS1104X-E original + mod
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg2326518/#msg2326518

And in other place can find my measurements also with both: 1104X-E and after mod.
Example in this thread after this message it start https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1624558/#msg1624558

But is it so that more is better? Not always.
All channels on 500MSa/s. SDS1104X-E have good DSP filter for reject aliasing!
~-36dB @ fNyquist
So, for avoid alias effects with measurements example with SDS1204X-E user need really know least some  basic fundamentals. - Mirror have also its backside.


But here bit more fair, both modified. Siglent for 200MHz and GoodWill for 300MHz

	SDS1104X-E mod	GDS1054B mod
	responds SDS1204X-E	with 300MHz license
Frequency	@rf-loop	@TK
10 MHz	636 mVpp	630 mVpp
100 MHz	561 mVpp	485 mVpp
200 MHz	494 mVpp	224 mVpp
300 MHz	342 mVpp	47 mVpp
350 MHz	256 mVpp	23 mVpp
400 MHz	171 mVpp	9 mVpp

[iMo]

A picture update..

[tautech]

A picture..

For SDS1204X-E..........a very accurate picture. 
From my own tests ~230 MHz @ -3dB.

[iMo]

Better the 1104X-E 200M should follow the shape of the 1104X-E 100M (or DSOX) above, but with the corner shifted to 200M, imho..

[tautech]

Better the 1104X-E 200M should follow the shape of the 1104X-E 100M (or DSOX) above, but with the corner shifted to 200M, imho..

I can do -3dB points for real SDS1104X-E and SDS1204X-E if you want........I have to do my pre-dispatch checks so no real problem to use my 510 MHz HP RF gen or SDG3021X SDG3032X.

[iMo]

Make it at all above freqs, I will update the picture then..

[rf-loop]

A picture..

For SDS1204X-E..........a very accurate picture. 
From my own tests ~230 MHz @ -3dB.

And previously (my previously made quality control's)  SDS1104X-E's  out from box: 
individual units I have measured have been between
~>120MHz   ---  ~<130MHz -3dB vs 1MHz

So this previous table is bit odd because TK did not measured any scope real -3dB point
Jump from 100 directly to 200MHz and then  between 300 and 400 also 350.
Proportionally jump from 100 to 200  is like jump from 300 to 600 or 200 to 400 so it is more than odd, specially because he test 50 - 100MHz nameplate scopes.  Peoples are some times ranting biased sellers but is it so that some times owners are even more biased or what to think. But yes, nice test.

[tautech]

A picture..

For SDS1204X-E..........a very accurate picture. 
From my own tests ~230 MHz @ -3dB.

And previously (my previously made quality control's)  SDS1104X-E's  out from box: 
individual units I have measured have been between
~>120MHz   ---  ~<130MHz -3dB vs 1MHz

So this previous table is bit odd because TK did not measured any scope real -3dB point
Jump from 100 directly to 200MHz and then  between 300 and 400 also 350.
Proportionally jump from 100 to 200  is like jump from 300 to 600 or 200 to 400 so it is more than odd, specially because he test 50 - 100MHz nameplate scopes.  Peoples are some times ranting biased sellers but is it so that some times owners are even more biased or what to think. But yes, nice test.

Agree 1000%
-3dB point is what matters however can be useful to know how high we can get stable performance albeit with significant attenuation. You saw me push SDS2202X-E to 520 MHz in another thread to make a point and I know the SDS1kX-E cannot be pushed that high.
Back to BW measurements..........

[TK]

I agree that the correct way is to measure -3dB but I saw several members asking for the max frequency that the scope (modded or original) can measure, so I tested to get to that value as soon as possible.

I will mod my sds1104x-e to sds1204x-e and edit my original results

[nctnico]

Still the graph in the 3rd reply post gives a good indication. Let's not get over obsessed with the precise -3dB point.

[tautech]

Scopes tested straight from box. Tests = warm up. 
SDS1104X-E
SDS1204X-E

Source SDG3021X (modded to 3.2 GHz model)
BNC cheap 1 GHz Siglent cable
Termination 50 ohm 1x Tek
Level 0dBm
Amplitude can be taken from OSD p-p measurements.

SDS1104X-E ....much as expected.
Images 27, 28, 29(200 MHz) and 30.(300 MHz)

SDS1204X-E..............pushed.  (Frequency counter becomes unstable @ 400+ MHz)
Images 31, 32, 33, 34, 36(350), 35(400), PNG450, PNG500, PNG550 and PNG600 

[rf-loop]

Still the graph in the 3rd reply post gives a good indication. Let's not get over obsessed with the precise -3dB point.

Yes. Only that it is quite common practice to measure -3dB point. 
But as you said "Let's not get over obsessed"

But because typically 100MHz scopes have BW limit slope between 100 and 200MHz it is bit rough only measure 100 and 200MHz point (and after then draw straight line between these points as @imo)

[rf-loop]

Scopes tested straight from box. Tests = warm up. 
SDS1104X-E
SDS1204X-E

Source SDG3021X (modded to 3.2 GHz model)
BNC cheap 1 GHz Siglent cable
Termination 50 ohm 1x Tek
Level 0dBm
Amplitude can be taken from OSD p-p measurements.

SDS1104X-E ....much as expected.
Images 27, 28, 29 and 30.

SDS1204X-E..............pushed.  (Frequency counter becomes unstable @ 400+ MHz)
Images 31, 32, 33, 34, 36(350), 35(400), PNG450, PNG500, PNG550 and PNG600 

What happen in image 29?
I expect roughly around  -6dB
Some thing is now wrong.

Oh... this trigger freq counter.. true signal is 200M..

[David Hess]

With manufacturers like Rigol producing DSOs with unspecified and variable full power bandwidth, these tests are close to useless without specifying the input sensitivity.  In the past this was not a consideration because the full power bandwidth was always higher than the bandwidth at any input sensitivity which was a result of operating the input stages over a much smaller signal range.

[tautech]

With manufacturers like Rigol producing DSOs with unspecified and variable full power bandwidth, these tests are close to useless without specifying the input sensitivity.  In the past this was not a consideration because the full power bandwidth was always higher than the bandwidth at any input sensitivity which was a result of operating the input stages over a much smaller signal range.

Specified in the OP and plainly visible in every screenshot.

[rf-loop]

With manufacturers like Rigol producing DSOs with unspecified and variable full power bandwidth, these tests are close to useless without specifying the input sensitivity.  In the past this was not a consideration because the full power bandwidth was always higher than the bandwidth at any input sensitivity which was a result of operating the input stages over a much smaller signal range.

Please explain.

If I measure freq response using (example) 5, 10, 20, 50, 100, 200, 500mV, etc/div and so on and show all these results. Or if I measure freq response for 100mV/div and show result. What is then wrong there. Can you say this response is not true with used setup.    Do you mean all this is bullshit. 

What is wrong. If it is wrong, please give explanation and instructions how to measure.
Or is it just only wrong because these are made in China.

I have repaired and calibrated several Tektronix old and also made frequency response measurements as told in Tek service manuals. Are these also useless.

Measured frequency response is just as result is in told situation. Or do you claim this result is wrong. If it is wrong please come here and show how it need measure so that also you are satisfied.

[David Hess]

With manufacturers like Rigol producing DSOs with unspecified and variable full power bandwidth, these tests are close to useless without specifying the input sensitivity.  In the past this was not a consideration because the full power bandwidth was always higher than the bandwidth at any input sensitivity which was a result of operating the input stages over a much smaller signal range.

Specified in the OP and plainly visible in every screenshot.

Signal level is not the input sensitivity and there are no screenshots associated with TK's post.

[nctnico]

With manufacturers like Rigol producing DSOs with unspecified and variable full power bandwidth, these tests are close to useless without specifying the input sensitivity.  In the past this was not a consideration because the full power bandwidth was always higher than the bandwidth at any input sensitivity which was a result of operating the input stages over a much smaller signal range.

Please explain.

If I measure freq response using (example) 5, 10, 20, 50, 100, 200, 500mV, etc/div and so on and show all these results. Or if I measure freq response for 100mV/div and show result. What is then wrong there. Can you say this response is not true with used setup.    Do you mean all this is bullshit. 

David has a good point. On the lower sensitivity settings there usually is a pre-amplifier which also affects the frequency response.  And the attenuators may behave differently at various frequencies as well. So if you want to do a really accurate frequency response graph then you'd need to do this using multiple v/div settings. But then again this is not a very useful exercise because near the bandwidth most (non-sinusoidal) signals will be distorted anyway.

[iMo]

A picture update..

[rf-loop]

With manufacturers like Rigol producing DSOs with unspecified and variable full power bandwidth, these tests are close to useless without specifying the input sensitivity.  In the past this was not a consideration because the full power bandwidth was always higher than the bandwidth at any input sensitivity which was a result of operating the input stages over a much smaller signal range.

Please explain.

If I measure freq response using (example) 5, 10, 20, 50, 100, 200, 500mV, etc/div and so on and show all these results. Or if I measure freq response for 100mV/div and show result. What is then wrong there. Can you say this response is not true with used setup.    Do you mean all this is bullshit. 

David has a good point. On the lower sensitivity settings there usually is a pre-amplifier which also affects the frequency response.  And the attenuators may behave differently at various frequencies as well. So if you want to do a really accurate

Yes, of course.
But still, if measure 100mV/div frequency respose using ok setup it IS 100mV/div frequency response in used setup and not bullshit. Bllshit is if after then claim it tell something what is not measured. Predictors and clairvoyants are a separate issue.
 
But, of course, and who even think other, it is NOT whole scope frequency response over whole signal level range with all sensitivity settings.  If it is measured using one V/div range with one signal input level and explained setup it is response just with it and nothing else. This is true what ever we test. Result tell only how it is with used setup and nothing else.

[TK]

The worst must be SDS1104X-E. The best?

I updated the original test to include the SDS1104X-E software hacked to 200MHz, so far it is the 2nd best on my test

[rf-loop]

A picture update..

Wrong, there is err...afaik.

Look tautech 1104X-E result. Where from you get 150MHz data point?
This around -18dB point is 200MHz!

[TK]

A picture update..

A picture is worth a thousand words (or numbers in this case)