Author Topic: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?  (Read 1139 times)

0 Members and 1 Guest are viewing this topic.

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
A question for owners of a Rohde & Schwarz RTB2000 with 70MHz bandwidth and a Siglent SDS2000x Plus with 100MHz bandwidth.
Have you ever measured the cut-off frequency at -3db ?
Is this artificial limit exactly at 70MHz/100MHz ?

Regards,
Frank
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 21413
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
A question for owners of a Rohde & Schwarz RTB2000 with 70MHz bandwidth and a Siglent SDS2000x Plus with 100MHz bandwidth.
Have you ever measured the cut-off frequency at -3db ?
Is this artificial limit exactly at 70MHz/100MHz ?

Regards,
Frank
SDS2104X Plus -3dB is ~185 MHz. Comes with 200 MHz probes.
Triple checked with 3 signal sources.

Edit
Actual screenshot:
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2958288/#msg2958288
« Last Edit: November 02, 2020, 06:54:14 pm by tautech »
Avid Rabid Hobbyist
 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
This is certainly very pleasing for the users of the Siglent with the 100MHz variant.
Rohde & Schwarz will probably not be as generous, I guess.

Thanks tautech
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3560
  • Country: cn
  • Born with DLL21 in hand
A question for owners of a Rohde & Schwarz RTB2000 with 70MHz bandwidth and a Siglent SDS2000x Plus with 100MHz bandwidth.
Have you ever measured the cut-off frequency at -3db ?
Is this artificial limit exactly at 70MHz/100MHz ?

Regards,
Frank
SDS2104X Plus -3dB is ~185 MHz. Comes with 200 MHz probes.
Triple checked with 3 signal sources.

Edit
Actual screenshot:
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2958288/#msg2958288

This is bit bad situation for start this from me but still I want rise one common question about how we measure oscilloscope frequency response and specially -3dB point what is normal.

This is also NOT any kind of complaint about your measurement, AFAIK we do it usually just using same principle what perhaps almost all do.

Usually we measure it using RF signal generator what have enough good flatness in its output connector. Then cable to scope and there scope internal 50ohm or external feed thru terminator what both are bit bad. It is simple method and if done carefully and well it may give somehow reliable result. About how to do it there can write whole book about things around this. It is not accurate method due to generator flatness errors and signal pathway impedance matching errors. But widely used and ok for normal purposes. But also this is not question.

How ever we do this test setup and so on,  we use same principle. We go more and more high freq and look oscilloscope response and where level drop 3dB related to our some reference point. If we continue rising freq we are sooner or later -6dB and -10dB and so on...  but principle is we are looking attenuated signal and watch how much it is attenuated until attenuation is value what we are interested, example -3dB or -6dB

Select some zero dB reference level say example some level what give 6div height signal and  example with 10k to 1MHz. (usually reference is more low than 1MHz)
Okay, we have now this level out from generator and after generator fine adjust we see exactly 6div peak-peak sine.
What we do next... if we are looking -3dB point. After then we start rise freq and go up until first time we see signal level on screen drops perfectly -3dB under reference level (6div)  or more low depending what we are looking.


But then, lets go bit more high freg (if there is enough room before Nyquist*0.8 or better if below Nyq*0.7). Lets think we find -12dB point. Signal is now attenuated 12dB...
Now, keep this frequency and rise generator output level now 12dB. Simple thing and of course we see now original reference level... 6div height signal. Or... after then we scratch our head and think what hell there happen... and why.
It not enough clear effect you can test in -20dB point or even more high frequency...  more high more weird.
Also it can of course do with all, -3, -6, -9, -12, and so on...




Some may think is it same. With 3dB mostly result is nearly same but not always...in some scopes difference may start even with -3dB point if reference level is least 6div. (this depends also what is scope display vertical and front end FS) 
If you want find -10dB point you may find big difference... if you go to -30dB point. You may get "even "no result" situation...    it is least good to try even once and then scratch head and know bit more about fact that all can see in scope screen is in some point unknown lies mixed with unknown truths.

Now, this other principle for test is:
Set signal level perfectly 6div height with some low freq just for reference point.
After then rise freq until -3dB drop. Rise generator level 3dB. Is level now 6div height. If not, drop frequency until signal is back to this 6div height.
Usually in 3dB point do not differ so much but if go higher frequencies difference is big or 6div height can not even reach how much ever you rise signal.

I do not remember if this method was in some old HP or Tek some oscilloscope service manual... or where... select reference level and freq, then rise test signal 3dB and rise freq until displayed signal drop back to this defined reference level.

If practice and theory is not equal it tells that used application of theory is wrong or the theory itself is wrong.
-
Harmony OS
 
The following users thanked this post: TurboTom

Online tautech

  • Super Contributor
  • ***
  • Posts: 21413
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Thanks for your comments rf-loop.

When I did the -3dB test on SDS2104X Plus I did not believe the first result so another RF gen was used.....and then another for absolute confidence that the results were real and IIRC some 7 MHz variance was observed between the 3 sig gens, SDG6022X, SSG3021X and old 500 MHz HP8654B.
So the ~185 MHz was near the middle of the 3 sig gen results albeit none were a properly leveled source.
Even my SDS5054X was dragged out to ensure no mistakes were made in what I measured.

Yes I have seen your sweeps and in time I will experiment with BW sweeps as I suspect with a little application they will be easy to do but now I wait for SDG6022X to be back in stock again as my demo was sold to a customer that needed one quickly.
Avid Rabid Hobbyist
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 3560
  • Country: cn
  • Born with DLL21 in hand
Thanks for your comments rf-loop.

When I did the -3dB test on SDS2104X Plus I did not believe the first result so another RF gen was used.....and then another for absolute confidence that the results were real and IIRC some 7 MHz variance was observed between the 3 sig gens, SDG6022X, SSG3021X and old 500 MHz HP8654B.
So the ~185 MHz was near the middle of the 3 sig gen results albeit none were a properly leveled source.
Even my SDS5054X was dragged out to ensure no mistakes were made in what I measured.

Yes I have seen your sweeps and in time I will experiment with BW sweeps as I suspect with a little application they will be easy to do but now I wait for SDG6022X to be back in stock again as my demo was sold to a customer that needed one quickly.

Sweeps and sweep yes..  sad I can not now do any examples about this. And do not worry, in same boat is all manufacturers, some more badly some less.


But just for your own fun... try... go as my previous text... take 6div ref level... go to as far as it drops  least over 10dB... perhaps more as far you can go with quite somehow trusted level can detect and measure from screen You can do it without trigger, trigger do not matter here anything. After then look how many dB it is down... example if its now 20dB down or even more do not change freq now. Only rise generator level 20dB or this dB amount what you see level have dropped and look how close to ref level it now rise. Think why -20 + 20 is not zero. ;)
Of course if input circuits are ideal ones it rise back to original reference level 6div... but... think what you see. How much you need rise level before you see 6div height.
If practice and theory is not equal it tells that used application of theory is wrong or the theory itself is wrong.
-
Harmony OS
 

Offline TurboTom

  • Super Contributor
  • ***
  • Posts: 1025
  • Country: de
I did this quick test as per @rf-loop's description with an MSO2072A-S and an SDG6022X, both "improved" to the max. Some time ago, I checked the generator for level accuracy with a HP power meter and it performed very well, within 1/10th of its spec. I used a level of -14dBm (approx. 126mVpp into 50R)

Due to the high bandwidth of the scope, I couldn't test any further than to the -6dB point which appars to be at ~450MHz. The funny thing: Nothing out-of-the-ordinary was to be observed -- when I increased the level by 6dB (to -8dBm), the displayed level on the scope screen went back right to where it was before, ~126mVpp. To do testing at higher dropoff levels (frequencies), I've got to pull out the RF signal generator... Maybe I'll try this later. Yet, at these frequencies, the trigger of the scope gets somewhat jittery, so I'll have to use single-shot mode and possibly cursor measurement to get accurate readings. @450MHz, it wouldn't make any difference if I run the scope single or dual channel (i.e. 2GSa/s vs. 1GSa/s) but I expect that any higher than that, the proximity of the nyquist frequency will have an effect.

Edit: Did some testing with the RF signal generator @ 12dB level reduction (which is reached at approx. 620MHz), and again didn't find any discrepancy when I increased the generator output level by 12dB (started at 0dBm, 10MHz). This may be related to the LMH6518 VGA that Rigol chose for the front-end. This amplifier is specified with a -3dB point of 900MHz, so maybe the bandwidth limit is arranged by classic R-C circuitry which is linear instead of dV/dt limited active circuitry.
« Last Edit: November 03, 2020, 04:23:05 pm by TurboTom »
 
The following users thanked this post: rf-loop, 2N3055

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
thanks for your detailed answers.   :)
Is there anybody else here who could measure the cut-off frequency of an 70MHz RTB2000 ?
« Last Edit: November 03, 2020, 07:59:59 pm by Frank_MV »
 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
nobody ?
 

Online Kean

  • Supporter
  • ****
  • Posts: 1225
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
I can't imagine too many people paying for an RTB2000 and only going with the 70MHz base version.  At least from the group of people who frequent this forum.
 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #10 on: November 04, 2020, 09:59:03 am »
Because 70MHz is too little bandwidth for a variety of measuring tasks  ?
Or too expensive for only 70MHz  ?

« Last Edit: November 04, 2020, 10:04:44 am by Frank_MV »
 

Online Kean

  • Supporter
  • ****
  • Posts: 1225
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #11 on: November 04, 2020, 02:12:49 pm »
Because 70MHz is too little bandwidth for a variety of measuring tasks  ?
Or too expensive for only 70MHz  ?
I'd say both, but mostly the latter.
RTB is not generally a scope for beginners, and those choosing an R&S scope are more likely to have higher BW requirements even if only occasionally, or larger budgets where the higher end models are a no-brainer.
The bundle deals were also really good value, and I believe they included BW upgrades.  I am not sure what was on offer elsewhere in the world though (vs AU).
 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #12 on: November 04, 2020, 02:28:08 pm »
Okay, thanks.
I have an offer for a new (4-channel) RTB2004 with all software options like i2c, SPI, UART decoding, Bode-Plot, Arbitary Generator for 2000€ ~ 2300 USD.
But bandwidth only 70MHz.
An upgrade to 100MHz I could afford financially, but not more.
Or should I better decide for a much cheaper Siglent SDS 2000x Plus, where I can get a BW hack as well ?
Very hard decision

" I am not sure what was on offer elsewhere in the world though (vs AU)"
>>Here in germany there is a "R&S Full Package RTB" offer/promotion at the moment.
    All software options including MSO and 300MHz bandwidth for 4000€ ~ 6500AUD.
    But for me a bit too expensive.
« Last Edit: November 04, 2020, 05:21:16 pm by Frank_MV »
 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #13 on: November 05, 2020, 08:07:41 am »
.
 

Offline switchabl

  • Regular Contributor
  • *
  • Posts: 71
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #14 on: November 05, 2020, 12:28:29 pm »
I have an offer for a new (4-channel) RTB2004 with all software options like i2c, SPI, UART decoding, Bode-Plot, Arbitary Generator for 2000€ ~ 2300 USD.
But bandwidth only 70MHz.
An upgrade to 100MHz I could afford financially, but not more.

Is that the one with the education discount? If you qualify for that, you could try asking for a discount on the bandwidth upgrade if you haven't already.
 

Online Kean

  • Supporter
  • ****
  • Posts: 1225
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #15 on: November 05, 2020, 01:31:15 pm »
I have no experience with the Siglent scopes, so I cannot compare.
Do you have any need for more than 70MHz?
You can always upgrade the BW later, but it will likely be more expensive (unless there is a future hack).

Oh, and I paid significantly more than 6500AUD when I got the bundle.  The cost here is not just a factor of exchange rate, but also additional costs to support the local sales & support organisation.
 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #16 on: November 05, 2020, 03:28:30 pm »
I have an offer for a new (4-channel) RTB2004 with all software options like i2c, SPI, UART decoding, Bode-Plot, Arbitary Generator for 2000€ ~ 2300 USD.
But bandwidth only 70MHz.
An upgrade to 100MHz I could afford financially, but not more.

Is that the one with the education discount? If you qualify for that, you could try asking for a discount on the bandwidth upgrade if you haven't already.

Yes, it is the Education version.
Thanks for the tip, do you have any experience with the discount at Rohde & Schwarz ?
Would this also be possible for the Logic Analyser Option ?

 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #17 on: November 05, 2020, 03:33:31 pm »
I have no experience with the Siglent scopes, so I cannot compare.
Do you have any need for more than 70MHz?
You can always upgrade the BW later, but it will likely be more expensive (unless there is a future hack).

Oh, and I paid significantly more than 6500AUD when I got the bundle.  The cost here is not just a factor of exchange rate, but also additional costs to support the local sales & support organisation.

Even more than 6500AUD, that is quite a lot.
I mean, 70MHz is currently enough for me, for sine signals it is enough, for non-sine signals I have to wait and see what I will encounter in the future
 

Offline switchabl

  • Regular Contributor
  • *
  • Posts: 71
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #18 on: November 05, 2020, 05:24:33 pm »
Yes, it is the Education version.
Thanks for the tip, do you have any experience with the discount at Rohde & Schwarz ?
Would this also be possible for the Logic Analyser Option ?

Not specifically, sorry. But it definitely won't hurt to contact either R&S sales or your distributor. Tell them you are really interested, but the offer you have doesn't quite fit your needs. I mean, it is their low-end line, so not sure how much flexibility there will be. The full package may still be out of your budget even with an additional edu discount. But in any case, I would always ask for a quote, rather than buying options at list price.
 

Offline Frank_MV

  • Regular Contributor
  • *
  • Posts: 99
  • Country: de
Re: Rohde & Scxhwarz RTB2000, Siglent SDS2000X Plus cut-off frequency at -3db ?
« Reply #19 on: November 05, 2020, 05:35:17 pm »
Yes, you are right about asking for offers.
Thanks
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf