R&S RTB2004 - Function Generator - Sweep option is badly broken!

[javadesigner]

The sweep in the R&S RTB2004 Function Generator (installed option) seems to be badly broken.

Let's start with the specs from the datasheet.

1. start frequency:  1 Hz to 25 MHz
2. stop frequency: 1 Hz to 25 MHz
3. sweep time: 1 ms to 10 s
4. sweep type: linear, logarithmic, triangle 1 to 1024

1)
Ok, so the sweep time can be set from 1 ms to 10 seconds (item 3), simple right ?

Well, no! The Sweep time always shows a weird arbitrary 4.979672 min time (!!) and I cannot enter anything less than this.

If I enter 4 seconds for example, or 10 ms or whatever, I get an error.

This is a glaringly bad bug #1. See attached screenshot.


2)
If I enter a Start frequency of 1 Hz, Stop Frequency of 3 Hz and a Sweep time of 5 sec - there is NO sweep.

The sweep does not work at all. .

This is glaringly bad bug #2. See attached screenshot.

I'm just gobsmacked at this. This feature is as simple and basic as it gets! Either I don't understand what I am doing (quite possible) or the R&S RTB2004 'scope is completely utterly broken - I'm discovering too many bugs that can hardly even exist in $100 scope, let alone $4000 like the RTB2004..

[2N3055]

Well, let's think about it for a second (pun intended).

You want to sweep from 1 Hz to 3 Hz. Let's simplify. So you need at least one period at 1 Hz, then one transition to 2 Hz, one period at 2 Hz, then transition to 3 Hz, and one period at 3Hz.

Since one period at 1 Hz take one second, that is : 1s+1s +0,5s+0.5s +0,3s+0,3s... 3,6seconds to do move that way.
That is, if they simply jump from 1 to 2 to 3 Hz like that.
Which they don't, but they gradually stretch phase of sinewave, so 4-5 seconds minimum to make transition from 1Hz to 3 Hz sounds about right. In fact, that is quite short too, for that transition.

So minimum time of 4-5 s for that sweep is not a bug, that is just fine.

As for no 2, try setting sweep to 10 seconds to see if there are any changes there. Also try different start and stop frequencies to find what does work and what doesn't. Where exactly things break?
When dealing with things like this, you need to investigate from all sides, gather details, and make sure what you saying is right and that you can demonstrate the problem if there is any.
Many times by analyzing you end up realizing there is no bug, but sometimes things simply work differently than what you think they should. And you learned something. Also win.

OTOH, but there I see a bigger problem you didn't notice.
10 second maximum for sweep is not nearly long enough. That is a problem, exactly because of what was said above.
If you want to make a sweep in audio range from 10 Hz to 20 Khz, more appropriate time would be 100 seconds for sweep.

Regards,

[javadesigner]

Hi 2N3055 - appreciate the reply!

Well, let's think about it for a second (pun intended).

You want to sweep from 1 Hz to 3 Hz. Let's simplify. So you need at least one period at 1 Hz, then one transition to 2 Hz, one period at 2 Hz, then transition to 3 Hz, and one period at 3Hz.

Since one period at 1 Hz take one second, that is : 1s+1s +0,5s+0.5s +0,3s+0,3s... 3,6seconds to do move that way.
That is, if they simply jump from 1 to 2 to 3 Hz like that.
Which they don't, but they gradually stretch phase of sinewave, so 4-5 seconds minimum to make transition from 1Hz to 3 Hz sounds about right. In fact, that is quite short too, for that transition.

So minimum time of 4-5 s for that sweep is not a bug, that is just fine.

In that case though, the (STOP frequency - START frequency) gap would be relevant, right ?

On this scope, if the START is 1 Hz, then this same limitation happens regardless of the gap between START and STOP frequency! Specifically:

START: 1 Hz

STOP:  3Hz   
|START-STOP|=2hz

STOP: 100 Hz         
|START-STOP|=99hz
..etc..

No matter, if the difference is 2 Hz, or 99 Hz or whatever, the exact same dialog box with exact same min time: 4.97967

Please see attached screenshot.

So...If the scope is gradually stretching between the START and STOP frequencies, then why does the 4-5 second period not change at all, regardless of STOP frequency ? (also regardless of horiz. scale btw)

As for no 2, try setting sweep to 10 seconds to see if there are any changes there. Also try different start and stop frequencies to find what does work and what doesn't. Where exactly things break?
[..]

It does not sweep at all - no changes - no sweep - at 10 seconds either.

Please see attached screenshot.

Edit: I let this run for about a several minutes and saw no change in the waveforms. Measurement shows a constant unchanging frequency of 2 Hz.

No sweep! Another screenshot attached.

Many times by analyzing you end up realizing there is no bug, but sometimes things simply work differently than what you think they should. And you learned something. Also win.

True, true.

OTOH, but there I see a bigger problem you didn't notice.
10 second maximum for sweep is not nearly long enough. That is a problem, exactly because of what was said above.
If you want to make a sweep in audio range from 10 Hz to 20 Khz, more appropriate time would be 100 seconds for sweep.

I agree - this is a big problem and makes this frequency generator a toy/useless.

I also don't see any options for:
(1) dwell time
(2) separate return time.
(3) frequency list (steps through and dwell on a list of frequencies, see how a a device is responding at specific ones).

This isn't in the spec sheet but it would have been nice to have.

[2N3055]

Well,

that is nice work.. That should be enough for R&S to have something to work with.

There is a user here, Rich (Rich@RohdeScopesUSA) which works for R&S.
Shoot him a PM about this.
He's outstanding guy and helped to many people here, even those from Europe, although he works for USA office. 
He will know what to do.

Take care and stay safe..

Sinisa

[Rydda]

Hi
I don't think you can call it  broken.

The minimum sweep time is set by the start frequency. Eg 2Hz will give 2.5 sec,  5Hz will give 998mS minimum time etc

Your problem with the 1Hz - 3Hz sweep may be a bug, but I think you will get your sweep if you change the stop frequency to 3.01 Hz. 
 
/rydda

[javadesigner]

Hi
I don't think you can call it  broken.
[..]
Your problem with the 1Hz - 3Hz sweep may be a bug, but I think you will get your sweep if you change the stop frequency to 3.01 Hz. 
 /rydda

3.01 does "nudge" it to start sweeping again.

Here is something interesting: try putting 2 for stop and see what happens. Well actually, you can't enter 2 because if you do, the scope changes it to 2.01 on its own!! (this happens when I enter 5 as well, scope changes it to 5.01).

And yes, at 2.01, it stops sweeping yet again and displays a constant unchanging waveform with a frequency of 1.5Hz. (watched over several minutes, tried 5 and 7 second sweep times both, different horiz. time scales, etc).

[Rich@RohdeScopesUSA]

Our R&D team has looked in to this - there is an issue with sweeps from 1-3Hz that will be fixed.

For the not rounded values (e.g. when it changes 2 to 2.01), we are not able to set every frequency due to technical reasons, so the R&D team chose values that were possible (like 2.01).

-Rich