new killer scope in town - a true game changer from R&S - RTB2002 & RTB2004

[nctnico]

Thanks R&S for letting me spend >2k$ on a scope that's broken by design! Luckily it was only 2k$ and not 8k$. I was aware that early adopting will give me buggy firmware, but this is quite a substantial hardware issue. Let's see how annoying this will be when using it for real. R&S care to comment?

I don't think it is broken by design.  If you watch the whole video, all the scopes does the same.

But not in the same amount. The Keysight and Siglent (GW Instek too BTW) need a really hard wack to distort the signal.

[EEVblog]

Now have a picture of people all across the world currently whacking their scopes, DMMS, PSU's, SMU's ....
Looking forward to future teardowns looking at whack-a-scope attenuation and for cushion mounted mainboards and connectors

The first manufacturer to come up with an isolated front end daughter board held in place by rubber bands will win the market I tell you.
A killing will also be had on 3rd party rubber band replacement kits in 10 years time.
Just don't try to patent it, I have prior art more than a decade back.
I also have prior art on using cotton wool, antistatic foam, sorbothane, and feminine hygine products (long story).

Pro tip, it's also possible to get helpful isolation using various jazzy PCB routing techniques.

Does whack-a-scope also cause peaks on any signal generator output? ... or DMMs and Power Supplies?
That could cause fun on SMU tests if they suffer too :p

The volt and frequency nuts will tell you all about that.

[EEVblog]

Thanks R&S for letting me spend >2k$ on a scope that's broken by design! Luckily it was only 2k$ and not 8k$. I was aware that early adopting will give me buggy firmware, but this is quite a substantial hardware issue. Let's see how annoying this will be when using it for real. R&S care to comment?

I don't think it is broken by design.  If you watch the whole video, all the scopes does the same.  You can avoid the full touchscreen interface and operate it remotely when you need to do really sensitive measurements.

Umm, why should you have to do that?
It's a pretty reasonable expectation to be able to use the touch screen feature on a touch screen scope without worrying whether it's going to impact your signal measurement.
I found this issue in normal use.

[EEVblog]

So we have to consider any scope with MLCCs in the channel frontend being cheap? The manufactures should choose types which are less susceptible to microphonics at least.

Or mount them on their side  (it works)
Not exactly production friendly though. Although if you want to really annoy your assembler, request it 

[Octane]

Thanks R&S for letting me spend >2k$ on a scope that's broken by design! Luckily it was only 2k$ and not 8k$. I was aware that early adopting will give me buggy firmware, but this is quite a substantial hardware issue. Let's see how annoying this will be when using it for real. R&S care to comment?

I don't think it is broken by design.  If you watch the whole video, all the scopes does the same.  You can avoid the full touchscreen interface and operate it remotely when you need to do really sensitive measurements.

So if you can't use a highly sensitive measurement gear as is it supposed to be used,but you need to have a workaround (in this case Remote Control) it is not broken by design? Maybe then they should have advertised this bug as a feature...   
But let's see how bad it really is for me in normal use. But if Dave finds that in the first few minutes it is probably not just a tiny thing....

[ci11]

It was blindly obvious to me in the first few minutes of using it, so I'd say it's a problem.
Tapping the BNC's, ok, every scope has a direct coupling at that point, fair enough.
Whacking the top of the unit, ok, don't do that, fair enough.
But a much touted touch screen being able to generate many division level impulses when tapped is unacceptable, even if some units aren't as susceptible.

Thanks Dave - for the quick 20Hz look over and then the tap test. This is interesting.

The reason for the 20Hz test is that it matters to audio. And the mechanical coupling issues is equally relevant because the excitation is well within audio range. Even if the user doesn't pound or tap on the case, the BNC does wiggle from time to time when the probe lead is moved. And that will find its way into the waveform. We can see from the HMO1202 teardown video that the BNC's are directly soldered onto the PCB, and if this construction is repeated on the RTB, this is clearly something to think about.

I have a theory why this may not have been caught by R&S and the others. It is simply that these scope at 70MHz - 300MHz BW were designed to compete well in signal diagnosis and bus decoding, and audio is not part of their consideration. A microphone preamp designer would never miss these design and construction considerations.

I can hear it now in München - "Es war nicht im Lastenheft" (it was not in the specs).

[Zbig]

Dave, now you know why the ASIC is SAD

Ok, enough bad puns for today from my side. Pinky promise.

[JoHr]

Thanks R&S for letting me spend >2k$ on a scope that's broken by design! Luckily it was only 2k$ and not 8k$. I was aware that early adopting will give me buggy firmware, but this is quite a substantial hardware issue. Let's see how annoying this will be when using it for real. R&S care to comment?

This is how 8k-20k€ equipment is working ... while hitting the inputs with fingernails. Slapping the chassis produces this as well but not that high, its mostly covered by its noise.
I´ve to hit my RTB unit pretty hard before its doing this.

In normal use there s no problem with my one.

[EEVblog]

If you want this problem fixed, just talk to marketing. They'll insist upon the world's first "Low Vibration Front End".
The sales people will love because they'll be able to do side-by-side demos tapping on the competitor BNCs and showing how crap they are!
I'm serious.

LowVib^TM 
or the even better NoVib^TM

[Octane]

Thanks R&S for letting me spend >2k$ on a scope that's broken by design! Luckily it was only 2k$ and not 8k$. I was aware that early adopting will give me buggy firmware, but this is quite a substantial hardware issue. Let's see how annoying this will be when using it for real. R&S care to comment?

This is how 8k-20k€ equipment is working ... while hitting the inputs with fingernails. Slapping the chassis produces this as well but not that high, its mostly covered by its noise.
I´ve to hit my RTB unit pretty hard before its doing this.

In normal use there s no problem with my one.

Thanks for pointing that out. It is clear that any scope is more or les susceptible for this. In case of the RTB it worries me that using the touchscreen as intended could lead to problems.
I would have thought a venerable company like R&S has better performance in terms of unwanted signals.

[JoHr]

Interesting is that the heavy scopes are doing this as well. Even if they are built like a Pzkpfw VI.

@Dave which ones have chassis- mounted BNCs?

[ebastler]

If you want this problem fixed, just talk to marketing. They'll insist upon the world's first "Low Vibration Front End".
The sales people will love because they'll be able to do side-by-side demos tapping on the competitor BNCs and showing how crap they are!
I'm serious.

LowVib^TM 
or the even better NoVib^TM

This whole topic makes me slightly nostalgic...
Microphonic tubes, anyone? 

[ci11]

I just tried the 2 of my scopes, one being quite expensive and the other at the "budget" end: a 400MHz LeCroy and a 60MHz Tek. Using the rubberized grip of my PB "Swiss Grip" #2 screwdriver, both yielded the same results: wacking on the case, nothing, even when it's a pretty hard wack; right on the BNC, yes, but it would have to be a deliberate and moderately hard wack.

[vokars]

Very interesting tests. I think the case-hitting sensitivity is a question of high or low quaility but the "touchscreen normal usage" sensitivity could be a warranty issue. R&S should fix that.

As already discussed here these electromechanical issues are well known in audio technology. And R&S sells audio analyzers. I recommend that the R&S audio developers have a look at the RTB2000.

[JPortici]

The first manufacturer to come up with an isolated front end daughter board held in place by rubber bands will win the market I tell you.
A killing will also be had on 3rd party rubber band replacement kits in 10 years time.
Just don't try to patent it, I have prior art more than a decade back.
I also have prior art on using cotton wool, antistatic foam, sorbothane, and feminine hygine products (long story).

Pro tip, it's also possible to get helpful isolation using various jazzy PCB routing techniques.

FWIW, Tek TPS2014 (1st series) is affected by wackyness. never had to open it but i don't think the acquisition boards are held in place by rubbers
infact it's pretty much worse than anyone in the video

[JPortici]

Or maybe you meant mechanically isolated?
Still, scope shot is always welcome

Knocked the enclosure in the picture, knocking BNC is ~ 3 div wide instead. Still, it's VERY much affected

[thanasisk]

Gotta love the age of the internet, enables you to make informed buying decisions !

Shocking discoveries by the way   

[GlowingGhoul]

Well this is very disappointing!

On the 16th March I paid for and placed an order for a Rohde & Schwarz RTB2K-COM4 package with TEquipment.
This was before any message appeared on their web page stating "SHIPPING TO USA ADDRESSES ONLY!".

The order was accepted and the next day I spoke to lady from TEquipment who confirmed my address and payment.

A couple of days latter I had an email conversations with TEquipment about shipping dates and was assured that I would receive a unit. In fact it was stressed that I was likely to receive one with much sooner than the quoted 3rd May.

Then just now I received a email saying:

"Unfortunately due to our contractual obligations with R & S we are unable to export any of the items sold by them and we will have to cancel the order you placed with us."

TEquipment then refunded my money that they had kindly kept in their bank for me over the last 13 days.

This whole operation by Rohde & Schwarz to only offer this promotion to the USA and Canada really stinks.

Grrr.

Based on what I'm reading regarding the touchscreen induced anomalies consider yourself lucky to have dodged this bullet.

[ebastler]

Well this is very disappointing!
[...]
This whole operation by Rohde & Schwarz to only offer this promotion to the USA and Canada really stinks.
Grrr.

Maybe you can buy Octane's scope -- he seems to be regretting his purchase now (see post #706 in this thread). Win-win! 

[TK]

So we have to consider any scope with MLCCs in the channel frontend being cheap? The manufactures should choose types which are less susceptible to microphonics at least.

Or mount them on their side  (it works)
Not exactly production friendly though. Although if you want to really annoy your assembler, request it 

Sorry if I am asking a dumb question... If the PCB is mounted vertical to the base of the scope (like the Keysight 1000X series), isn't it equivalent to having the multilayer caps mounted on their side?

[Octane]

Well this is very disappointing!
[...]
This whole operation by Rohde & Schwarz to only offer this promotion to the USA and Canada really stinks.
Grrr.

Maybe you can buy Octane's scope -- he seems to be regretting his purchase now (see post #706 in this thread). Win-win! 

Let's first see how it behaves once I get it. I'm building up my lab here and need a scope. And since it's likely not a problem with higher V/Div I probably can live with it. Although it is disappointing to get something from a higher end manufacturer with such a stupid issue.

[pm.llb]

Knocked the enclosure in the picture, knocking BNC is ~ 3 div wide instead. Still, it's VERY much affected

My old Tek TDS2024 have been  beaten for 5 min and nothing.
It seems that I can't join  to this privilaged club of "mikeScopes".

Anyway. tequipment has canceled my European order after they
promissed delivery on 3th May. I saw new 7 units ready to sell at their's page.
Lucky ( or not considering bugs found) those US folks who will get another chance ...

Now I have extra 2k$ to spend on the equipment:
Let's play and try to collect equal ( functionality)  or better ( price) equivalent
equipment:

Goal:
Budget: 2080USD+
Components ( could be separate):
DSO or MSO 4 channel , 200MHz+ , FFT ( for small frequecy filter graphing, EMI detection ),  I2C, Parallel, SPI, Serial decoding, ( CAN,LIN will be  +)
Arbitrary generator for at least 25MHz+
Not Rhode & Schwarz ( start to do not like this company - guess why )
Cherry on a cake - DMM 5+ digit LoZ, ESR ( covers multimeter RTB function)

Priority:
1) Quality ( mid is enough) 
2) price ( have to save some money for a beer)
3) features ( everyone likes to play)

Types:
DSO/MSO : Rigol 1054Z ( what options to buy, what options to hack?)
Arbitrary generator: ?
DMM: SANWA PC5000A ( almost the same as brymen 857)

Could you help with con/pros other proposals ?
Is there something left for descend 1GHz+ spectrum analyser ?

[Octane]

So we have to consider any scope with MLCCs in the channel frontend being cheap? The manufactures should choose types which are less susceptible to microphonics at least.

Or mount them on their side  (it works)
Not exactly production friendly though. Although if you want to really annoy your assembler, request it 

Sorry if I am asking a dumb question... If the PCB is mounted vertical to the base of the scope (like the Keysight 1000X series), isn't it equivalent to having the multilayer caps mounted on their side?

I would guess it depends on the relative orientation between the cap and the PCB. Not the overall orientation. If you excite flexing vibrations in the PCB then bending a "flat" cap is easier then a cap that's higher then wide. See beams with different aspect ratios...

[vokars]

So we have to consider any scope with MLCCs in the channel frontend being cheap? The manufactures should choose types which are less susceptible to microphonics at least.

Or mount them on their side  (it works)
Not exactly production friendly though. Although if you want to really annoy your assembler, request it 

Sorry if I am asking a dumb question... If the PCB is mounted vertical to the base of the scope (like the Keysight 1000X series), isn't it equivalent to having the multilayer caps mounted on their side?

I would guess it depends on the relative orientation between the cap and the PCB. Not the overall orientation.

Capacity depends on the distance between the layers. Stress on a capacitor therefore has different effects if applicated parallel or perpendicular to the layer-area. Also as Dave mentioned piezo effects of the ceramic depend on the stress direction.

This is like a condensor or piezo mic. The direction of the sound wave must have a perpendicular component to the capacitor area for changing the capacitor layer distance.

Addendum: To clarify: That means the orientation of the capacitor on the PCB is relevant because the PCB transfers the stress to the capacitor.

[TK]

So we have to consider any scope with MLCCs in the channel frontend being cheap? The manufactures should choose types which are less susceptible to microphonics at least.

Or mount them on their side  (it works)
Not exactly production friendly though. Although if you want to really annoy your assembler, request it 

Sorry if I am asking a dumb question... If the PCB is mounted vertical to the base of the scope (like the Keysight 1000X series), isn't it equivalent to having the multilayer caps mounted on their side?

I would guess it depends on the relative orientation between the cap and the PCB. Not the overall orientation. If you excite flexing vibrations in the PCB then bending a "flat" cap is easier then a cap that's higher then wide. See beams with different aspect ratios...

I am testing my EDUX1002G scope and it is affected by the Wack-a-Scope feature.  But if I connect a 50ohm BNC terminator in CH1, then it is Wack-immune.  If I turn both CH1 (still with 50 ohm terminator at input) and CH2 ON, then I Wack the scope and I only see the issue on CH2.