Magnova oscilloscope

[2N3055]

I'm also noticing something I call 'trigger lensing'. Around the trigger point, the trace gets thinner but beyond that, it get much thicker. I don't think this is a true representation of the signal as noise around the trigger point makes the trace move left/right more than it should. A trigger condition should be targeted at an edge of a signal (IOW a section of the signal) around the trigger level crossing which to some extend (=the trigger hysteresis boundaries) cancels noise, not just the single point where the signal crosses the trigger level. Maybe this can be adjusted with a trigger hysteresis setting.

Thats interesting. Is there a paper about that topic to dive a bit deeper?
I saw that issue on various oscilloscopes but never saw that as an issue.
I thought its something related to the sample rate / waveforms per second update rate ratio and never thought about it in detail.

But I put our mighty Tek Series 5 scope at the office though the same challenge and even here its visible (but of course at a much lower extend):

There is no paper. That is artefact stemming from noise and large magnification of that portion. It is normal and cannot be avoided unless you have no noise at all.

[nctnico]

I'm also noticing something I call 'trigger lensing'. Around the trigger point, the trace gets thinner but beyond that, it get much thicker. I don't think this is a true representation of the signal as noise around the trigger point makes the trace move left/right more than it should. A trigger condition should be targeted at an edge of a signal (IOW a section of the signal) around the trigger level crossing which to some extend (=the trigger hysteresis boundaries) cancels noise, not just the single point where the signal crosses the trigger level. Maybe this can be adjusted with a trigger hysteresis setting.

Thats interesting. Is there a paper about that topic to dive a bit deeper?
I saw that issue on various oscilloscopes but never saw that as an issue.
I thought its something related to the sample rate / waveforms per second update rate ratio and never thought about it in detail.

But I put our mighty Tek Series 5 scope at the office though the same challenge and even here its visible (but of course at a much lower extend):

There is no paper. That is artefact stemming from noise and large magnification of that portion. It is normal and cannot be avoided unless you have no noise at all.

Then explain to me why most of the A-brand scopes don't show this effect (or at least much less of it)? And if you look carefully, there is no magnification. The reference is a much much lower frequency used as a baseline to check the bandwidth. IMHO the problem is interpolating the trigger point between two actual samples instead of using a curve fitted onto the signal to determine the time shift needed to plot the acquisitions on top of eachother.

[2N3055]

I'm also noticing something I call 'trigger lensing'. Around the trigger point, the trace gets thinner but beyond that, it get much thicker. I don't think this is a true representation of the signal as noise around the trigger point makes the trace move left/right more than it should. A trigger condition should be targeted at an edge of a signal (IOW a section of the signal) around the trigger level crossing which to some extend (=the trigger hysteresis boundaries) cancels noise, not just the single point where the signal crosses the trigger level. Maybe this can be adjusted with a trigger hysteresis setting.

Thats interesting. Is there a paper about that topic to dive a bit deeper?
I saw that issue on various oscilloscopes but never saw that as an issue.
I thought its something related to the sample rate / waveforms per second update rate ratio and never thought about it in detail.

But I put our mighty Tek Series 5 scope at the office though the same challenge and even here its visible (but of course at a much lower extend):

There is no paper. That is artefact stemming from noise and large magnification of that portion. It is normal and cannot be avoided unless you have no noise at all.

Then explain to me why most of the A-brand scopes don't show this effect (or at least much less of it)? And if you look carefully, there is no magnification. The reference is a much much lower frequency used as a baseline to check the bandwidth. IMHO the problem is interpolating the trigger point between two actual samples instead of using a curve fitted onto the signal to determine the time shift needed to plot the acquisitions on top of eachother.

Brown signal is reference signal. Horizontal noise is order of 300 ps P-P.

What we see is exactly the trigger jitter. And yes, even 1x10E6 USD scopes have it. To different extent.
I know it is easily visible on MSOX3104T. Noise and high WFMs/s make it clearly visible.

You don't understand . At this scale, noise looks like squiggly line, single shot signal is not some clean sinewave.
Trace is that wide anyways, when you overlap thousand passes over...

[CosteC]

I'm also noticing something I call 'trigger lensing'. Around the trigger point, the trace gets thinner but beyond that, it get much thicker. I don't think this is a true representation of the signal as noise around the trigger point makes the trace move left/right more than it should. A trigger condition should be targeted at an edge of a signal (IOW a section of the signal) around the trigger level crossing which to some extend (=the trigger hysteresis boundaries) cancels noise, not just the single point where the signal crosses the trigger level. Maybe this can be adjusted with a trigger hysteresis setting.

I am pretty sure the "trigger lensing" is exactly caused by signal noise, oscilloscope noise and signal jitter. To prove/disprove this hypothesis noise/jitter shall be measured with more precise device, which is not so easy for 12 bit oscilloscope, spectrum analyser may be helpful.
Interesting may be also measuring same signal on two channels, but one with some filtering to reduce signal noise and observe result.

Trigger which eliminates noise is tempting yet it is unclear how it shall be done. Each implementation has its own pros and cons.

[tautech]

I'm also noticing something I call 'trigger lensing'. Around the trigger point, the trace gets thinner but beyond that, it get much thicker. I don't think this is a true representation of the signal as noise around the trigger point makes the trace move left/right more than it should. A trigger condition should be targeted at an edge of a signal (IOW a section of the signal) around the trigger level crossing which to some extend (=the trigger hysteresis boundaries) cancels noise, not just the single point where the signal crosses the trigger level. Maybe this can be adjusted with a trigger hysteresis setting.

I am pretty sure the "trigger lensing" is exactly caused by signal noise, oscilloscope noise and signal jitter. To prove/disprove this hypothesis noise/jitter shall be measured with more precise device, which is not so easy for 12 bit oscilloscope, spectrum analyser may be helpful.
Interesting may be also measuring same signal on two channels, but one with some filtering to reduce signal noise and observe result.

Trigger which eliminates noise is tempting yet it is unclear how it shall be done. Each implementation has its own pros and cons.

The 'lens' is only relevant and apparent on the triggered channel.

[joeqsmith]

I'm also noticing something I call 'trigger lensing'. Around the trigger point, the trace gets thinner but beyond that, it get much thicker. I don't think this is a true representation of the signal as noise around the trigger point makes the trace move left/right more than it should. A trigger condition should be targeted at an edge of a signal (IOW a section of the signal) around the trigger level crossing which to some extend (=the trigger hysteresis boundaries) cancels noise, not just the single point where the signal crosses the trigger level. Maybe this can be adjusted with a trigger hysteresis setting.

I am pretty sure the "trigger lensing" is exactly caused by signal noise, oscilloscope noise and signal jitter. To prove/disprove this hypothesis noise/jitter shall be measured with more precise device, which is not so easy for 12 bit oscilloscope, spectrum analyser may be helpful.
Interesting may be also measuring same signal on two channels, but one with some filtering to reduce signal noise and observe result.

Trigger which eliminates noise is tempting yet it is unclear how it shall be done. Each implementation has its own pros and cons.

The 'lens' is only relevant and apparent on the triggered channel.

I would expect it to behave this way.  Noise is not going to be identical across all channels and with a stable trigger, I would expect it to tighten up as shown. 

[nctnico]

I'm also noticing something I call 'trigger lensing'. Around the trigger point, the trace gets thinner but beyond that, it get much thicker. I don't think this is a true representation of the signal as noise around the trigger point makes the trace move left/right more than it should. A trigger condition should be targeted at an edge of a signal (IOW a section of the signal) around the trigger level crossing which to some extend (=the trigger hysteresis boundaries) cancels noise, not just the single point where the signal crosses the trigger level. Maybe this can be adjusted with a trigger hysteresis setting.

I am pretty sure the "trigger lensing" is exactly caused by signal noise, oscilloscope noise and signal jitter. To prove/disprove this hypothesis noise/jitter shall be measured with more precise device, which is not so easy for 12 bit oscilloscope, spectrum analyser may be helpful.
Interesting may be also measuring same signal on two channels, but one with some filtering to reduce signal noise and observe result.

Trigger which eliminates noise is tempting yet it is unclear how it shall be done. Each implementation has its own pros and cons.

If a longer section of signal is used to determine the trigger point (= finding the actual 'edge of the signal'), noise is cancelled better resulting in less trigger jitter but it will increase the minimal signal amplitude needed for stable triggering. I might have the opportunity to play a bit with the Magnova myself in the near future; I'll see if I can dig a little bit deeper into this.

[FloBX]

Maybe just some insight on how trigger-point interpolation behavior is implemented on the Magnova, what we expect you to see and ways to influence the general trigger result, if the interpolated „actual“ waveform shall not be the desired reference for triggering:

As specified in the datasheet, the maximum (Sinc) interpolated sampling rate inside the Magnova processing system is 512 GSa/s at 1.0 GSa/s ADC sampling rate and 819.2 GSa/s at 1.6 GSa/s ADC sampling rate, respectively. This on the one hand defines the temporal resolution of the determination of the actual trigger point. On the other hand the interpolation method itself already implies, that calculation of the trigger point is based on not only two succeeding points but a whole series of samples to achieve a proper interpolation/signal reconstruction behavior. The observed „trigger lens“ corresponds to a precisely matched overlay of thousands of waveforms based on a specific trigger level. We decided, that this is the behavior, we would like to provide: Highest precision triggering allowing to trigger even on slightest signal changes (which also holds for noise). This temporal precision is still visible at up to 200 ps/div (via Zoom feature or general waveform display). A „distributed“ trigger point was not intended. You might intentionally „worsen“ the displayed (not the acquired) result by switching to linear interpolation in the Display menu, which might at some point lead to a seemingly less precise match of waveform trigger positions (with all further consequences).

Some manufacturers provide(d) dedicated bandwidth filters for their trigger paths. We (and seemingly other manufacturers) decided that configurable trigger hysteresis, versatile trigger types (e.g. slope) and additional optional and configurable low-pass and HiRes filter stages preceding the trigger system should suit most requirements when influence of noise on triggering must be further reduced.

The Magnova has a trigger rate of currently up to > 14M wfms/s (simply start history mode at corresponding time scales with proper input signal and press Run) with precise determination of the trigger point taking place for all acquired waveforms. Only a very limited number of trigger types remains as an exception (based on their general concept) meaning the positions of corresponding trigger events always match sample positions / are never interpolated.

Best regards,
Florian

[egonotto]

Hello,

The Magnova introductory discount will expire soon (valid until 30/09/2024).
I think it would be good if you could see something before then.

Best regards
egonotto

[joeqsmith]

Hello,

The Magnova introductory discount will expire soon (valid until 30/09/2024).
I think it would be good if you could see something before then.

Best regards
egonotto

As far as I know, those of us in the USA still can't order it.  They never respond to me and I'm surprised that their pre-sales support hasn't been more responsive.  Hard to say what their post sales support would be. 

[Sorama]

@egonotto
What do you mean by “you” could see something?

Sorry, I don’t get it.
BR
Sam

[nctnico]

@egonotto
What do you mean by “you” could see something?

From the context my guess would be seeing an in-depth review before the sale ends so people can decide whether to buy or not.

[Sorama]

I see.

And I agree as I will “have” to place an order next week and would prefer to have my decision based on something more than a wild guess.

The scope looks nice, but so did my wife.
After “buying” her, things looked less nice.

But I guess this is part of the game…introduction price.
I mean the scope 🤣

[egonotto]

Hello,

I'm sorry, my text is very unclear. It's partly due to my poor English but also to my vague wishes.

I just wish I had more information.

For example, a video showing a few applications. This doesn't have to be a full review.

Then the data sheet could be updated.

But measurements and the corresponding pictures would also be good.

Best regards
egonotto
 

[Martin72]

It seems that the only user who has a Magnova scope is in no hurry to test it or present it here.
That wouldn't have happened to me, believe me.
As for the promotion:
Now there have been delays, maybe this promo will be extended a bit because of that?
We´ll see.

[tv84]

This scope looks very nice (I wish Batronix a big success!) but I must confess this image



was quite a disappointment for me, as a right-handed guy. Am I overreacting, guys?

[Martin72]

Ah, it took a while, but now...

[Sorama]

As for the promotion:
Now there have been delays, maybe this promo will be extended a bit because of that?
We´ll see.

If I were Batronix, I wouldn’t prolonge.

An introduction price is like a Kickstarter, you invest without knowing anything for real.
Why whould Batronix let the price be dependent of some analysis by someone?
How long should they wait?

Just playing the devils advocate.

[Sorama]

This scope looks very nice (I wish Batronix a big success!) but I must confess this image



was quite a disappointment for me, as a right-handed guy. Am I overreacting, guys?

Hmm, not sure I get it…
The bending probes ?

[egonotto]

Hello,

without an introductory discount, you can buy a Batronix Magnova BMO200 or a Siglent SDS3034X HD with Siglent SPL2016 Logic Probes and Siglent SAG1021I for almost the same money (Siglent is about 60 € more expensive).
Or for half the money you can buy a Micsig MHO3-2504, which gives you the highest resolution of the screen at 1920 x 1200.

Best regards
egonotto

[Sorama]

I admit I would not buy the Magnova at its full price.

On the other hand, the Micsig (you have) is not comparable in terms of math, fft,…
I think the magnova is way more performant in these fields, definitely when the scope really is a platform that Batronix will continue to finetune.

[tv84]

Hmm, not sure I get it…
The bending probes ?

Yep, the placement of the connectors and consequent cable crossing on the bench in an area where I place the forearm...

[pmcouto]

This scope looks very nice (I wish Batronix a big success!) but I must confess this image



was quite a disappointment for me, as a right-handed guy. Am I overreacting, guys?

Congratulations on your new scope! 

When did you place your order?

[tv84]

without an introductory discount, you can buy a Batronix Magnova BMO200 or a Siglent SDS3034X HD with Siglent SPL2016 Logic Probes and Siglent SAG1021I for almost the same money (Siglent is about 60 € more expensive).

I guess you meant BMO350, right?

[Sorama]

Hmm, not sure I get it…
The bending probes ?

Yep, the placement of the connectors and consequent cable crossing on the bench in an area where I place the forearm...

Is that really an issue?
What about the amount of space you win in front of the scope for your DUT ?