Magnova oscilloscope

[2N3055]

We already explained before.

Memory mode must be in AUTO.

Trigger mode must be in Normal.

Apply 10-20 MHz signal to input, with vertical set it to nicely fit on screen vertically. Set edge trigger, and make it trigger reliably.

Then measure.

If you set scope to manual memory mode of 300 MPTs you command it to always capture full 300 Mpts. Regardless the timebase setting.

I tried with inexpensive SDS800xHD. At 100µs/div, it gets 200 triggers per second at 2Mpts and 287 triggers/s at 1MPts (1GS/s). Magnova numbers are faster than that scope, and two orders of magnitude out in your measurements.

All settings were already set 1:1 as you wrote
Some results are mentioned in my post, but I'll add them to my table!

p.s. 3rd column appeared for Magnova. 10MHz input.

My short conclusion:
a. there is a 20-year-old Agilent that have independant FIFO memory buffer (hardware solution) and a fast/independent trigger;
b. there is a ''modern digital'' sampling tool. Ok, I understand that small manufacturers like Batronix don't have access to custom-made FPGAs (this results in well-known problems at high sampling rates). But, sorry, at slow sampling rates, there are software solutions for memory management.

Yeah, you are doing something wrong. Those numbers are not right.
How are you measuring triggering frequencies?

[kaspich]

We already explained before.

Memory mode must be in AUTO.

Trigger mode must be in Normal.

Apply 10-20 MHz signal to input, with vertical set it to nicely fit on screen vertically. Set edge trigger, and make it trigger reliably.

Then measure.

If you set scope to manual memory mode of 300 MPTs you command it to always capture full 300 Mpts. Regardless the timebase setting.

I tried with inexpensive SDS800xHD. At 100µs/div, it gets 200 triggers per second at 2Mpts and 287 triggers/s at 1MPts (1GS/s). Magnova numbers are faster than that scope, and two orders of magnitude out in your measurements.

All settings were already set 1:1 as you wrote
Some results are mentioned in my post, but I'll add them to my table!

p.s. 3rd column appeared for Magnova. 10MHz input.

My short conclusion:
a. there is a 20-year-old Agilent that have independant FIFO memory buffer (hardware solution) and a fast/independent trigger;
b. there is a ''modern digital'' sampling tool. Ok, I understand that small manufacturers like Batronix don't have access to custom-made FPGAs (this results in well-known problems at high sampling rates). But, sorry, at slow sampling rates, there are software solutions for memory management.

Yeah, you are doing something wrong. Those numbers are not right.
How are you measuring triggering frequencies?

Magnova doesn't report auto-triggered events via the Aux output; I am using their own wfps value on the screen.
But, trust me, refreshrate is vvveeerrryyy sslooowww.
For sure, everything is fine for ns time scale - 100+k wfps

[2N3055]

As a nugget how things can be misread when you setup things wrongly in deep menus:

On Keysight MSOX3104T, you go to Utility->Rear Panel and select ->TrigOut:Trigger source.
Apply 20 MHz input to scope. And on TrigOut BNC you measure 20 Mhz.
In this mode scope patches out signal from analog trigger comparator before triggering engine.
Basically you get squared signal from input, with front end being variable preamp. I used this few times to do exactly this: I took signal from sensor and created 2,5V digital signal from it. It works until some frequency of 50-100MHz with progressively worse pulse shape. Up to 10-20 MHz works well.

You go back and set Utility->Rear Panel and select ->TrigOut:Triggers,
Now you get a pulse every time scope starts acquisition. This gives you Wfms/s

If you had this setting wrong, you would think your scope has 20 MWfms/s trigger rate.

[jusaca]

[...] is there a need for extra deep analysis? That's a disaster...

Either a disaster, or just plain old user error... Which it certainly looks like in this case
Please learn to use the equipment before filling this thread with walls and walls of text 

[kaspich]

As a nugget how things can be misread when you setup things wrongly in deep menus:

On Keysight MSOX3104T, you go to Utility->Rear Panel and select ->TrigOut:Trigger source.
Apply 20 MHz input to scope. And on TrigOut BNC you measure 20 Mhz.
In this mode scope patches out signal from analog trigger comparator before triggering engine.
Basically you get squared signal from input, with front end being variable preamp. I used this few times to do exactly this: I took signal from sensor and created 2,5V digital signal from it. It works until some frequency of 50-100MHz with progressively worse pulse shape. Up to 10-20 MHz works well.

You go back and set Utility->Rear Panel and select ->TrigOut:Triggers,
Now you get a pulse every time scope starts acquisition. This gives you Wfms/s

If you had this setting wrong, you would think your scope has 20 MWfms/s trigger rate.

wow. something new to me

I am talking about Magnova. But, go ahead

[kaspich]

[...] is there a need for extra deep analysis? That's a disaster...

Either a disaster, or just plain old user error... Which it certainly looks like in this case
Please learn to use the equipment before filling this thread with walls and walls of text 

ok, smart guy. Please show the result of your test!

[Andre77]

Let's start with very simple things. All we know - Magnova is a fast/responsive scope. Let's check the situation outside the sweet spot!
Please see the attachment. I am lazy, I didn't use a calculator, Excel formulas - hopefully, if there are no math mistakes.

Hi Kaspich,

I'm sorry, but the values you mentioned can't be correct.

Please try the following steps:
1. Settings -> Default Settings
2. Acquire -> Disable Extended Capture (This is a useful feature, but it slows things down by a factor of two, so to ensure the measurement is comparable to those from the other scopes, please disable it.)
3. Apply a 10 MHz test signal to the channel as the trigger source
4. Check whether the trigger level matches the signal.

You should now see approximately 357 wfms/s, as shown in the following screenshot.

If you don't achieve a similar speed after following the steps above, please post a screenshot (or send it to me).

In your table, you wrote only 4 wfms/s at 100 µs/div. This could be because you did not set the memory setting to 'Auto', but rather to a fixed, very high value. This forces the oscilloscope to capture a wide time range. If you set 327 MPts, the oscilloscope would have to record 204 ms at 1.6 GSa/s (327E6 pts/1.6E9 Sa/s = 0.204 s). With that, 4 wfms/s is realistic; even with zero dead time, 5 wfms/s would not be achievable.

[kaspich]

Andre, thanks for the feedback.
a. memory is/was set to Auto(fast).
b. in the attachment - table for faster time scale settings, too.

I'll try the steps recommended by you soon, but I get what I get
Could be some problems/bugs.

p.s. As we see, everything is fine at higher time scale settings!

[kaspich]

ok, guys, please see the attachment.
My scope remains on 3.8M points @ 100 us/div.
As we see, the input frequency is 20MHz
38 wfps (as in my previously posted pictures).

4 wfps was an untriggered situation!

p.s. 3..4wfps/100us also attached.

Yes, with the disabled extended capturing feature, the memory depth dropped down by 2; now there are 360 wfps @ 100us, if/when the signal is applied.
But, untriggered refreshrate still is 4wfps...

[Sorama]

ok, guys, please see the attachment.
My scope remains on 3.8M points @ 100 us/div.
As we see, the input frequency is 20MHz
38 wfps (as in my previously posted pictures).

4 wfps was an untriggered situation!

p.s. 3..4wfps/100us also attached.

What about the other, requested settings?

[kaspich]

ok, guys, please see the attachment.
My scope remains on 3.8M points @ 100 us/div.
As we see, the input frequency is 20MHz
38 wfps (as in my previously posted pictures).

4 wfps was an untriggered situation!

p.s. 3..4wfps/100us also attached.

What about the other, requested settings?

please be a little bit patient
I upgraded my post

[core]

But, untriggered refreshrate still is 4wfps...

And this is normal.
Without a signal it has no reason to trigger, it is just the auto-trigger rate.

[Andre77]

Yes, with the disabled extended capturing feature, the memory depth dropped down by 2; now there are 360 wfps @ 100us, if/when the signal is applied.

Okay, so with the default settings, your Magnova is now as fast as expected.

But, untriggered refreshrate still is 4wfps...

That’s actually normal behavior and not specific to the Magnova.

In auto-trigger mode, the oscilloscope intentionally generates a fallback trigger if no valid trigger event occurs. This is done via an internal timing mechanism so that the display doesn’t remain blank when no signal is present.

This fallback is deliberately slow (typically a few Hz, ~3–5 wfms/s) on essentially all oscilloscopes. The reason is a trade-off:
- Auto-trigger must still work reliably for very low-frequency or sporadic signals
- So the scope needs to wait long enough to give a real trigger event a chance to occur
- If the auto-trigger timeout were much shorter, the scope would constantly self-trigger and you would never be able to properly capture slow signals

In other words: A slow auto-trigger rate is by design, not a performance limitation.

So the behavior you’re seeing (~4 wfms/s without a valid trigger) is exactly what you would expect from other vendors as well.

[kaspich]

But, untriggered refreshrate still is 4wfps...

And this is normal.
Without a signal it has no reason to trigger, it is just the auto-trigger rate.

oh, ok, I am a strange guy

Could I ask why my other scopes have a higher refresh rate?
Maybe there are some problems with my other scopes?

[hotze]

oh, ok, I am a strange guy

ehm .. yep.

[woody]

As much as I detest (the tone of) this thread lately, I must admit there is something strange with regard to the wfps. When I set my scope as Andre suggests I see the expected 357 wfps. As soon as I set a frequency measurement however, this drops to the 40 wfps Kaspich sees.

No idea if that is to be expected, as I realize my knowledge is limited.

[kaspich]

Yes, with the disabled extended capturing feature, the memory depth dropped down by 2; now there are 360 wfps @ 100us, if/when the signal is applied.

Okay, so with the default settings, your Magnova is now as fast as expected.

But, untriggered refreshrate still is 4wfps...

That’s actually normal behavior and not specific to the Magnova.

In auto-trigger mode, the oscilloscope intentionally generates a fallback trigger if no valid trigger event occurs. This is done via an internal timing mechanism so that the display doesn’t remain blank when no signal is present.

This fallback is deliberately slow (typically a few Hz, ~3–5 wfms/s) on essentially all oscilloscopes. The reason is a trade-off:
- Auto-trigger must still work reliably for very low-frequency or sporadic signals
- So the scope needs to wait long enough to give a real trigger event a chance to occur
- If the auto-trigger timeout were much shorter, the scope would constantly self-trigger and you would never be able to properly capture slow signals

In other words: A slow auto-trigger rate is by design, not a performance limitation.

This is also why manufacturers generally recommend:
- Auto mode → for finding signals / unknown conditions
- Normal mode → for proper measurements and maximum waveform update rate

So the behavior you’re seeing (~4 wfms/s without a valid trigger) is exactly what you would expect from other vendors as well.

Andre,

not so simple.
Default settings mean - extra capture is on.
Second - that's what I am talking about (internal mechanism/timer, that can be easily modified).

If I am inspecting an unknown signal. I am waiting for a fast/immediate response. Not 1 sec delay and 99% ''off'' time. At this moment, I have no clue how to set my trigger. I need a fast-responding scope that is always ''on''. Like an analog scope (that automatically enables trigger if sync is not identified).
That's how Agilent works. That's how Siglent tries.

And, even with this extra capturing off, Magnova is about/around 30% ''live'', instead of 80..90% of UNtriggered Agilent.

Ok, again, if I'm unique, no problems. I'll send my device back.

[kaspich]

As much as I detest (the tone of) this thread lately, I must admit there is something strange with regard to the wfps. When I set my scope as Andre suggests I see the expected 357 wfps. As soon as I set a frequency measurement however, this drops to the 40 wfps Kaspich sees.

No idea if that is to be expected, as I realize my knowledge is limited.

It's a normal situation because measurements require additional computing resources/time. Other topic - why such a significant drop is noticed at very slow time base settings.

[john74]

If I am inspecting an unknown signal. I am waiting for a fast/immediate response.

In these situations I always use the "instant/always" trigger setting.
Here you have the fastest response when the exact trigger condition is not known yet.

[kaspich]

If I am inspecting an unknown signal. I am waiting for a fast/immediate response.

In these situations I always use the "instant/always" trigger setting.
Here you have the fastest response when the exact trigger condition is not known yet.

ok, so, every time to set this, I need to enter the Trigger menu, then the Type submenu (why basic settings are not in the same menu - God knows), then change settings. Then press Yes. Then repeatedly press Yes. And, so on and on.
Btw, with all this done, the "on'' ratio is about 30%.

Guys, I have a MEGA proposal. Super Innovation! Please, create an ''intelligent'' trigger! Trigger, that works as a "usual'' one (selected from the list), but is activated by the timer, too (+ time scale) if the "real'' trigger is not activated!

 

[Andre77]

Second - that's what I am talking about (internal mechanism/timer, that can be easily modified).
If I am inspecting an unknown signal. I am waiting for a fast/immediate response.

Okay, now it’s becoming clear what you’re actually getting at. You're not concerned with the Magnova's waveform update rate, but rather with its behavior in auto-trigger mode - specifically the delay we intentionally built in when there is no valid trigger event.

The current implementation uses a relatively long auto-trigger timeout to ensure that even very slow or sporadic signals can still be captured reliably.

However, this is not a fundamental limitation - the auto-trigger timeout can be significantly reduced in software.

We could also consider adapting the timeout dynamically, for example shortening it when no valid trigger events are detected over a longer period of time.

We’ll take this into internal review and follow up with an update that significantly reduces the auto-trigger delay or makes it configurable, including some adaptive optimization. This may take a little time as we’re currently finalizing another release (and some of my colleagues have just left on vacation), but we will extend your evaluation period accordingly to ensure you can properly test the improvements.

I would also like to suggest that we discuss the details directly via email rather than continuing the discussion here in the forum, as some of the posts here might be quite confusing to casual readers.

Default settings mean - extra capture is on.

I only mentioned that to ensure comparable recording times between different scopes. It does not affect the auto-trigger timeout.

[Sorama]

I look forward to the upcoming firmware this week.
DVM, hardware counter…

[KungFuJosh]

Lonnnnng story short, not a bug, but a feature request.

[kaspich]

Lonnnnng story short, not a bug, but a feature request.

Josh, I am a real fan of yours!

Every single step I am trying fails.
I tried Instant trigger.

100Hz input, square wave.
Please see what I got (when I stopped the scope). Ok, I understand why I got this, but question (especially to you) - it's still not a bug?

Sorry, guys, for a brain f&^$*.

[rf-messkopf]

Oh crap... priceless