Magnova oscilloscope

[Martin72]

I had both of my scopes connected earlier to compare speeds.
What you can clearly see is that although fewer lines are visible at the same time in the Siglent table, the data packets are displayed horizontally in their entirety (whole words) instead of each letter running down vertically.
This is much better for clarity.
Oh, and the speed is the same for both.

[jusaca]

Yes, I prefer the way Siglent displays the data as well.
But the Magnova way might be helpful if you want to determine the timing of the individual bytes. The Siglent way allows only to get the timing of the single packets (I mean from the table, the waveform obviously has all the timing information in both cases).
By the way, does the Siglent allow to set the timeout for the packets?

[tautech]

Yes, I prefer the way Siglent displays the data as well.
But the Magnova way might be helpful if you want to determine the timing of the individual bytes. The Siglent way allows only to get the timing of the single packets (I mean from the table, the waveform obviously has all the timing information in both cases).
By the way, does the Siglent allow to set the timeout for the packets?

Yes with Trigger Holdoff the scope won't trigger within a packet.
A Falling edge is normally the right one to use for the correct starting point of a packet.

Set the Holdoff to a little more than a packet width.

[Sorama]

Hi,

• Improved the decode table layout.

What exactly has been changed there? I don't see any difference at first glance...

Link to the decoder pics 10/2024:
https://www.eevblog.com/forum/testgear/magnova-oscilloscope/msg5688091/#msg5688091


EDIT:
Request:
It would be good if there were a blank line between the data packets in the table for clarity.
At the moment, there are only flowing numbers/words.

@martin72

See message #875.

I reported the misalignment of the columns and the data/errors and this was rectified in the latest update.

Edit: do you have a better/sharper picture of the display and what the exact issue looks like?

[Martin72]

Hi,

It's not an issue, just a request to insert a blank line.
I attached two screenshots to this post, one of which shows the table:

https://www.eevblog.com/forum/testgear/magnova-oscilloscope/msg5885728/#msg5885728

[Sorama]

very good suggestion to insert some blank line for readability, but it would indeed be even better if packets are shown horizontally, like Siglent does. 

[nctnico]

very good suggestion to insert some blank line for readability, but it would indeed be even better if packets are shown horizontally, like Siglent does.

Typically there is an option in oscilloscopes to show character based protocols (typically UART decoding) as lines or single characters. Does the Magnova have that?

[Martin72]

Hi,

If I haven't overlooked anything, you can only show/hide the table.

[Martin72]

I haven't overlooked anything; you can show or hide the table, but that's all.

Does anyone know where I can find the power on settings on Magnova?
I can't find anything under “Settings.”
I have to re-set everything every time I start up.
What else I found:
The thresholds on the decoders...
They are set to 23V and 10V by default when the probes are plugged in, otherwise to 2.3V and 1V.

[egonotto]

Hello,

you can really set 500 ks/div? This then takes almost 70 days to fill the screen.

Best regards
egonotto

[Martin72]

Hi egonotto,

I didn't know what you meant at first, but yes, you can set it to 500ksec.
The trigger then triggers in 34d 17h.....

Martin

[woody]

For now I am eagerly awaiting the hardware that will enable the 'Digital' button I keep seeing in the pictures on the Batronix site 

Either someone backed their vehicle over a box somewhere or the logic probe proves more difficult to create than initially thought. Earlier this week the availability showed 1-4 days but now it's back to 10-30 days again.

Ah well, I'll have to be patient.

[Andre77]

What you can clearly see is that although fewer lines are visible at the same time in the Siglent table, the data packets are displayed horizontally in their entirety (whole words) instead of each letter running down vertically.

That's right – depending on the application, this format is more useful. We're planning to offer it as an optional display mode in an upcoming update.

Does anyone know where I can find the power on settings on Magnova? I can't find anything under “Settings.”
I have to re-set everything every time I start up.

Thanks for pointing out this unexpected behavior. Under normal conditions, the Magnova should recall all previous settings at startup. There was, however, a configuration that caused the settings not to be saved properly. Thanks to the additional information you provided via email, we were able to reproduce and fix the issue. It will be resolved in the next update.

The thresholds on the decoders... They are set to 23V and 10V by default when the probes are plugged in, otherwise to 2.3V and 1V.

Thanks for this valuable feedback as well. Currently, the internal levels of the frontend are stored at this setting. If the probe attenuation is changed afterward, the displayed thresholds are scaled accordingly. We'll improve it in a future update.

Either someone backed their vehicle over a box somewhere or the logic probe proves more difficult to create than initially thought. Earlier this week the availability showed 1-4 days but now it's back to 10-30 days again.

Sorry for the delay. There are no issues with the logic probes themselves or the software. The delay, frustratingly, is due to the measurement cable.

We're having it made by a company here in Germany, based on specific requirements: highly flexible cables, very durable connectors, and custom-labeled heat-shrink tubing. But of course it's not rocket science.

Unfortunately, the supplier repeatedly postponed delivery, apparently due to delays in manufacturing a special mold. However, we've now received a firm commitment that the cables will be ready by the end of May.

[kmoonwalker]

Hello,

I was following this thread from some time and it is really great to see that fixes are coming on regular basis

Scope looks nice and all but I have two questions:

- how "greatly" (according to firmware updates notes) was increased waveforms update rate?

- if it has a possibility to control it by mice (I mean clicking some button to activate and use a scroll wheel as an alternative to rotating any encoder via hand)?
I know it is strange but I could have some use for it.

Best regards,
KM

[0xdeadbeef]

Really nice looking scope with very sleek GUI. I love the large Full-HD screen and the active support. Letting aside the BNC connectors at the side (which would be a showstopper for my personal setup) and the unclear active probe situation, my main concern would be the sampling rate, though. For me, nothing beats sampling rate. I chose the SDS5034X mainly because it had the highest two channel sampling rate (5 GSa/s) in the~3k€ price range. Compared to that, a maximum sampling rate of 1.6GSa/s seems pretty low. For a 350Mhz signal, this would mean only ~4.6 sampling points per period (while e.g. my SDS5034X would have ~14.3 points).
Combined with the high horizontal resolution, this kinda means that for any signal with more than 1MHz or so, there are more interpolated than measured horizontal values on the screen.
So, I wonder if there will be Magnova models with higher sampling rates in the future or if this is a hurdle for such a small team with limited resources they find difficult to overcome.

[kmoonwalker]

Really nice looking scope with very sleek GUI. I love the large Full-HD screen and the active support. Letting aside the BNC connectors at the side (which would be a showstopper for my personal setup) and the unclear active probe situation, my main concern would be the sampling rate, though. For me, nothing beats sampling rate. I chose the SDS5034X mainly because it had the highest two channel sampling rate (5 GSa/s) in the~3k€ price range. Compared to that, a maximum sampling rate of 1.6GSa/s seems pretty low. For a 350Mhz signal, this would mean only ~4.6 sampling points per period (while e.g. my SDS5034X would have ~14.3 points).
Combined with the high horizontal resolution, this kinda means that for any signal with more than 1MHz or so, there are more interpolated than measured horizontal values on the screen.
So, I wonder if there will be Magnova models with higher sampling rates in the future or if this is a hurdle for such a small team with limited resources they find difficult to overcome.

I think they wanted to make the scope on some reasonable budget and as a learning curve too
From my experience with processing data it is very easy to say lets sample with x speed but hardware needs to handle it and without some special (pricey) ASICs or DSPs making some calculations in one to few clock cycles it is never easy to estimate how much computing power you really gonna need. They actually done not bad job at all to splendid on some aspects and it is a very new product whatever we can say.

My thinking is that perhaps this may not be the scope one could hope for on some aspects but I was thinking when I first saw it this way:
when it comes overall polished lets try it because it feels like an investment in new oscilloscopes maker in EU and cheaper than "metrology standard brand for VHF" - both things we need - especially nowadays
(and perhaps it may not be great for capturing erratic/rare occuring signals signals due to interpolation but still on 100/200 MHz range it should do pretty decent - after all not so long ago all we could get was 1GS/s scope with few MB of mem tops We cannot treat it by price for now as placement on current prices ladder of scopes - we are getting some things fancy like metal chassis, made in EU, etc so straight comparison should be in my opinion to scopes with sampling speeds on range of 1-2GS/s from which most is incapable of keeping real 1+ GS/s per channel with four channels same time - comparing it right now with 1GHz mobos limited by software and thus having better DAC score than full throttle is not fair for me even if price range may look the same)

[Andre77]

Scope looks nice and all but I have two questions:
- how "greatly" (according to firmware updates notes) was increased waveforms update rate?

Some performance improvements that came with the latest updates (among many other improvements):
- v1.3.4 (March): Decoding speed improved by up to 2x (depending on settings)
- v1.4.1 (April): FFT update rate doubled (depending on settings)
- v1.5.0 (coming soon): FFT speed will increase by another ~50%.

For example with 1 CH, 5 µs/div x 12 div, 1.6 GSa/s = 96 kpts (FFT with 131072 pts,  65536 bins):
- v1.3.4: ~10 FFT/s
- v1.4.1: ~20 FFT/s
- v1.5.0: ~30 FFT/s

Of course, if you're using more memory or longer timebases, the FFT rate drops accordingly. For example, with 1 Mpts, you'll see around 4 FFTs per second.

- if it has a possibility to control it by mice (I mean clicking some button to activate and use a scroll wheel as an alternative to rotating any encoder via hand)?
I know it is strange but I could have some use for it.

Yes, full control via mouse is absolutely possible – and even common. Magnova is specifically optimized for button-free operation, so mouse control (including using the scroll wheel as a virtual knob) works especially well.

Really nice looking scope with very sleek GUI. I love the large Full-HD screen and the active support.

Really appreciate that – thanks a lot!

For me, nothing beats sampling rate. I chose the SDS5034X mainly because it had the highest two channel sampling rate (5 GSa/s) in the~3k€ price range.

The SDS5034X is of course a good choice if your main focus is on the sampling rate. And this series goes up to 1 GHz bandwidth, for which the sampling rate have to fit for. The sampling rate has a major influence on the price of an oscilloscope. Therefore, an oscilloscope series designed up to 350 MHz bandwidth will not have the same sampling rates as another series with up to 1 GHz bandwidth.

We focused on processor and FPGA performance, as well as on features and user comfort - such as the 15.6" Full HD display and a metal housing that serves as a heatsink, allowing for silent, fanless operation and a lot of other points. The sampling rate exceeds the requirements of the Nyquist criterion. With a well designed sin(x)/x interpolation, a sampling rate of around 2.5 times the signal bandwidth is typically sufficient to accurately reconstruct a sine wave. And near the upper end of the bandwidth, signals tend to resemble a sine wave anyway.

So, I wonder if there will be Magnova models with higher sampling rates in the future or if this is a hurdle for such a small team with limited resources they find difficult to overcome.

A new series with a higher sampling rate and bandwidth is technically feasible for us. However, we want to be transparent: there is no new Magnova series coming out in the near future. Our current focus is on expanding the Magnova as a platform – which already delivers strong performance and still offers plenty of potential for new features. The good news: future series will directly benefit from this continuously evolving software ecosystem.

[kmoonwalker]

Scope looks nice and all but I have two questions:
- how "greatly" (according to firmware updates notes) was increased waveforms update rate?

Some performance improvements that came with the latest updates (among many other improvements):
- v1.3.4 (March): Decoding speed improved by up to 2x (depending on settings)
- v1.4.1 (April): FFT update rate doubled (depending on settings)
- v1.5.0 (coming soon): FFT speed will increase by another ~50%.

For example with 1 CH, 5 µs/div x 12 div, 1.6 GSa/s = 96 kpts:
- v1.3.4: ~10 FFT/s
- v1.4.1: ~20 FFT/s
- v1.5.0: ~30 FFT/s

Of course, if you're using more memory or longer timebases, the FFT rate drops accordingly. For example, with 1 Mpts, you'll see around 4 FFTs per second.

- if it has a possibility to control it by mice (I mean clicking some button to activate and use a scroll wheel as an alternative to rotating any encoder via hand)?
I know it is strange but I could have some use for it.

Yes, full control via mouse is absolutely possible – and even common. Magnova is specifically optimized for button-free operation, so mouse control (including using the scroll wheel as a virtual knob) works especially well.

many thanks for prompt reply

I am aware it is not a special ASIC scope so there are limitations - thou I have big apetite for that 8 Mp FFT

[tautech]

thou I have big appetite for that 8 Mp FFT

You greedy sod ! 
SDS6000A territory but even the little 800X HD can provide you with 2 Mpts FFT

[kmoonwalker]

 

[kmoonwalker]

what I often wanted is alike during old days (looking on some high end HP scope options) but to have a possibility to replace a computing module for example today I am purchasing a mobo and later on I can add an ASIC to a socket or even never one after few years.

and I know it will not happen being a nightmare in terms of software versions and keeping all up to same standard, etc and for such things we need those big boxes from Rohde but if you think I am greedy with 8 lets think how 32 would be suitable for catching a faint signals in background heh - one can only dream - perhaps in next 10 years something with enough computing power will be price reasonable

I would love more than few k waveforms/s to have it logging and to have that 0,5% chance of catching an erratic signal... so like two weeks of recording or so...

[tautech]

...............
but if you think I am greedy with 8 lets think how 32 would be suitable for catching a faint signals in background heh - one can only dream - perhaps in next 10 years something with enough computing power will be price reasonable

Your dream is available also with a 15.6" display......SDS7000A !

[kmoonwalker]

...............
but if you think I am greedy with 8 lets think how 32 would be suitable for catching a faint signals in background heh - one can only dream - perhaps in next 10 years something with enough computing power will be price reasonable

Your dream is available also with a 15.6" display......SDS7000A !

only that aprox 40k euro price tag is not very appealing value but I am open for donations xD

[KungFuJosh]

...............
but if you think I am greedy with 8 lets think how 32 would be suitable for catching a faint signals in background heh - one can only dream - perhaps in next 10 years something with enough computing power will be price reasonable

Your dream is available also with a 15.6" display......SDS7000A !

I think you proved their points. He wanted the 8M FFT and 15.6" FHD screen. You suggested Siglent's cheapest and most expensive scopes.

Fact is, there isn't a model that directly competes with the Magnova for FFT + big FHD screen + budget price...plus active probe interface.

I (obviously) love Siglent products (6 on my bench including 2504X HD and 5054X scopes), but this is the area that annoys me about their new product releases. Why would they release anything with potential for window splitting, and max out at 1280x800??

Want the screen size and/or resolution, and don't need the higher bandwidth? Why pay 10 times as much for the 7000A series?

The new 5000X HD series absolutely should be FHD, and they're ridiculous for doing anything less. I'd still love one either way though. 😉

Thanks,
Josh

[kmoonwalker]

Well as fancy FHD is there are reasons why manufacturers are sticking to lower resolutions - it is not really the major thing with display or something like most people think.

Main reson is that if you have for example 800x600 display you have 480k points that you need to update on that screen very fast with scope and such amount is really easy to do.

The higher we go ie. 1280x800 => 1 M points

FDH => ca 2 M points

with all math aplied this can go up to 4-8-16-32 times more computing power to even show something on display (or more with math then smoothening, approximation of curves etc)

Magnova processor needs to work way harder to present same thing for us as observers, this takes more memory for even system/GUI so it is not even possible to directly compare scope with low res to big res one sometimes. Some save on RAM some save on mem depth (this also requires ca x^2-x^3 more computing power with amount increase and then you reach the point where you for example feeded a good cpu with enough ram but thing start to be sluggish due to its overall size and you need to cut some things a bit)
 
For me responsiveness of scope usually matters more than ability to surge it to do some complex thing and during that time it is inoperable or barely reacting (also usually blind) and yet I want my big depth FFT in cases I may need to call it...