........... the 6000A would come into question with Siglent, possibly the new SDS7000A, which according to siglent eu should be on the market this year, like the larger lecroys it is a pc-based scope.
It might be beyond tooki's budget however as the base model will be 2 GHz BW which for SDS6204A (top 6000 model) is already $9990.
https://www.eevblog.com/forum/testgear/siglent-sds7000a-dsos-coming/
Since Siglent is closely entwined with LeCroy, it’s my understanding that Siglent scopes lean more towards being “analytical” scopes like LeCroys, as opposed to “interactive” or “real-time” scopes like Keysight (and Rigol, which more closely copies them). Since I already have a heavy-duty analytical scope, I think I should be focused more on a real-time model. Does Siglent have something more like that?
Logic of Siglent scopes is very similar to logic of LeCroy.
Slightly slower refresh rate is not show stopper. It simply means sometimes you will see the glitch every 5 seconds instead of every 2 seconds. It is slower but still works. But lack of memory will prevent you from using high sample rate for longer timebase. There is no workaround for that. Your 500MHz scope soon becomes a 20 MHz scope.Or less. Or undersamples....
I have here 2 Siglent scopes and also Keysight MSOX3104T.
I don't see much advantage from super high refresh rates on Keysight, but I do see problems because of short memory on Keysight that makes it drop sample rate much quicker as you go to longer time-bases. 2MPTs compared to 100MPTs is large factor.
Let's say you want to measure a parameter, like risetime, and want to do statistics on 1000 edges for an average.
Keysight takes only one measurement per trigger. On first edge from the left side of the visible screen, not in relation to trigger.
You will need 1000 trigger events until you get your 1000 measurements. So you shorten timebase so only one event is on screen. And since it uses decimated data, short timebase will increase timing resolution too. You need to always keep short timebase anyways, because otherwise sample rate drops. It needs super fast triggering because otherwise......
On Siglent or LeCroy, enter deep measurements.
You take one, single, capture with 1000 edges in it. Since you have long memory you keep maximum sample rate and retain same timing resolution as with short timebase. Scope will go into that capture and measure each and every edge and add them to the stats. All 1000 of them from a single trigger. Or 10000 from a single trigger.
In this particular case, which one is faster?
Also hardware decode in Keysights is both good and bad. It is fast, but you have to setup everything perfectly up front, before capture. You cannot capture some nice long UART data, and you realize it is not 8bit no parity but 7bit with parity and it all looks scrambled. You cannot go into settings, set it right and it will then decode properly. It was decode at capture time and that is it. You need to update settings and recapture new capture from begining. OTOH, this scope actually captures data with a comparator built in analog channel. Because of that it will sometimes decode data that in analog view looks unrecognizable... But that is a crutch that works despite it's short memory..
On scopes with software decode, you can capture some data even without decode being on. You can enable/disable decode at will and play with settings until it shows propper data...
Also Siglents/LeCroy have always running History of captures in background. It is similar to segmented capture but while normally using scope...
Every concept will have bad and good sides...