70Mpts Memory Depth:
Now we actually get up to 35Mpts per channel and 70Mpts interleaved, but in terms of sample rate preservation, it is ‘only’ 28Mpts and 56Mpts, respectively.
With all channels on, we can have 1GSa/s up to 2ms/div, which translates to a record length of 28Mpts. At 5ms/div, the memory increases to 35Mpts, but still the sample rate drops to 500MSa/s, as we would actually need 56Mpts in order to preserve the 1GSa/s.
Likewise, in interleaving mode, i.e. just channel 1 or 2 plus channel 3 or 4 turned on, we get 2GSa/s up to 2ms/div, which results in a record length of 56Mpts. At 5ms/div, the memory increases to 70Mpts, but the sample rate drops to 1GSa/s
With the digital channels enabled, things get a little more complicated. Generally, the sample rate on digital channels cannot exceed 500MSa/s.
With all analog channels on, we can maintain 1GSa/s up to 1ms/div, which means 14Mpts. At the same time, digital channels have 500MSa/s and 7Mpts.
At 2ms/div, analog channels drop to 500MSa/s, thus now being equal to the digital sample rate, and memory depth is still 14Mpts on analog and increases to 14Mpts for the digital channels as well.
In interleaving mode, we can maintain 2GSa/s up to 1ms/div, which means 28Mpts. At the same time, digital channels have 500MSa/s and 7Mpts.
At 2ms/div, analog channels drop dramatically to 500MSa/s, and memory depth is s14Mpts for both analog and digital channels.
Slower timebase settings don’t affect memory depth any further, but sample rate on both analog and digital channels drops accordingly.
Conclusion:
When using the SDS2000 as a DSO, we actually benefit from the increased memory. As an MSO, nothing seems to have changed, so one might assume that a major portion of the additional 42Mpts memory comes from the digital part and consequently is only available when digital channels aren’t in use.
Anyway, I think this is a clever move and I’m quite comfortable with it.
EDIT: Corrected analog samplerate on interleaved mixed mode.