The *real* record length of Agilent Infiniivision 3000 X series?

[michkhol]

I'm totally confused by datasheets about the record lengh of these scopes. I actually have two datasheets, the old:

Maximum sample rate	2 GSa/sec, 4 GSa/s half-channel interleaved
Maximum record length	1 Mpts (digital only), 250 kpts (analog and digital channels) standard 2 Mpts (digital only), 500Kpts (analog and digital channels) with 4M memory upgrade (DSOX3MEMUP)
Maximum duration of time captured at highest sample rate (all channels)	2 Mpts (digital only), 500Kpts (analog and digital channels) with 4M memory upgrade (DSOX3MEMUP)

and the new:

Maximum sample rate	4 GSa/s half-channel interleaved, 2 Gsa/s per channel (100 to 500MHz models) 5 GSa/s half-channel interleaved, 2.5 Gsa/s per channel (1GHz models)
Maximum record length	2 Mpts half-channel interleaved, 1 Mpts per channel (Standard) 4 Mpts half-channel interleaved, 2 Mpts per channel (Optional with DSOX3MEMUP)
Maximum duration of time captured at highest sample rate (all channels)	500Kpts (analog and digital channels) with 4M memory upgrade (DSOX3MEMUP)

Besides the funny way of measuring time in Kpts it is very hard to make any sense of it. So far I could come up with the maximum of only 250Kpts of data that can be actually explored on the screen at any point of time in the standard configuration. Any other opinions?

Tek 3000 series says 5M per channel. Period.

[T4P]

I believe they actually mean Digital ON or Analog ON not at the same time.

PS: The Tek 3000 is slow as molasses

[w2aew]

The Tek spec is simple, because it has 5M on each channel - the four analog channels and the 16 digital channels (on the MSO models). The amount of memory used at any given time is user adjustable. This lets you decide which sample rate and record length you want.

In my experience the Tek scopes are certainly not slow as molasses. Not that one big difference is that all measurements are made using the acquired samples, not on the decimated display points.

[Rufus]

I'm totally confused by datasheets about the record lengh of these scopes. I actually have two datasheets, the old:

It is confusing even when you have the scope because it only shows the current sample rate.

Basically I think it has 2M (or 4M with the upgrade) of memory per 2 channels which it can write at 4GB/s.

The memory is halved in auto/triggered modes because it captures into one half while processing the other.
 
A digital channel being on halves the sample rate and uses half the memory. The only 'waste' is digital and one analog channel where memory for the other analog channel isn't used and digital on a 4 channel scope were memory for digital on channels 1 and 2 isn't used on channels 3 and 4.

That probably left you confused but basically I think it does the best it can with the limitation of 2M (4M) of memory and 4GB/s write rate.

[mikeselectricstuff]

The memory is halved in auto/triggered modes because it captures into one half while processing the other.

It is also reduced (on the 6000 by a factor of 4) in peak display mode as it stores min & max for each sample. Don't know why the reduction is /4 not /2 - not sure if the X series is different.