On Tue, Dec 22, 2020 at 03:24 AM, Jeff Dutky wrote:
My post really was meant tongue-in-check, in reaction to Roy Thistle's earlier post. We've seen prices of some models of equipment increase after more attention was paid to them on the 'net. It's a familiar market process.
The second point in my post referred to the relative ease of repairing a (switching) power supply as opposed to the difficulty to be expected when trying to repair an acquisition board, exacerbated by the lack of replacement components. The internal PC often can be replaced.
Regarding BW needed: There is a rule of thumb that says that at least 5 x the highest frequency component that you want to see in some detail determines the BW you need.
As regards DSO's, especially for digital applications, a few points:
I don't personally know the TDS5000 series and haven't studied their characteristics. Some say that the fact they "run on" a PC is a disadvantage. I have no reason to agree without having further information.
For analog 'scopes, BW is (almost) everything. Much less so with digital 'scopes. With digital 'scopes, samples/second is of paramount importance because it's the one feature that determines the ability to see short, one-shot events in any detail. As an example, a 400 MHz DSO with a maximum real-time sample rate of 200 Msa/s (like the old TDS400 series) is ok for watching a periodic square wave with a rise time of say 5 ns but useless for a one-shot observation in that range. "Equivalent Time" sampling won't help in real-time situations of course. If you want to see an edge or inspect a possible setup violation in a digital system, even running at only 1 MHz (your 6800), you need to resolve that time in enough detail, especially in one-shot or jitter situations.
Next come trigger capabilities. I rate those above memory depth because triggering makes acquisition selective.
Another important factor is "number of waveforms per second" (wfm/s). In contrast with analog 'scopes, DSO's, and especially early ones, had a very long time between subsequent "scans", since they operated in a sample one dataset - then store - then calculate - then present - and sample again. Repetition frequencies of less than 100 Hz weren't uncommon for 1 GHz DSO's (TDS680)! Most of the time, the 'scope was "blind", causing events happening during that time to be missed. Modern DSO's can perform at least at 10,000 wfm/s, many at 100,000 or more, so very little blind time. With strong triggering capabilities, the disadvantage can be mitigated.
There's a limit to the attractiveness of analog metaphors in digital 'scopes. I'm quite fond of the late-90's / early 2000's LeCroy 'scopes. They're really digitizers with piles of ways to process the digitized data. They're nothing like an analog 'scope - gone digital but are quite suited to a scientific approach, based on their digital nature.
I recommend you start by looking for one of the modern affordable Rigols or Siglents. They'll give you lots of fun and measurement info and convenience. Analog 'scopes are still worthwhile (apart from fun in their repair and restoration) because there's nothing more useful with newer models but with DSO's, development has been steep rather recently and cost/feature has come down. Just be prepared to meet bugs...