Rigol DS1074Z

If I'm understanding what you are saying:

"Scroll mode" can be nice at low sweep rates. Analog 'scopes have a "holdoff" time that limits the maximum trigger rate.

100ns is unbelievably fast. To have a longer record, or reduced inter-record time could be done with special-purpose hardware.

If you simply want the sweep to restart as soon as possible after the last one has completed - you could simply feed a fast clock into the trigger circuit.

I have a ds1054Z, and have used the expanded trace feature a lot to split a waveform up, and scroll through the expanded section, which may be several 'screens' wide. Never noticed any anomalies as showing gaps across screen boundaries.

There are a few oddities in the way it stores waveforms for later display, that seems to be on a screen-by-screen basis in some modes, but the main memory seems to do it all in one piece, set by the trigger point.

Maybe the 1074 is different, but I rather doubt it.

Frank - the ds1074z is a digital sampling scope with a purported memory of

12 million "points", not an analog type.

Hul

Frank Miles wrote:

Yeah I've got the 1052e, I don't use the expanded display much.

Maybe try the eevblog forums? Say is the x-y mode any better than the 1052? George H.

Say

I never used either 1052 or 1054 in X-Y mode, so no idea.

The advanced triggering functions in the 1054 are a huge advantage over the 1052, but mostly for digital analysis work, which is what I do.

I am aware of that. However most digital scopes have a longer interval between trigger acceptance - the comparison is still useful in many circumstances.

Adrian - the scope I'm using has currently a high frequency (100 hz) square wave on one channel to which the trigger is set. Another channel shows a square wave of about 10 cps. The hi freq. channel starts with 1/4 cycle low, then several 1/2 high - 1/2 low sequences and enda with 1/4 cycle high. This is the apperence of the high freq. signal in all "frames". The trigger is not altered from normal use, but the mode selecter in the trigger group on the panel, ie the 3 push buttons, was tried on "single" mode. Your mentioning the "trigger point" being the control for how the memory opperates sounds like what is needed. Do you happen to remember specifically how you set the "trigger point"? Whether you do or not, you've pointed the way to getting this scope to work as expected, for which I'm most grateful.

Hul

Adrian Jansen wrote:

You might get some gasps by calling a 100 Hz wave 'high frequency' :-)

Ok, so assuming the 1074 has similar triggering facilities to the 1054, hit the trigger menu and then select pulse width triggering on the 'soft' buttons, rather than the standard edge triggering, and set the max pulse width to just over 1/4 of a cycle ( ie 1/4 of 5 msec, for a

100 Hz period ). You also need to select whether you want positive or negative going pulses. then the scope will trigger only on a leading 1/4 cycle of the correct polarity, and ignore the longer 1/2 period pulses. That sets where the memory trigger point is. Memory is actually filled both before and after that point, which is usually set at the horizontal centre of the screen. You can also set the hold-off period to longer than the total pulse train length, so it wont re-trigger on the trailing 1/4 period pulse as well. Alternately you can use single trace mode to just trigger on one of the pulses as you have set up above, and then any subsequent pulse sequences are ignored, until you push the Run/Stop button again to re-arm the trigger.

Once you have all that set up, and captured a waveform, you can 'inspect' bits of it just by pressing the sweep speed knob, which splits the screen to show the full capture, and allows to expand any part of it by changing the 'sweep' speed for the split part, scrolling along it using the sweep position control.

Its worth reading the manual on how to set up triggering too. Triggering on a DSO is vastly more sophisticated and useful than on the older analog CROs.

+1! Re: DSO triggering, If you've got a noise signal you want to measure the spectrum of try, triggering at the tippy-top of the signal (normal mode) and average. You get something that looks like the auto correlation function. And a cleaner FFT. Warning: The FFT falls off faster than expected at the edges. the 3db point of a filter is 6dB down on this "averaged" then FFT.

George H.

sci.electronics.design Rigol DS1074Z scope

A little more looking into Rigol's DS1074Z scope has found both the horizontal sweep rate and the "memory depth" can be adjusted manually. The sample rate is determined by the scope's computer according to the formula, as stated in their manual:

mem. depth = sample rate * (horz. rate per grat.) * 12

With the horizontal rate set at 10 millisec and the memory depth at 12 million "memory points", the sample rate is shown on the screen as 250 million per second, which doean't quite agree with the above formula, possibilty because it's derived from a list's closest terms. Anyway, the terms as shown, result in 2.5 million samples per graticle. Awk! The length of recorded data doesn't change with a change in "memory depth", only the number of "memory points" in which it's held. More "memory points" means higher sample rate with Rigol's scope, which is contrary to "more memory" normally implying more data and a longer recoring time. As Rigol's scope appears, it's only good for 10 slightly random "frames" of recorded data which would, in the intended application, provide about 1/2 second of data rather than the 5 or 6 or 7 seconds needed. It could be that Rigol isn't working a con with the DS1074Z and I have simply missed something that would make the scope useful as expected and, if anyone see's what that is, do mention it.

Hul

Hmmm, I'll have to look at mine but something tells me I could set the Max deapth which will get readings of a single sweep. Setting the sweep time to slow!!!! will capture more data but reduce resolution when you scroll back and expand it.

So If I do a 10 sec sweep, if it can, with max memory setting, I should be able to capture a lot of data?

I've never maxed mine out to the end but I do think this is what I did the last time I was looking for data with in a frame of a burst, using the other channel to start the sweep.

Jamie

OK I have no idea if this will work, but you are doing fairly slow stuff, so in a roll mode... voltage scrolls across screen.. you may be able to read the data out fast enough to keep up... otherwise get a labjack, or other A-D ->usb and read it out "forever".

George H.

For those specs, and if he does not need DC, a basic sound card logger will do.

Jamie

Jamie - I tried setting the time base to 1 sec per grat. with the memory depth set at 12m, as you suggested. With a 100 cps signal on chanel 1, the hold button was pressed then the "delayed time base" expansion showed the signal in recognizable form. Basically what is needed and I thank you for pointing the way. Rigol's "delayed time base" is easily, for me at least, mistaken for "delay time base".

Hul

M Philbrook wrote:

I thought we already went through this in an earlier thread.

If you set the main sweep to 10 seconds total, ie right across the screen in 10 sec. Then that records 12 million samples over 10 seconds, or 1.2 uS per sample. So you can expand the trace using the split screen control ( agreed that is wrongly named 'delayed sweep", but that is not the issue ) as far as you like, the limit of resolution you can see is 1.2 uS. Surely that is adequate for looking at millisec pulses ?

Actually it occurs to me we may be talking about 2 different sorts of storage here.

There is the main storage of the DSO in fast RAM, with its 12 million points of data storage, from which you can view on screen some or all of it by using the split screen functions. That acts somewhat like a circular buffer, and is filled/refreshed continuously as signals come in. You can select a 'trigger' point with a whole variety of constraints, which let you set where a particular piece of the captured waveform appears on screen ( usually in the middle ), but can be set elsewhere using the position control. You dont need to do anything apart from set the trigger conditions and the sweep speed to to operate the scope in this mode, memory is filled and displayed automatically.

Then there is the 'storage' function which allows to record part or all of the main memory into internal or external flash memory as a file ( USB stick, etc ). That is set up to record as 'screen shots' of image files, or as data in csv files. And you can select whether you want either the entire memory dumped, or just an individual screen, in various formats, and with or without the screen setups etc. That has really nothing to do with how the CRO gets the data, only how you can record it to permanent storage.

Now, which of these two functions are we talking about ?

Adrian - unfortunantly, we were talking about both of Rigol's storage functions. Initialy going through the manual, I recognized only the multi "frames" storage capability and took the "delayed time base" to be a description of a delayed sweep mechanism & disregarded it. Later it was mentioned as a possible method and indeed it was.

Hul

Adrian Jansen wrote: