Have a look at my new Oscilloscope (New to me that is!)

[Hero999]

Are you talking about the hold-off?

Set that to a minimum.

I suppose you're right, there will be a short delay between triggering and a trace but if the frequency is very low, it will be negligible.

So on reflection 1MHz is probably too high. I normally use the mains frequency to calibrate my 'scope and I do so in the manner I described above so the hold-off effect should be minimal.

 

[crutschow]

Are you talking about the hold-off?

No. I'm taking about the retrace time. The time it takes to reset the sweep circuit from the end back to the start of the sweep. At 60Hz it's probably not significant. At 1MHz it could be.

 

[Speakerguy]

Modern digital scopes are power houses for debugging. I own a Tek DPO3034. While DSO scopes might not match up in some areas to analog scopes, like debugging glitches and the like, DPO scopes blow everything away. The triggering modes, memory, analysis options, it's just crazy how much they can do. Mine will even decode I2C and SPI and display the info on the screen in ASCII, binary, , hex etc and store the data transmissions in an event table for me to look at. Modern DPO's really do rock.

But old analog scopes like the one in the OP are still very, very usable tools. They will do 95% of what my DPO will, in terms of features that I regularly use. The other 5% costs a ton of money though.

 

[Hero999]

Do some digital 'scopes have a Fourier Transform function?

That would be cool having a spectrum analyser function even if it did only work for audio frequencies.

No. I'm taking about the retrace time. The time it takes to reset the sweep circuit from the end back to the start of the sweep. At 60Hz it's probably not significant. At 1MHz it could be.

That's what I meant, the hold-off controls the retrace time but there's a minimum setting. Some 'scopes don't even have a hold-off function so the retrace time is fixed.

I agree, it won't matter at low frequencies but at higher frequencies it will. Whether it's a problem at 1MHz depends on the scope, the one described by the original poster has a bandwidth of 20MHz so 1MHz might be significant. The only way of telling it to measure the waveform against the scale.

I've found the synchronous method to be far more accurate (at low frequencies of course) because it's easier to fine tune than matching up the waveform to the scale.

 

[Nigel Goodwin]

I don't believe that will work in general. You are assuming that the flyback time of the trace is negligible with respect to the sweep time.

It is negligible with respect to sweep time, wouldn't be much use if it wasn't.

 

[Dean Huster]

Retrace is the time it takes for the sweep to return to the left side of the screen. Holdoff is the time that the triggering system is disabled to give the triggering and sweep electronics a chance to "recover" before another trigger is accepted. Holdoff time on some scopes is set and non-adjustable. On other (often better) scopes, there's a front-panel holdoff time adjustment that can be used to help trigger the scope in the presence of a non-repetitive signal such as some digital serial word trains or television signals.

You absolutely cannot correctly adjust or check timebase accuracy by free-running the sweep and adjusting things until the display stabilizes. Trigger the sweep normally at 1ms/DIV sweep speed, apply a known-accurate 1KHz signal and adjust the main timing adjustment (an internal calibration adjustment) for exactly one cycle per division on the screen over the center eight divisions. In other words, ignore the two end divisions on the screen as the specs for most analog scopes do not include these because the CRT is not as linear at the extremes as it is in the center.

Faster times such as 1µs/DIV and faster have one or more adjustments that are made after the basic sweep time adjustment is made.

Dean

 

[Speakerguy]

Do some digital 'scopes have a Fourier Transform function?

Yes, it is a quite common feature, even on the cheap ones these days I think. My oscilloscope has an FT function in the Math menu, and you can adjust its range (MHz/div) by adjusting the time base knob. I was getting some low level noise in my measurements, and the FT mode in my scope showed me a real nice elevation from ~200 to 204MHz on an unconnected probe. Turns out the tower two blocks up the street must be broadcasting channel 11.

 

[Hero999]

Dean,

How come I managed to do it using the mains frequency? I disabled the trigger and synced it so the waveform stood still. It measured exactly 20ms per cycle (50Hz mains).

 

[solis365]

Yes, it is a quite common feature, even on the cheap ones these days I think. My oscilloscope has an FT function in the Math menu, and you can adjust its range (MHz/div) by adjusting the time base knob. I was getting some low level noise in my measurements, and the FT mode in my scope showed me a real nice elevation from ~200 to 204MHz on an unconnected probe. Turns out the tower two blocks up the street must be broadcasting channel 11.

yes being able to FT your data right there in real time is a great boon to analysis. though a digital scopes FT will never compare to a spectrum analyzer (unless you have the REALLY expensive scopes...).

The small digital scope I used (tektronix 2024B) is a pretty solid unit with lots of features. Couple thousand bucks. But the FFT function on it just simply does not compare to the features, accuracy, and noise performance of a 20 year old signal analyzer we ended up getting to perform the measurement.

 

[crutschow]

It is negligible with respect to sweep time, wouldn't be much use if it wasn't.

The scope would still be quite usable even if the retrace time were a significant part of the sweep time (which it likely is at high sweep speeds). You just miss seeing the part of the signal that occurs during the retrace. For example, if the retrace time were equal to the sweep time, you would miss half the signal. But for a repetitive signal, that's not significant and you'd never notice.

 

[Speakerguy]

The small digital scope I used (tektronix 2024B) is a pretty solid unit with lots of features. Couple thousand bucks. But the FFT function on it just simply does not compare to the features, accuracy, and noise performance of a 20 year old signal analyzer we ended up getting to perform the measurement.

Scopes will always be limited in this regard, since I believe most all digital scopes (except for the very high end ones) use 8 to 10 bit ADCs.

 

[solis365]

Scopes will always be limited in this regard, since I believe most all digital scopes (except for the very high end ones) use 8 to 10 bit ADCs.

Yes that's the problem, I think there was a round-off error floor on the FFT at something like only -40dBm. Not so useful when you're trying to look at something in the nanovolt area.

It was also limited by sample size and record length, which severely limited frequency resolution and bandwidth.

 

[Speakerguy]

Yes that's the problem, I think there was a round-off error floor on the FFT at something like only -40dBm. Not so useful when you're trying to look at something in the nanovolt area.

It was also limited by sample size and record length, which severely limited frequency resolution and bandwidth.

Definitely no good for nanovolts, but still useful for many other things. I will frequently look at the FT of the output of a switching converter to get an idea of its quality. But while the noise floor is certainly high, the resolution on my scope is certainly adequate. I think I've got a max record length of 5M samples.

 

[shimniok]

Nice scope! Should be useful for all kinds of stuff! I know mine has been real handy.