Do oscilloscope graticules have to be square? What's your opinion.

[Neganur]

So, I sold off my analogue scope and use an Agilent 3000X series now. With the very first screenshot I take, I notice that the grid isn't square - big deal?


I don't know. I'm used to a perfectly square graticule from the old Tektronix and I have this strange idea that curves don't 'look' as they're supposed to. For example a perfect sinusoid just looks...weird, like someone elongated the graph to emphasise or falsify data. The dimensions of the grid is ca 1.5x1.1 mm and I wonder why it isn't square. It probably doesn't matter much and it's just something I have to get used to.

I asked this on the Agilent forum too, but the guy doesn't seem to understand me.

If you look at other brand's oscilloscopes, Rigol, older Tektronix and some Hameg have it square, R&S RTOs and all newer Agilents seem to have it rectangular.  The grid of the 3000X  is actually switched to a square shape when you change the into XY-mode, but according to the guy in the Agilent forum "nobody uses XY-mode anyway", which I think is a really dumb reasoning.

Enough of my  talk, what do you think - should it be square like engineering paper or is this a blast from the past and cursors & auto-measurement readouts do the job today?

[grumpydoc]

For example a perfect sinusoid just looks...weird.

Why? If you stretch a sine horizontally on screen it has the same appearence as one of a lower frequency but it's still a sine wave.

[wraper]

I don't know. I'm used to a perfectly square graticule from the old Tektronix and I have this strange idea that curves don't 'look' as they're supposed to. For example a perfect sinusoid just looks...weird, like someone elongated the graph to emphasise or falsify data.

Yes, your idea is strange. Just probe 1,5 times higher frequency sinusoid on the same time/div setting. It will look "right".

[kg4arn]

IMO it only needs to be square in XY mode.

[rsjsouza]

Despite it does not matter for precision or measurements, I agree with the OP. You can blame on subjectivity or optical illusion, but I prefer a perfectly squared grid.

[AndyC_772]

It's just a regular 10 square grid, all that's changed since your Tek is that the aspect ratio of the screen is wider. It could be argued that adding more squares would have made more sense, but it's an entirely arbitrary decision either way.

The idea that a sine wave looks 'weird' or 'stretched' makes no sense to me at all, I'm afraid. The horizontal time base is entirely up to you, there's no 'right' or 'wrong' scale to use in the time axis.

In X-Y mode, the units on the two axes are the same, so it's reasonable to expect the grid to be square - though I can't say I'd ever even noticed it being redrawn in this mode until now.

[Neganur]

I guess it's hard to explain a subjective sensation. A square does not look square when you make it a rectangle. (sounds dumb, I know...implying, that I looked at square paper all my life)

I'm also not saying that it's not allowed to be rectangular, but I was surprised how much it seemed to bother me.

If it doesn't matter as you say, why are all those old instruments perfect squares? (because it was invented on a Friday! ) Is it maybe because that's how the landscape-shaped instruments screens have changed our preference nowadays? Why did Agilent not just make it 12 x 8 divisions (10x8 division seems to be the classic screen size, which they did stick to). 

[jpb]

My WaveJet has square graticules 50x50 pixels (500 pixels wide and 400 pixels high for 10x8). I like it but can't think of any good reason why it is better than rectangular ones.

It would probably make more sense to have a more direct mapping to the 8bit dynamic range of the A2Ds, i.e. have it 256 or 512 pixels high or perhaps 500 pixels high with a bit of overlap.

Having said that, I noticed the other day that the dimensions change if you have measurements displayed (that is rise times and so on beneath the graph)- so perhaps it is only square sometimes.

[madires]

The best solution would be to make it selectable in the menu and to also support 2-3 sizes, like small, normal and large. That should make most EEs happy.

[grumpydoc]

The best solution would be to make it selectable in the menu and to also support 2-3 sizes, like small, normal and large. That should make most EEs happy.

Possibly, but the basic effect of that is the same as switching from one sensitivity to another (or timebase). It would probably be worth offering a graticule line every 0.5 division as well as every division.

I have to say regardless as to whether it makes sine (on any other) wave "look wierd" I would have thought going from a 10x8 to 12x8 screen just sticking with square divisions would have been a lot easier.

[Hypernova]

If it doesn't matter as you say, why are all those old instruments perfect squares?

It's a carry over from the days when you had the grid printed on the screen. If you want to support XY mode there you have to stick with a square grid.

Unless you're OCD when you make a plot of two different units in excel do you go the extra mile and tweak it until it's square?

[Mark_O]

why are all those old instruments perfect squares?

It's a carry over from the days when you had the grid printed on the screen. If you want to support XY mode there you have to stick with a square grid.

   Bingo!  We have a winner.

[zapta]

Square graticules are left over from the times oscilloscopes had round CRTs, just like the diagonal screen size specification we still use.