How to buy an Oscilloscope?

Yanir

I would like to get an oscilloscope to help with a video based project but i'm not quite sure how to choose one. My basic notion was to buy one that has a bandwidth of 50mhz becuase thats what my SX is running on for this project. But is sampling time more important? Can i buy one that runs at 20mhz but samples in ns?

My project is generating b/w video signals. it's in it's infancy, im just trying to generate a solid white screen, and i am trying to generate horizontal sync pulses, i think it's generating the right signal but the screen isn't showing me what i expect.
Anyway, if anyone can make a recommendation it would be greatly appreciated, price is also a factor. I'm also in nyc if anyone knows of any retailors that sell this equipment.

thanks in advance

yanir

Paul Baker

I have BW video code that is in its infancy, Ill be hapy to post it when I get home. It, at least for my system, correctly produces the horizontal and vertical synce pulses properly. I deliberately used my VCR which is pretty forgiving in timing requirements to make development easier, plus the blue screen is a great indicator when your too far off, so I don't know if it is correct enough to drive a more finicky monitor. I have gotten it to display a black screen, a white screen, a grey screen (theres only 2 bits and value 0 is the sync) and various numbers of vertical black and white stripes. I have coallesed many different sources including the official international standard for what the timings should be (they all seem to have some varience unfortunately). I chose the handout for a college course because of it seems to be in close agreement with the official standard and is really easy to read and interpret, I'll post that as well, the only noticible issue with is is that the 20 lines it specifies as the vertical sync leaves a black band at the bottom, so there is a fudge factor of 8 lines in the code to correct for this. Also the code is in development so there isn't much commenting at this point, Ill see what I can do about inserting some to make it more readable.

Post Edited (Paul Baker) : 3/17/2005 6:54:57 PM GMT

Yanir

Thanks Paul, that would be great, i'm currently using a radioshack rf modulator as my input device, maybe it's pickier than a vcr... something new to try.

Bean

Some scopes actually have a "video triggering mode" for looking at video waveforms.
BTW The schematic and source code from my video display module are available on the website below. Take a look...
Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Check out· the "SX-Video Display Module"

www.sxvm.com
·

pjv

Hi Yanir;

As far as scopes are concerned it's mostly a matter of money. More channels and more bandwidth are always desirable.

For interest, there is a "rule of thumb" most folks are not aware of, and that is the approximate relationship between bandwidth and rise-time.

Typically, and roughly, it works out that bandwidth (in Megahertz) multiplied by the vertical amplifier's rise-time (in microSeconds) equals 0.3

So if you wish to (somewhat faithfully) observe a pulse with a 10 nano Second rise-time (as is the case of an SX), then you would want a bandwidth of 0.3 divided by 0.010 micro Seconds equals 30 MHz.

A calculation like this will help to determine the minimum bandwidth you should have.

As for channels, I believe that for serious work you need AT LEAST two channels plus an external trigger. If you budet is limited, I would sacrifice excess bandwidth for more channels. In the complicated digital word, more channels will be more valuable than more bandwidth.

Another little known issue is probe capacitance. At 10 nano second rise-time speeds the probe capacitance, typically 10 or so pico farads, can seriously degrade the rise-time of a pulse. The best solution for that, albeit an expensive one probably beyond hobby budgets, is to have one special low capacitance probe of one pico farad. They have an isolating amplifier right in the probe tip.

Hope this helps.

Peter (pjv)

Yanir

pjv,

that definately does help, i have a much better understanding of what i need now. thanks

and btw, i switched to my hauppage tv card and i'm getting a much better result, not solid white but mostly white with some weird black lines, easier to debugg this way. guess that radio shack rf modulator was too picky.

thanks all again, everyone is always very helpful.

Paul Baker

One thing, are you using the SX-Key as your clock source? It was recently talked about the skew in the clock signal produced by the SX-key will produce weird squiggly black lines in your signal, try switching over to the resonator to see if that takes care of it.

SteveW

"Some scopes actually have a "video triggering mode" for looking at video waveforms."

True, but it's overrated (or, at least, not a strong influence in buying a scope, for me).
If I want to do video triggering, I tend to knock up a quick test board with a
http://www.national.com/pf/LM/LM1881.html - sync decoder chip, can pull the H and V syncs from pretty much anything. 8 pin package, and takes a couple of capacitors and a resistor. All very convenient, and a very low component cost.
(A SX, with a serial port, allowing me to pick out a given field / line / bunch of pixels is something that I think would be a sweet project, if I ever had the time...)

Steve