Electronics > Beginners
Looking for a oscilloscope
shrubbie:
I've been reading a lot about oscilloscopes and recommendations, actually about every thread here.
But I still can't decide what oscilloscope to buy.
All the threads seem to have same problem, no one defines what they expect from the scope so you'll only get the general stuff.
So, here we go:
Requirements:
- Power supply design/debugging
- Audio amplifier design/debugging
- µC stuff (I'll probably try first to build somekind of LA from scratch first, so not really requirement at this time)
- Possibility to dump data to PC for analysis
- Detect relatively small delays between signals (measure speed of light stuff)
- And as I'm studying physics I'd also like to do some semi-scientific measurements of everything I think off (just for own amusement)
Basically general engineering and measurement stuff.
And before you recommend old analog scopes: already looking for one, but one just can't have enough gadgets.
I've been looking at rigol but the small screen really turns me off.
Another alternative could be owon, but the inner nerd screams for agilent, but I'm having hard time justifying the price of agilent.
So is agilent really worth the investment and why?
I'm most interrested (2000 series) in 2x100MHz (possibly 4x100MHz or 2x200MHz, but which if I decide to invest more?).
With agilent you seem to get free wavegen if you buy before 2012-08-31, so that's a plus.
I know I'll probably get by for a while with old analog scope and cheap rigol/owon, but I'd really like scope that'll last.
Basically I'd like to know what more I get with agilent (or possibly other brand) than with cheap rigol/owon,
and whether or not it's worth it as an serious hobbyist (and perfectionist).
Budget could be around 1500€, cheap is good but man's gotta do what a man's gotta do..
quantumfall:
I'm no expert so this is just me talking "Budget could be around 1500€, cheap is good but man's gotta do what a man's gotta do" Yeh you can justify it like that and I've done that myself with cars before and yes they are fond memories as well so it is not just about the cost.
You could start with a basic Rigol, Owon or Hantek the cost is a lot less and you could move on to an Agilent later.
I do not have a DSO but am considering a basic Owon , Hantek etc only for cost reasons, but the Agilents are a class act, in the UK you can get the 2000 series with a free signal gen upgrade untill August I think.
If you are going to use it and only want to buy once my vote is for the Agilent, but thats not based on experience of them just a practical concern for support and quality build over the long term.
Measuring the speed of light, well I'm not sure how close you will get with a scope, is that practical ?
DonRon:
Hi shrubbie,
what about a combiscope, for example the Hameg HM1008 http://www.hameg.com/0.357.0.html
Both worlds (analog + digital. scope) in one instrument. Okay - the digital part cannot competite with the Agilent, but it might be a first good scope capable to do a lot of different stuff.
And looking at the audio amplifier stuff the analoge scope might be much better then the digital.
I have the HM1008 for several weeks now and I am really satisfied. It is also very interesting to compare analoge and digital.
I use it for power supply design, audio stuff and also ham radio stuff.
I knew it will not be that scope for the rest of my life, but it will help me for the next few years. During that time the modern DSO's will getting more and more powerful at lower cost (i hope so).
And watch the new Rigol "low cost" scopes which will come in a few month. They will be much better then the actual DS1102 series.
So cheers for now,
Ronald
ejeffrey:
--- Quote from: shrubbie on April 05, 2012, 07:03:48 pm ---- Power supply design/debugging
--- End quote ---
A true differential probe would be pretty useful here. A current probe would be nice, but they are _expensive_.
--- Quote ---- Audio amplifier design/debugging
--- End quote ---
Almost any scope will do. You will also want a function generator. Audio work could take advantage of a high dynamic range ADC, but there aren't many options for 12 bit scopes. The most common high dynamic range scopes are the low speed USB scopes, but they are pretty limiting for general purpose use. Ideally you would like to have a good FFT/spectral analysis mode, but that is pretty much garbage on every scope I have ever seen, so no point worrying too much about it.
--- Quote ---- µC stuff (I'll probably try first to build somekind of LA from scratch first, so not really requirement at this time)
--- End quote ---
More than two channels, serial protocol decode, and an MSO option would be handy here, but not critical if you are going to use a logic analyzer.
--- Quote ---- Possibility to dump data to PC for analysis
--- End quote ---
Pretty much any DSO should support this, but if you get an older low-end unit it may be an optional feature, and older equipment of any stripe may be GPIB only. Newer scopes will have USB, ethernet, or serial, and will usually be able to write waveforms to a USB flash drive as a CSV and/or screen capture, although some of them (I am looking at you, TDS2000) are unbearably slow. I rarely use the manufacturers client software, I generally program it directly or use the LabView instrument drivers from the NI dev zone (be warned, some of those are pretty buggy).
--- Quote ---- Detect relatively small delays between signals (measure speed of light stuff)
--- End quote ---
Any scope should be able to do this. A scope with bandwidth B will have risetime 0.35/B, and you should be able to resolve by eye ~ 1/4 risetime, better if you capture data and fit it. Obviously more bandwidth helps, but if your only goal is to show the propagation speed, it is cheaper to use a longer cable :)
--- Quote ---I'm most interrested (2000 series) in 2x100MHz (possibly 4x100MHz or 2x200MHz, but which if I decide to invest more?).
--- End quote ---
Nothing you are doing appears to be bandwidth intensive. At least if the micros you are interested in are things like atmel/pic/msp430, and low speed serial protocols they typically operate at a few MHz. On the other hand, 200 MHz is not fast enough for ARM and high speed USB. For digital systems you don't really need the analog bandwidth to capture 10 harmonics, although you do want the sample rate that high to avoid aliasing. A 500 MS/s or 1 GS/s scope with 50 MHz bandwidth will do fine. For analog work like audio and SMPSs you want the bandwidth to see the harmonics and to make sure your circuit isn't oscillating above its operating frequency, but 50 MHz is enough to get you the 10th harmonic of any audio DAC sample clock and most switch-mode supplies. I suspect that there are people in the world who have made audio band circuits oscillate in the UHF, but that takes a special kind of talent.
I love me a 4 channel oscilloscope, but some people don't see the point.
shrubbie:
--- Quote from: quantumfall on April 05, 2012, 07:34:53 pm ---If you are going to use it and only want to buy once my vote is for the Agilent, but thats not based on experience of them just a practical concern for support and quality build over the long term.
--- End quote ---
Indeed I was looking for a longer time investment. But there's no point paying for agilent if cheapo DSO can do everything I need.
Or in other words agilent (or other name brand) doesn't offer any cool bells and whistles I feel I must have.
So the real problem is does agilent offer something cool/critical/"any other excuse here" that cheapo DSOs doesn't.
--- Quote from: DonRon on April 05, 2012, 07:51:30 pm ---what about a combiscope, for example the Hameg HM1008 http://www.hameg.com/0.357.0.html
--- End quote ---
Specs look fairly good, but I'm afraid that I'm in love with agilent.
Though you never know, have to keep options.
--- Quote from: ejeffrey on April 05, 2012, 08:20:49 pm ---A true differential probe would be pretty useful here. A current probe would be nice, but they are _expensive_.
--- End quote ---
Could one just make those? Or probes in general?
Would be fun to see how good probes one could make and even try to beat some commercial ones.
--- Quote from: ejeffrey on April 05, 2012, 08:20:49 pm ---Almost any scope will do. You will also want a function generator. Audio work could take advantage of a high dynamic range ADC, but there aren't many options for 12 bit scopes. The most common high dynamic range scopes are the low speed USB scopes, but they are pretty limiting for general purpose use. Ideally you would like to have a good FFT/spectral analysis mode, but that is pretty much garbage on every scope I have ever seen, so no point worrying too much about it.
--- End quote ---
That's the reason I was looking at the agilent, I'd get basic wavegen and 12bit high resolution mode.
Dunno how usefull that 12bit mode is in real life though..
Would it be possible to record a good chunk of data directly to PC and do FFT on PC?
And would it even make a difference?
I was though toying with the idea that I'd use my old X-FI soundcard as an spectrum analyzer, but we'll see.
--- Quote from: ejeffrey on April 05, 2012, 08:20:49 pm ---Any scope should be able to do this. A scope with bandwidth B will have risetime 0.35/B, and you should be able to resolve by eye ~ 1/4 risetime, better if you capture data and fit it. Obviously more bandwidth helps, but if your only goal is to show the propagation speed, it is cheaper to use a longer cable :)
--- End quote ---
Indeed, measuring speed of light is relatively easy and cheap if you've got long cables.
I was thinking more about how fast transient pulses I could detect and how small delays between them.
Or possibly use two channels and detect glitches and delays between them?
For example agilents datasheet says it's able to detect glitches as narrow as 500ps,
so could one use for example two phototransistors measure speed of the projectile?
Prolly stupid question since I'm pretty sure that it's doable, but could one use same method for something faster like delays between radiolinks?
or in general one could detect and measure delays between 500ps long pulses?
--- Quote from: ejeffrey on April 05, 2012, 08:20:49 pm ---Nothing you are doing appears to be bandwidth intensive. At least if the micros you are interested in are things like atmel/pic/msp430, and low speed serial protocols they typically operate at a few MHz. On the other hand, 200 MHz is not fast enough for ARM and high speed USB. For digital systems you don't really need the analog bandwidth to capture 10 harmonics, although you do want the sample rate that high to avoid aliasing. A 500 MS/s or 1 GS/s scope with 50 MHz bandwidth will do fine. For analog work like audio and SMPSs you want the bandwidth to see the harmonics and to make sure your circuit isn't oscillating above its operating frequency, but 50 MHz is enough to get you the 10th harmonic of any audio DAC sample clock and most switch-mode supplies. I suspect that there are people in the world who have made audio band circuits oscillate in the UHF, but that takes a special kind of talent.
I love me a 4 channel oscilloscope, but some people don't see the point.
--- End quote ---
So 100MHz is prolly more than enough for my use.
I can see how 4 channels would be nice but on the other hand I can see my self doing fine with just 2.
Can't really decide which way to go.
Navigation
[0] Message Index
[#] Next page
Go to full version