Which oscilloscope?

[Gibson486]

Rigol DS1102E or Instek GDS1102U....

I'd get a textronix 2000 series, but i do not have the bling to get one and $400 is pretty much the max i would spend anyways on this. I would buy an old one off of craigslist, but they only have old analog ones that are just way to big....

 

[William Gaatjes]

What are you planning to use the oscilloscope for ?

That is the most important question.
These devices always come with a limitation.
But that limitation would be non existent depending on what you are going to use it for. Bandwidth limitations, internally generated noise. Fast acquisition of samples but slow lcd. All kinds of problems.
Oscilloscopes are expensive for a reason.

Best thing is that you start thinking about what you want and then read about the limitations. If the limitations are not an issue, you know what to buy.

http://www.rigolna.com/products/digital-oscilloscopes/ds1000e/ds1102e/

http://www.gwinstek.com/en/product/productdetail.aspx?pid=3&mid=7&id=1299

The rigol gives me a better impression from a quick peek at the specs given.

Rigol honestly states that the samplerate is only half in 2 channel mode. Which of course makes sense in this price range.

I think it is best that you compare all specifications to what you want.

 

[Gibson486]

I'm aware of the limitations. For bandwidth, I am not going to be anywhere close to the max. I am just going to be using this for general stuff.

 

[William Gaatjes]

What i always find handy is the ability to make screenshots.
To do a measurement and then save it upon an usb memory stick.

The rigol seems better suited. It has more memory to store data from a acquisition. Higher sample rate. All digital oscilloscopes use staggered ADC's.
This means that there are 4 adc present that are each sequentially activated. If these are 250MS each, you would get an effective samplerate of 4*250MS.
A 4ns conversion time would be turned into a 1 ns conversion time. This asks for a very critical clocking system with extremely good phase control. Could be 4 clocks at 250MHz but 90 degrees shifted apart.


I also do not understand this from the gwinstek website :

INTERFACE
USB Slave Connector USB1.1 & 2.0 full speed compatible (printers and flash disk not supported)
USB Host Image (BMP), waveform data (CSV), and setup (SET)

It has a slave usb connector but cannot use memory sticks ?
What would that slave usb stick be used for then ?

EDIT:
Nevermind. I got confused. Of course the usb slave cannot be used for memory sticks. The oscilloscope must be an USB host in order to do so.

 

[William Gaatjes]

From reading , i have found a rumor that Rigol makes the low end digital scopes for Agilent. When it comes to Agilent oscilloscopes, i am sold. We have 2 of those 4 analog channel + 16 channel logic analyzer at work. Great instruments to work with.

I would buy a Rigol based on that rumor.

But at the moment i am always proud to have this one at home and will never replace it, although i would buy digital oscilloscope as extra instrument when i have solved some current matters :

 

[William Gaatjes]

Last spam post but fun to read :

http://www.eevblog.com/forum/projec...l-stuff/tektronix-465-repair-and-restoration/

Look how happy it is after being repaired :

No microcontrollers were used , just using the delayed sweep on a square wave together with a sine wave signal.

 

[Gibson486]

Yeah, they have lots of analog ones on craigslist for $100. Thing is, my desk place is very little...Also, I think the USB would be good. i have always used that feature to capture waves. As most with EE's, I have only used Textronics stuff because it is what they have at school and what all big corporations use.

 

[Jarhead]

Being an EE, I've used Agilent/HP, Tektronix, LeCroy, Hitachi, my Dad's BK Precision, and a variety of others. I'd always take an Tek over an Agilent/HP, unless I was doing just digital stuff. That said, LeCroy makes some nice scopes. I'd avoid the Tek 22xx series, unless they were cheap. The storage function for capturing one time events(like glitches) is very useful, otherwise the analog phosphor scopes are just as handy if you have a digital camera, as you can take a picture (some macro cameras one can full screen and get much higher resolution than you will get with an LCD. But a clever hint, put the camera on movie mode, and now you have the storage function.

Those old analog scopes will give you much more for your money, and you won't get the same quality/functionality until you go to a true analog modern scopes that cost more than modern cars.

If you go the old route, one feature I'd make sure an old oscilloscope has is that the scope does have a triggered sweep.

Once you get into the 30,000.00 range, the Agilent/HP scopes can capture what you easily see on a Tek, but you have to go and set various options in the menus- and no, you have to set up things differently for different signal types.

Most all digital scopes can save the screenshot as an image file, and one can save waveforms as .csv, if you'd like to drop the waveforms into Excel for further processing, or creating your own graph.

Keep in mind that an oscilloscope's bandwidth is the point at which a signal like a sine wave, at that frequency, is 3dB down, or only shows 70% of the actual signal's amplitude. If your digital signal edges are faster than the reciprocal, the signal's rise and fall times will be stretched out (be really slow).

For digital work you want a scope that is much faster than the edges you need to look at, take a look at the apparent 100MHz sinewave, which is actually a rectangular looking square wave, and even the 500MHz scope makes the clock signal look really messed up:
http://gpete-neil.blogspot.com/2011/10/determining-how-much-oscilloscope.html

IMHO for digital, to really see what is going on, you need a scope with a Bandwidth that is 10x higher than what you need. If you are designing or troubleshooting a design, this is especially important, otherwize you can get away with just 5x the bandwidth.

However if you are just doing lower speed hobby stuff, one can turn their sound card into an oscilloscope, or even go to some of the PC USB/PCI based scopes. Be sure to look at the independent reviews on the USB scopes, there are some really killer ones that will give you 10x the performance of a similar priced stand-alone scope. Tipically

This is decent beginner primer on oscilloscopes for the first time buyer:
http://www.radio-electronics.com/info/t_and_m/oscilloscope/buying_scope.php

I looked at some reviews of the low end scopes, people are saying the Rigol and Instek have a big problem with the interleaved ADCs that introduces copious amounts of timing jitter (if you are timing sensitive, this could be a gotcha).

If you are primarily looking for a logic analyzer, you get a lot more bang for your buck with one of the USB based devices.

 

[Modelworks]

I love the 465 scope, one of the best tektronix ever made. I use a BK precision scope a lot of the time as well. Both are analog but I don't use scopes for digital work, for that I use logic analyzers.

If you are just interested in working with digital data , then one of the best is the ones from salae. They start at $149.
http://www.saleae.com/logic

If you want to capture low frequencies then a pc sound card can do it. I say sound card and not regular pc audio jacks because you really need to remove the input capacitors to whatever chip the card uses, otherwise you will get poor results.

Lastly there is the DIY analyzers like the open bench one for $50. It is very capable but you have to provide the case, wires, etc.
http://dangerousprototypes.com/docs/Open_Bench_Logic_Sniffer

 

[harrkev]

I actually work for Agilent so I may be biased ;-) I helped a little on the ASIC that is in the latest DSOX2000 series, so I can recommend those... But they are a little out of your price range -- they start at around $1200, but feature a built-in function generator (pretty cool stuff) -- but that is an add-on that cost more money. Save up for one.

Now, since your price range IS limited, you have a few options. If you really do not NEED screen shots, I happen to really like the old Tektronix analog scopes. Other have mentioned the Tek 465. If you can find one in great shape in your price range, that might be an option. The layout of the buttons and knobs is confusing at first, but completely awesome once you get used to it. Also, analog scopes have an awesome update rate (see below). If I only had $400 or so, I would probably rather have a nice analog scope over a cut-rate digital one.

** BEGIN EDIT **
I really DO love the Tek 465. Unlike analog scopes, everything you need is on a knob or button right at your fingertips. No menus, trying to remember which menu has the option you are looking for, etc. Once you get use to it, you can play that thing like a violin. They really are a pleasure to use, given the limitations of being an analog scope, and being old and having limited bandwidth.
** END EDIT **

One thing to keep in mind is the ***update rate***. Let's assume that you are running the scope VERY FAST trying to capture a very short period of time. For example, you are trying to capture an event that happens every 1000 ns. This means that each second, you could, in theory capture ONE MILLION events every second. Now, lets assume that you are trying to capture a glitch that only happens about every second (around once every million events). How likely are you to see this?????? With an analog scope you stand a pretty good shot. One of the advantages of a good analog scope is that they can re-arm very fast. Now, let's assume that your digital scope that actually grab and display 100,000 waveforms every second. You would probably notice the glitch after 10 seconds (50% chance after 5 seconds, but the chance is close to 100% at 10 seconds). With a 10,000 waveforms/second scope, it would take around 100 seconds. If your scope can do 1,000 waveforms/second, it would take around it would take 1000 seconds (16 minutes). For just looking to see what a waveform looks like, update rate does not really matter. If you are looking for occasional glitches, it really matters. Note that Agilent-made scopes really excel in this department. In the range of $1200 to $3000 scopes, Tek cannot even come close to touching Agilent in this specification.

Be cautious buying/using low-end digital scopes. Many of them have MAJOR aliasing problems. Not a deal breaker, just something to be aware of. Put in a sine wave of some frequency (say 1 MHz, for example), and then start dropping your sample rate. At some point (with almost all cheap scopes) if you sample at less than the Nyquist rate (in this case, 2 MHz), what you see on the screen is NOT what the scope is actually seeing. Note that every agilent-made scope does NOT suffer aliasing problems (unless you explicitly go through menu options to try to get this problem back). I cannot speak for other manufacturers, but I suspect that the other major brands (Tek, LeCroy, etc.) have probably taken care of this problem too.

I have used PC-based scopes before, and I have seen problems with the front-end. For example, setting an offset (for example, trying to zoom in on a 1V AC signal riding on a 10V carrier, without going into AC mode), the shifting is done digitally, so it is easy to over-run the ADC, so the top of the waveform gets chopped when you zoom in vertically.

There are a **LOT** of things involved in making a scope do the *RIGHT* thing under all circumstances. It is a lot harder to get right than you would think. New digital scopes can be cheap because they take shortcuts -- update rate, aliasing, triggering, DSP, etc. You just need to decide how important all that stuff is to you. How important is it to your if your scope is lying to you (even just a little bit). If you are a hobbyist doing this for fun, then you can probably get by with less.