Thoughts on an entry level scope?

  In addition to buying new, you might consider: A/D converter technology, sold as a' front-end' package, along with SW for the PC... Or, condsider my fave: buy USED on E-bay. 100MHz Teks go for around $100.-, and 60Mz go for around $60.-.  plus shipping.   For my area, I began by searching by distance to my zip, and then asked the vendor if they would drop shipping/handling fees (-assuming I'm the winning bidde, an paid woth pay-palr), IF I picked up the 'scope at their convenience... this can save you lot of money, and also shift the price-point to your favor since you won't be paying shipping- as would perhaps, the other bidders.  You can get advice as to which used scopes are great deals c/o another post.

We have used DS1000D, which is similar scope but with the logic analyzer capability. Okay, so I've been spoiled by some really good Agilent / Tek scopes...

The price is really tempting on the Rigol and I will say it worked. The fan is a bit noisy, which could be an issue for some applications. The screen is quite small, so the logic analyzer functionality on the 1000D is kind of wasted.

If I were getting a new cheap scope today, I would spring the extra $$$ for an Agilent 2000 series one. Otherwise, I think I would be happier if I could source a used Agilent or Tek. The Rigol had some nice features vs. Tek 1000 series, such as the colour screen. But there are a few useability issues I have with the Rigol so that it wasn't intuitive or a joy to use. I'd rather have a good used 'scope than a crappy new one. (Afterall, everywhere I've worked, I've been presented with "existing/used" equipment).

Or if you want to stay cheap, external USB boxes connected to laptop (such as those from MCC or  DataTranslation) could serve as a reasonable substitute for similar coin, but their capabilities aren't worth it for the $$ in my opinion

Electronics gear is driven by price sensitivity, everyone strives for bang for buck.  If money were never a key parameter you'd get the top of the line of everything, and thus, not worry about whether it will ever meet your needs; as likely there aren't any other gear above it.

While I love the new Agilent Infinivision?? series scopes with the very high waveform/s rate, you're talking about a factor or 3x more cost, to start.  There is always a better tool above what you intend to buy, if you are willing to pay more.  So, its prudent to buy precisely for what you need, then upgrade as the time comes.  Otherwise, the ceiling for costs can be fairly high.

Over at eevblog.com, we discussed these 2 scopes at length, if you wish to read the details.  The Instek 1062A sustains a higher sampling rate when both channels are used, and it has more memory for its memory length, compared to the Rigol.  There are a lot of posts there from owners of both scopes.    If I could do it again, I would buy the Instek, its far better a 60 MHz scope than the Rigol.  The Rigol is infamous for the 100 MHz hack, but this is misleading, since the sampling rate remains unchanged, and using Nyquist limits as a criteria, will only be adequate for 100MHz on single channel usage, using its lowest memory length of 6k.  Reports of folks with scope problems have shown mix results with Rigol's tech support, whereas Instek is more established and has a better reputation for support.

As for other competitors from Chinese makers: Hantek, Tekway, Silgent aka Atten, etc., their price is too close to Rigol to consider, if there are no direct sales vendors in your country, should there be a product defect, returning it would be difficult.

A very workable alternative to the suggested models is the Uni-Trend 2000 series oscilloscopes - Farnell stock them under their Tenma brand, if you do a web search for UTD2062C (60MHz 500smps) or UTD2062CE (60MHz 1gsps) you'll find details of the manufacturers original branded products - 60MHz is the best bang for buck in this range IMHO coming in at around 300UKP - of course there's lower cost options if budget is a real issue (UTD2025C = 25MHz @ 250msps around £230) and at the other end there's the 4000 series (e.g., UTD4102C = 100MHz @ 2gsps around £800) which also have a built in 16 channel logic analyser.

William

http://www.labtronix.co.uk

Rigol is simple, ok for school demonstrations maybe, but you will  quickly find it's limits as your skills improve.  An analogue scope with  delayed sweep would be a better first choice.  GW-Instek is reasonable  and well thought out, but the screen resolution is small.  Vertical  resolution with DSOs is usually 8 bits, needing 256 vertical levels to  properly display.  Most cheap DSOs have only 230-240 vertical pixels, so  will always show jaggy waveforms - not very nice.  Currently, the  Hantek DSO1062B, 1102B and 1202B all offer much better technical  performance and cost (see eBay).  Tekway is identical, but probably more  expensive in Aus.  Neither are perfect, but offer more technical  capability, and you won't find their limits as easily.  The 800x480  screen is much nicer to look at all day.  Most low cost and some  not-cheap DSOs, don't have a compression algorithm to squeeze their long  acquisition memory onto the short width screen - you miss seeing stuff  that might be in the memory.  The only low cost scopes with proper display compression systems are GW_instek, Hantek/Tekway.   Don't buy anything with a short acquisition memory, 10k is an absolute    minimum, and much longer is essential if you are doing anything with    serial ports, USB etc.  When you get some experience, the new Agilent  2000 series have good reviews and professional software, but you pay for  that.  Since it's all digital, the quality of internal software matters  a lot to the usability of a modern DSO.

Hi Sara,

Thanks for mentioning your employer, I appreciate the clarity.  It's also timely, as I have a question about the Tektronix  DSOs you mentioned: Since the MSO/DPO2000C Series have very long  acquisition memory, I was wondering how that memory is handled when it  is output to the screen.  It's easier to propose an example so hard  numbers are being used.  If you set the timebase to 0.1 milliseconds per  division, the total time recorded will be 1 millisecond across the  screens 10 divisions.  At the maximum sample rate, the whole memory  buffer will be filled, or 1 million samples will be taken and stored in  the acquisition memory.  Since the display is 480 pixels wide, and with  typical edge surrounds, it means that the display part of the screen (10  divisions) looks to be around 450 pixels.  This means that 1 million  data points has to be represented by 450 pixels.  How is this done?  Is  the acquisition memory decimated, showing only 1 data point out of every  1,000,000/450=2,222 (approx) points?  Or is there some  acquisition_memory_to_display_compression_algorithm that runs in the  background to make any amplitude variation visible on the normal  screen?  I'm not talking about using waveform zoom to see the pulse, but  when you first set the scope to 0.1 millisecond per division.  As an  example, if there was a single pulse of 0.1 microseconds width somewhere  in the memory, it would occupy about 100 memory words.  If this 100  points fell in between the displayed points, every 2,222 points, would  this always be displayed on the main screen without using zoom?

Thanks, Colin

Hi Colin,

You are exactly right. There is some display compression occurring with more samples than pixels (which happens a lot, especially for long records). As you imagine, long records aren't useful for visual inspection and debug. To get better detail you would use the zoom function. In fact, on the MSO/DPO2000 Series you could use the Wave Inspector controls which allows you to quickly zoom and pan with one set of knobs (kinda like a Digital Video Recorder controller for pausing/playing/fast forwarding TV shows).

One more thing to point out is that part of the display compression algorithm accounts for unique or low duty cycle events. Say you have a clock that has a glitch appear this glitch would be more prominent while the other normal cycles would be "blended" into each other. This is also true with MSO digital channels where glitches, runts and other signal anomalies are highlighted in white. Again, it can be difficult to observe events without zoom but we do try to make easier with some built-in display tools like highlighting and weighted signal display features. I hope this helps.

~Sara