Another Scope + Logic Analyzer Question

[Electro Fan]

If you had a choice between an (Agilent 2000) scope with 2 analog channels + 8 LA channels or an (Agilent 2000) scope with 4 analog channels but no LA channels, which would you prefer for a combination of general scope work plus occasional I2C work?  (Assume in both cases you had the DSOX2EMBD I2C decoder.) Thanks

[Smokey]

4 analog channels.  All day long.  But that's just me.

[dfmischler]

Me too.

[benemorius]

Depends on your needs, but consider this: Get 4 analog channels now and you can add the 8 digital channels later; get 2 analog channels now and you'll be stuck with 2 analog channels for the life of the scope.

I cringe every time I see a 2 channel 2000X/3000X on youtube or in lab pictures. I can associate very well with having insufficient funds, but this is one shortcut that will haunt you relentlessly for many years to come. Unless you never get very far into electronics, you WILL want 4 analog channels.

For general work plus occasional I2C work as you say, it sounds like you can easily get away with just 4 analog channels.

Don't forget you can also get a cheap USB logic analyser/sampler for <$100 USD if you want the convenience of having lots of channels.

[echen1024]

I2C is just 2 wires, SCL and SDA, so I would just go with a 4 analog channel, and add the LA later on if you need it.

[Berni]

Go for the 4 channel version. When debugging more more comple stuff I often find my self using over 2 channels, besides 4 channels is enough to look at SPI and the scope will do serial decode from analog channels if you have that feature. Buy a USB logic analyzer for digital for a few hundred bucks. I prefer those due to my pc having a larger screen on a pc and a mouse.

[echen1024]

Exactly. The 4 channel is much more important than the LA. It can be added on later, plus a whole shitload of other funtionality.

[Electro Fan]

Hey, Thanks Everyone!  In baseball terms it looks like we might be pitching a shutout:  something to zero, in favor of 4 channels.

I'd still be up for hearing from anyone who has a preference for 2 channels (or at least thinks you could live with 2 channels while doing some basic scope work including some I2C projects).  Not that I think 2 channels is really highly preferrable, but here's why:

If we need 4 channels we rule out an Agilent 2002A (which seems pretty cost-effective), and we also rule out a Rigol 2072 which is very cost-effective.

One of the reasons for even considering/liking the Agilent 2002 (or any Agilent) so much has been the ability to add the built-in LA (now or later). 

But if we decide that a built-in LA might not even be needed for I2C/SPI because we can get by with the analog channels, but only with 4 channels, then we are pushing Rigol to a 4000 series.  Kind of doubles the price (almost triples the price for Rigol).

Then, once we push up into the $2 - 2.5k range it opens the idea of "well, since we only want to do this once, how about an Agilent 3000?"  Yowza

- Price before add-ons (and maybe some discounts):

Can't buy one of these 'cause you guys don't approve of 2 channels?
Rigol 2072 $839
Agilent 2002A $1258

So now we need one of these:
Agilent 2004A $1916
Agilent 2014A $2246
Rigol 4014 $2399

And once we get to about here, we look at:
Agilent 3014A $3816

Then we check our sanity.

It seems like scopes don't change as fast as computing and networking equipment (instead of 1-3 year product cycles scopes feel more like 3-5 year product cycles).  If scopes did evolve a little faster then it would make sense to live with a Rigol 2072 for maybe 2-3 years and and then see what comes down the pike, but when you start thinking about owning a piece of equipment for 5 years or more you'd kind of like to get the one you want.

Rigol really should figure out how to introduce a 4 channel version of the 2000 for about $1499 with a $400 LA, and then offer a bundled package with the scope, LA, and decodes for $1999

The practical answer is to live with 2 channels on a Rigol 2072 with the SD-DS2 for about $1k but maybe practical isn't the right answer

[true]

This is EEVBlog forums, where you are advised that you should be spending $30k on scopes and other test gear yearly.

Based on what little you said you don't need the LA.

Keep in mind you originally did not give the option of 2ch alone, only 2ch+LA or 4ch.

[benemorius]

Here's the thing: the 4 channel recommendations are primarily intended to look after your long term interests and your bang per buck. If you aren't convinced that you're still going to be using the scope regularly 5 years down the road, these recommendations carry little weight. If you're willing to consider a bottom of the range scope that doesn't cost much and doesn't even have a 4 channel version, the recommendations again carry little weight. The primary objection is to spending a whole bunch of money on a scope and then not spending a little more to get the 4 channel variant.

I completely agree with your "what about an Agilent 3000X" reasoning, even if you had intended to hold it up as a negative thing. If you can afford it and if you're damn sure that you're going to be using the scope weekly 6 years from now then I'd insist that you'll be happier with that than with anything else you're considering. It's a slippery slope and I'm not sure how much of a case there is for stopping mid-way down it if you KNOW you're serious about electronics.

It sounds, however, as though you may not be so certain about these prerequisites. If that's the case then I'd advise a different strategy: try to get away with as little as you can until your certainty goes up. From what you've said, you can easily get away with using a $400 Rigol, or even a $50 analog scope plus a $100 USB logic analyser. Later on, if you decide you still want to, you can get a much nicer second scope. If that happens, you will not regret having two scopes and you will not look upon the first one as a wasted purchase.

This is EEVBlog forums, where you are advised that you should be spending $30k on scopes and other test gear yearly.

That's silly. See above. Recommendations from seasoned professionals here are by default going to be based on the assumption that the inquirer is asking for advice on how to skip over the lesser, more regrettable purchases and jump straight to the stuff that you'll be happy with for most of the rest of your life. If that assumption isn't correct, you have to explicitly say so upfront.

[Electro Fan]

This is EEVBlog forums, where you are advised that you should be spending $30k on scopes and other test gear yearly.

Based on what little you said you don't need the LA.

Keep in mind you originally did not give the option of 2ch alone, only 2ch+LA or 4ch.

You are correct, I didn't suggest the option of a 2ch alone.  I think I would like a LA, but I'm very sure I'd like to look at digital signals as well as analog signals - and what I'd really like to be able to do is to observe the relationships between analog and digital signals, as precisely in time as is reasonably possible.  It seems like just 2 channels would be marginal, but if I could get the job done with only 2 channels and save a $1k - $2k it would be something to consider.  I'm still learning about the alternatives and trade-offs - and considering what the right short term and long term solution(s) might be.  I hear you about guys liking to recommend stuff here on the EEVblog Forum but I think most of the recommendations are done with both good intentions and based on good experience.  Some people here have more knowledge and experience than others, but I don't mind any advice (it's not like we're paying a lot for the advice we get here ) - and definitely it's fun to shop.  Shoping for gear might be almost as much fun as using it.  So, I don't mind shopping and I don't mind anyone enjoying shopping on my list .  And I definitely appreciate everyone here who takes the time to help with information and suggestions.  Thx!

[MrAureliusR]

You are correct, I didn't suggest the option of a 2ch alone.  I think I would like a LA, but I'm very sure I'd like to look at digital signals as well as analog signals - and what I'd really like to be able to do is to observe the relationships between analog and digital signals, as precisely in time as is reasonably possible.  It seems like just 2 channels would be marginal, but if I could get the job done with only 2 channels and save a $1k - $2k it would be something to consider.

Unfortunately, observing the relationship between analog and digital signals using 2 channels on a digital scope isn't really going to happen. If you really want to do that, maybe look at buying a used MSO on eBay. You'll probably pay ~$1000 or less and you'll get the functionality you want... 2 or 4 analog with 8 to 16 digital channels... Check this one out, for example: http://www.ebay.ca/itm/Agilent-HP-54645D-100-MHz-Mixed-Signal-Oscilloscope-with-54650A-/360698318363?pt=BI_Oscilloscopes&hash=item53fb4b8e1b&_uhb=1

Of course, the downside to a scope like that is no built-in decoding of serial protocols. I don't know about handing the data to a PC via RS-232 or GPIB, or if that's even possible in (near) real-time to decode, but hey -- at least you can watch both signals on the screen and see the relationship a lot better.

[Electro Fan]

Here's the thing: the 4 channel recommendations are primarily intended to look after your long term interests and your bang per buck. If you aren't convinced that you're still going to be using the scope regularly 5 years down the road, these recommendations carry little weight. If you're willing to consider a bottom of the range scope that doesn't cost much and doesn't even have a 4 channel version, the recommendations again carry little weight. The primary objection is to spending a whole bunch of money on a scope and then not spending a little more to get the 4 channel variant.

I completely agree with your "what about an Agilent 3000X" reasoning, even if you had intended to hold it up as a negative thing. If you can afford it and if you're damn sure that you're going to be using the scope weekly 6 years from now then I'd insist that you'll be happier with that than with anything else you're considering. It's a slippery slope and I'm not sure how much of a case there is for stopping mid-way down it if you KNOW you're serious about electronics.

It sounds, however, as though you may not be so certain about these prerequisites.

Sorry I wasn't able to be more clear but I think you are pretty close to dialing-in my questions and thinking.  My questions are driven by the fact that I don't know exactly what I'm going to do or need to do or want to do with the scope initially, and even less sure what I will wish I could do with it later. 

For the record I have a 100 MHz 4 channel Tektronix analog scope and I enjoy it a lot but so far I'm mostly teaching myself and doing small projects; definitely it's not something I use to earn a living.  But every day I feel like I'm learning more and I've come to believe that test equipment is a fantastic "tool box" and "lab" with which to learn by trial and error.  I study, hypothesize, set up simple tests, learn from the results, and keep moving.

I suspect that if I could spend a day with each of these scopes it would be a lot easier to know which is the appropriate tool.  It's kind of like reading about a Honda vs a Lexus vs a BMW, at some point you can only get so much from reading the manufacturers' spec sheets and manuals, and from reading reviews; a test drive would be pretty enlightening, I'm sure.  In lieu of a test drive it's great to be able to frame questions (however marginally well) and get answers from people who share the passion. 

You are definitely correct that I'd like to avoid regretting having two (digital) scopes and look upon the first one as a wasted purchase; and I'd like to enjoy the first one from the get-go - but without having a test drive it's hard to know whether a $1k scope is sufficient or a $2.5k scope might be penny wise but pound foolish or a $4k scope is a great investment, or just dumb (for me - for others I'm certain it is a tremendous scope).

So, I'm trying to break the problem down into piece parts - one of which is how many analog and/or digital channels makes sense.  If we didn't have the Rigol 2072 anchoring the low end of the pricing I'm pretty sure I'd be saying just go with an Agilent 2004A because from there almost everything can be added incrementally without much of a premium for adding it later than sooner.  And from there we get to "just get it over with and go for an Agilent 3014A..." Did someone say Ground Hog Day?  Thanks for all the advice - it's all welcome!

[Electro Fan]

Unfortunately, observing the relationship between analog and digital signals using 2 channels on a digital scope isn't really going to happen. If you really want to do that, maybe look at buying a used MSO on eBay. You'll probably pay ~$1000 or less and you'll get the functionality you want... 2 or 4 analog with 8 to 16 digital channels... Check this one out, for example: http://www.ebay.ca/itm/Agilent-HP-54645D-100-MHz-Mixed-Signal-Oscilloscope-with-54650A-/360698318363?pt=BI_Oscilloscopes&hash=item53fb4b8e1b&_uhb=1

Of course, the downside to a scope like that is no built-in decoding of serial protocols. I don't know about handing the data to a PC via RS-232 or GPIB, or if that's even possible in (near) real-time to decode, but hey -- at least you can watch both signals on the screen and see the relationship a lot better.

I am starting to get the opinion that the opinion here is that 2 channels is wishful thinking, so if it needs to be 4 channels then so be it.  I am pretty sure that I would like to have the serial decoding capability built-in - it is part of one of the key projects I have in mind.  The only reason this is even slightly hard is the Rigol 2072; you could purchase the thing for $1k with the I2C/SPI and if it sufficed, great; if it didn't and you couldn't sell it for a nickle you'd be out $1k if you never used the scope again.  But you'd probably either find a use for it or sell it used for $400-$500 in a couple years.  I don't think there are any bad choices here; it's just a matter of figuring out the best choice.  The more I know about the differences between the scopes under consideration the more I'm likely to make a good choice, I think .

[true]

I am starting to get the opinion that the opinion here is that 2 channels is wishful thinking, so if it needs to be 4 channels then so be it.

I do primarily digital stuff and DS2072 is more than enough for me right now. Have LA but haven't even used it yet.

Your choice just depends on what you are doing. Since you don't know what you want to do, I don't know what to say you should do.

You did mention you have a 100MHz Tek; if that's enough for your analog needs why not use that? Or just get an LA for your digital stuff?

[Berni]

Well for digital having 4 channels is more important because digital buses usually come with a lot of signals. But don't overestimate the usefulness of having LA and analog channels aligned, usually you can use an analog channel  to trigger from some digital line. I also prefer seeing digital things in analog because it also shows me any signal integrity problems like ringing or crosstalk as I often work with stuff past 10MHz.

Personally I have a Agilent MSO6034 that I got from eBay for about 2500 USD. It's a 4ch 300MHz scope with 16 digital channels.Its basically the same as the x3000 scopes, just not as thin, no signal gen and the UI is blue instead of black, but it does boot faster and has 8M points of memory. You can also unlock software features on it trough a trick(like the overly expensive serial decode)

A good scope is good for 10s of years as they don't change that much, but they do come down in price. The equivalent of that X2000 would have cost more than 10 grand in the past. But if you are lucky you can score one of those old scopes for cheep on eBay like I did.

[Smokey]

A good scope is good for 10s of years as they don't change that much.....

This has been true in the past, but look at how fast DSOs have been evolving recently.  I'm not saying something made now won't be any good in 10 years, but the features on a scope made in 10 years will completely blow away anything made now.  Just saying.

::edit:: typo

[echen1024]

The biggest advances in analog scopes were cursors and built in measurements. Now, with DSOs, it's like smartphones. Yearly model releases, bigger screens, larger memory...

[ben_r_]

Def the four channel! Pick up a separate LA like a Saleae or a used Agilent LA on eBay or something. Then if you still feel you want or need an LA built into the scope you can always add it later in the Agilent X-Series scope.