Autonerdz forums

Alright, I'll take it upon myself to ask the questions and start the thread to end all threads  !  (Okay...I really hope it doesn't end all threads  ) 
What specs should we be concerned with?  Okay, maybe a better question is to start from the beginning.  What should we be looking for in a DSO?  What can't we give up for performance to gain, say, support?  Just how important is sample rate?  What is bandwidth and just how much do we need?  Why don't most companies publish the record point length, or buffer size?
I'd be lying if I led you to believe that I don't know the answers or have my own opinions on some of the questions I've written down.  But I want you guys to take aim with your own answers, questions, and opinions.  Iatn, lately, has become very testy in the forums about DSO's...but the fact remains that there is so much dis/mis-information out there and still, according to a poll in Underhood Service, only about 50% of shops have a DSO.  There are so many to choose from, so many misnomers  that is can't be anything but overwelming to be in the position of wanting to purchase and or upgrade DSO's.  Later, Matt.
Oh yeah....in all cases, make sure you buy atleast one Picoscope!      

That Ferret deal was a joke.  I remember when it first got posted on iATN...it pretty much got ripped apart. 
Did you find what you were looking for?  What scopes were in the running?  What scopes were no longer in the running after you started learning more about DSO's?  If you did it over again, what would you be looking for in a new scope?  Later, Matt.

I have seen that ferret article around.  I saw it on iATN too.  I think the reason John had a hard time understanding it is because it is nonsense.  Here are a few points:

Channels:  How sample rate is affected depends on if there is a converter for each channel or if it is shared.  This goes for buffer too.  So, you can't use this factor without knowing these things.

Update Rate:  This is an important issue, but has nothing to do with how many samples you get in a capture.  Therefore, if you factor this in, you are not calculating sample rate.

AD Converter Bit Resolution:  Again, this has nothing to do with sample rate.  True, the more bits the more vertical detail, but samples are plotted horizontally.  There is no such thing as vertical sampling on a scope.  Two vertical samples would mean that you had two voltage values at the same point in time.  That's impossible.  More vertical resolution plots each sample with a more accurate voltage value, but there are not more samples.  Imagine a horizontal grid on the screen.  An 8bit scope would have 256 horizontal lines representing voltage values where a sample point could be placed.  Samples cannot be plotted between the grid lines.  If a sample point had a voltage between the grid lines, it would be plotted at the voltage for the next lower grid line.  This is why a sine wave appears to have jagged steps in it with a low resolution scope.  Each point has a straight line connecting points between grid lines.  A 12bit scope would have 4096 grid lines, therefore each point plotted has a much more accurate voltage value.  Much smaller voltage changes can be detected and displayed.  This means more detail.  So...factoring bit resolution into the formula just makes for a larger number.  The number has nothing to do with samples collected.

Pixel Width:  I think they are trying to say buffer size here.  Essential ingredient for sample rate.  On fixed display scopes, with x number of pixels, it makes no sense to have a buffer larger than the fixed display width.  On scopes that can expand the display, many more samples can be collected.  For example, Pico can collect many more samples that any PC screen resolution can display.  You have to expand the capture to reveal the detail. 

There is no mention of max AD converter speed in the article.  This is the other essential ingredient to calculate sample rate.  So, the Ferret article leaves one of the two essential pieces of information needed to calculate sample rate out of their formula, and adds in a bunch that have nothing to do with sample rate.  The result is nonsense.

There is a third factor, screen time.  We factor that in based on our own time base settings.  We only need the two factors of buffer size and max speed to factor in with our known screen time to calculate true sample rate.

BTW, the Lab Scope Tutorial, on the Ferret link that John provided, also has many errors, but at least it's not total nonsense. 

You'll have to scroll down the page alittle, it's actually called:  XYZ's of Oscilliscopes.

Fair warning this is a long response.

Your thread may not have recieved great response since many members are pico biased (I know I am)

I have gone through a couple of scopes that I still own before purchasing the pico.  The first was a velleman one channnel scope from radioshack.com it was inexpensive at about 100 bucks but very limited in abilties it did due sensor and fuel pump waves ok but resolution was weak and the max voltage is 20 with no attenuator.  Overall it was limited but better than no scope at all.  I have not used it in a long time.

Later I upgraded to a mac scopemeter  it was a improvement by far.  It is a handheld 2 channel scope it was more expensive than the velleman at about 5 bills but had far more ability but the resolution was still weak.  I still use it but mainly for quick checks of alternator diodes since it has ac coupling.

I recently aquired a snap on power grapghing meter
This is post pico and that was mainly purchased because my dvom died I demoed a unit and was very impressed but it is no compare to the pico.  I use it mainly for the dvom abilities and the database but I have compared many signals on this unit and the pico at the same time.  The leads from the pico are well suited to stacking to be connected to 2 units at once.  The main difference I noticed is the PGM show large failure fairly well but does not show subtle failures well at all.  The area the PGM does shine is real long time bases such as a pressure transducer hooked into the fuel line to observe fuel pressure on cars that die when hot.  You hook it up set the long time base and walk away to your other work.   But in all fairness you could probably do the same with pico in a save on trigger but I use too many programs on my computer to have it tied up for long periods of time.

I finally saved up enough and bought a pico 212-3.  What an improvement  over both my prior scopes.  It is a longer set up time but I overcame that by setting up a diagnostic cart.  One of the coolest things about the cart is I have a jump box that I use to power the pico making it truly portable.  I have to say this is the best tool I have bought as far making me money.  My two favorite abilities are the expand option and the composite view(s) option.

I have been in the situation where I was without my pico for awhile but Tom helped me fix it fairly quickly.  The ground connecter burnt through one channel. May be tom could post a pic.   I suspect I hooked up something wrong more than once and it burnt up.  The time I was without my pico I tried using my other scopes for diagnosing cars.  It was a truly trying time since going from the pico to my old scope was like going back to the dark ages. 

If and when I purchase another scope it will be another pico 212(3/50/100) series. Mainly for ability and the excellent product support from tom and pico.  I recently upgraded to a new computer and still have the old notebook that I ran the pico on.  The thought is that with both computers running and external trigger to both you could effectively have a true 4 channel scope for a investment similar to many 4 channels and less than some 4 channels that are common ground style.

Just a couple of other thoughts on your post.  In deciding what scope to buy the application it is to be used in should be considered.  You did not mention scope accuracy  the pico 212 is very good at +/-1 dc but  the new 3000 series accels at many things but is not well suited to auto applications it has an accuracy of +/- 3  and has a very limited voltage range. 

One last note one of my coworkers has been wanting a pico for a long time.  He is kind of an ebay addict he has been watching for pico's  for at least a year.  In that time there has only been one.  How many other auto test tools can say that?  Seems to me most people are keeping them.  Of course he did not buy it and regrets it.