YA scope advice topic: Decision tree for Siglent (and/or MSO5000) scope purchase

[pipe2null]

Sorry about the epic, I get wordy when I'm attempting to be thorough.  I've tried to include all the typically requested info I've seen from other posts when people ask similar questions.

I've wanted my own scope for a long time, and I'm currently working on a few projects that finally give me a good enough excuse to take the plunge.  And, like so many other people visiting this forum, I could really use a little guidance.

Any future replacement of whatever equipment I buy is far enough down the road that I have to consider this a "one time" purchase/investment, so while my budget is tight, if spending a few-ish hundred $ more now will better enable the breadth of projects I can do, then I should go ahead and spend the extra dough.  Since 99% of the projects I have in mind are primarily for re/learning stuff myself or intended to be released as open source (if I hit a good 'nough quality bar), I don't have any qualms about hacking a (NEW) scope to get more capability, limited to available step by step hacking guides and a (literal) Radio Shack soldering iron.  Hopefully soon I'll have a good excuse to pickup a hot air rework station.  A reflow toaster oven is on my project list too, but it's a wayz down. 

To start with, I need to scope various serial connections at the max my hardware supports (125MHz SPI, SWD/JTAG *signals* for ARM close to theoretical max, etc...), but I can just use a 4ch scope for the time being.  Down the road I want to add the "minor" upgrade of an MSO digital probe.  The next round of projects use more analog (characterizing serially-connected analog sensors, etc.).  I have a long list of projects that I want to do, this is just a quick list of starters.

Question #1:
I have had a difficult time trying to find any specs on Siglent's MSO digital probe options (specific models mentioned below), like actual sampling rate or memory depth or BW...  To put it simply, what is the maximum clock rate the MSO digital channels can reliably acquire?

I am REALLY rusty on the EE side of things, but I *THINK* I'll be ok with a scope that has 4ch, 200MHz BW, 1GSa/s, serial decode included (or hackable), and MSO digital channels as a supported "purchase/DIY when necessary" minor upgrade, as my minimum requirements given how comparatively cheap scopes are these days.

Question #2:
Temporarily ignoring other specs/features I should be looking for, are these minimum specs appropriate?  Too low/high?

But.....  I would really like to be able to do some projects with USB, hopefully a little at the physical layer.  Anything USB3.0 and above is cost-prohibitive for my budget equipment-wise, but I'd *really like* to be able to use at least high speed USB (480Mbps) in projects.  So:

Question #3:
What are the minimum specs for a scope to do "meaningful/useful" work with high speed USB?  Put another way, if you exclude engineering tasks that require a $10k+ scope, what is a "reasonable" minimum spec for the other stuff?

As much as I liked using HP/Agilent back-in-the-day, Keysight is way outside my price range.  I've heard good things about Siglent, so I previously attempted to isolate my decision tree to that brand.  If that is a questionable approach, please question.

My recent decision tree *WAS*:
First choice: The mythical, non-existent Siglent SDS2204X-E or SDS2354X-E
Reluctant second choice:  Provided the BW firmware hack still works, the SDS1104X-E
Reluctant third choice:  SDS1204X-E

But then I found a discounted SDS2304X: 4ch+trig, 300MHz BW, 2GSa/s, 140Mpts...

Question #4:
Is the SDS2304X hackable, if so where should I look for guidance?  It's a stretch, but I can kinda afford the discounted price, but not if I have to tack on the serial decode license.  Based on the info I've seen, I have the impression that the AWG and MSO hardware are built in, so am I correct in thinking that a firmware hack plus some form of DIY digital probe is all that's needed?  Better BW, sampling rate, and memory depth seem like a really good trade off for the built in web server of the X-E series, at least for the currently existing 4ch models.

And then there is the Rigol MSO5074 which is apparently very hackable and the MSO port seems prime for a DIY digital probe.  Although I don't know if they are closing the loopholes for hacking...

Question #5:
At first glance, a Rigol MSO5074 hacked to max spec seems like the overall best option, but then I watched Dave's 40 minute rambling rant video on all the bugs and problems.  So.......  Headaches, bugs, and reliability versus specs versus budget?  I guess my Question #5 is the overall question:  Given the info in this epicly long post, what sage guidance can you provide?

Question #6 (Bonus Round):
What is the airspeed velocity of an unladen swallow?


Answers, comments, and any appropriate head smacking are much appreciated, especially on #4 and #5.  Just to reiterate, I have a tight budget, but since I need to make this a "one time" purchase/investment, shelling out a little more dough now makes sense.  I'm not looking for the biggest bang for the buck, I'm going for the minimum buck for maximum feasible development potential with room to grow, if that nuanced difference makes sense.

[nctnico]

You should try a few scopes and add the GW Instek MSO-2204E and R&S RTB2004 to the list. Siglent is quite decent nowadays but what I don't like about Siglent scopes is that they only decode what is on screen (there is a workaround using the zoom function but that takes away screen real estate and it still is a crutch) and the automatic memory length selection. Both make the oscilloscope more cumbersome to operate and Siglent should really alter their thinking about how an oscilloscope with deep memory should work. You are right about the Rigol MSO5000; it is not ready at all and it won't be for the next few years.

However given your high frequency requirements you may want to look at a used oscilloscope with a >1GHz bandwidth to look at USB signals. Then again... without the right software there is not much to see. Getting the circuit board right based on impedance calculators is a better method to create a board which just works. Lot's of people over here who are willing to check your layout if you have doubts.

[tautech]

#1, Datasheets have the MSO sampling and memory specs: https://siglentna.com/resources/documents/digital-oscilloscopes/#sds5000x-series
Scroll down through the models.
#3  for your first choice.....would be mine too ! I really hope they build those models. 
#4 Check the Siglent .ads thread. 
Yep the MSO and AWG HW is inbuilt however a standalone AWG is always better.
Don't do a DIY MSO probe as this HW is top notch nice stuff and you'd be struggling to build anything as good.
Pic of the one I had for my SDS2304X:

They've used this HW again for the new SDS5000X models and I've bought SPL2016 again.
#6.............~24MPH ! 

[2N3055]

Question #6 (Bonus Round):
What is the airspeed velocity of an unladen swallow?


Answers, comments, and any appropriate head smacking are much appreciated, especially on #4 and #5.  Just to reiterate, I have a tight budget, but since I need to make this a "one time" purchase/investment, shelling out a little more dough now makes sense.  I'm not looking for the biggest bang for the buck, I'm going for the minimum buck for maximum feasible development potential with room to grow, if that nuanced difference makes sense.

#6 African or European swallow ?

As Nico said, to look at 125 MHz squarewave, you need AT LEAST 1GHz bandwidth. Or better to say, on that signal you will have sub 500 ps edges, in order for signal resemble square not triangle. With 1 GHz scope you will see transitions, but would need 2GHz + to actually estimate edge quality.. You might be able to decode something with 350-500MHz scope, but will have no clue about signal integrity..

[jemangedeslolos]

Hello,

If I may...I have a beautiful SDS2204X with all genuine software licence for sell.
I "hacked" it to 300Mhz BW with an hardware hack ( 4 caps to remove ).

In perfect condition ( like a new one, with the original box, original probes, no scratch ).
If you are interessted you know what to do 

[Fungus]

To start with, I need to scope various serial connections at the max my hardware supports (125MHz SPI, SWD/JTAG *signals* for ARM close to theoretical max, etc...), but I can just use a 4ch scope for the time being.

I am REALLY rusty on the EE side of things, but I *THINK* I'll be ok with a scope that has 4ch, 200MHz BW, 1GSa/s,

For 125MHz SPI? Not even close.

Question #5:
At first glance, a Rigol MSO5074 hacked to max spec seems like the overall best option, but then I watched Dave's 40 minute rambling rant video on all the bugs and problems.

Dave had one of the very first ones and firmware can change (that's the whole point).

There's been a few firmware updates since that video. You can see release notes on the download page and there's a thread for that.

PS: Siglents have bugs, too.

[nctnico]

To start with, I need to scope various serial connections at the max my hardware supports (125MHz SPI, SWD/JTAG *signals* for ARM close to theoretical max, etc...), but I can just use a 4ch scope for the time being.

I am REALLY rusty on the EE side of things, but I *THINK* I'll be ok with a scope that has 4ch, 200MHz BW, 1GSa/s,

For 125MHz SPI? Not even close.

Actually I managed to decode 125MHz SPI with my GDS-2204E. https://www.eevblog.com/forum/testgear/gw-instek-gds2204e-(200mhz-4-channel-dso)-review/msg941933/#msg941933

[Fungus]

To start with, I need to scope various serial connections at the max my hardware supports (125MHz SPI, SWD/JTAG *signals* for ARM close to theoretical max, etc...), but I can just use a 4ch scope for the time being.

I am REALLY rusty on the EE side of things, but I *THINK* I'll be ok with a scope that has 4ch, 200MHz BW, 1GSa/s,

For 125MHz SPI? Not even close.

Actually I managed to decode 125MHz SPI with my GDS-2204E.

Decoding ain't the same as "scoping" but maybe the OP can clear up what exactly he wants to do.

[nctnico]

True but an older high frequency oscilloscope likely doesn't have SPI decoding. So you need two tools to do signal verification and functional verification when the budget is limited (and most oscilloscopes with decoding can't do 125MHz SPI anyway).

[tv84]

Question #4:
SDS2x04X - hackable to 300MHz (BUT via HW, so you loose warranty). Old Blackfin architecture. The new X-E scopes are ARM architecture so any bug squashing or new addons are more likely here.

A SDS2x04X-E line would be great for a lot of people! For now, it seems Siglent isnt capable of inserting the 4-ch in this model line.

But, others indicate that you may need BWs > 300 MHz...

Question #5:
MSO5074 is totally hackable via software. Extremely unlikely the loopholes will ever be closed. The FW may not yet be perfect but it's the best HW for the buck, in it's price range. And will most probably benefit from Rigol's investment in the new line of MSO8000 FW developments.

Question #6 (Bonus Round):
If you hack the african into a european species then both speeds will be the famous 24mph.

As a conclusion: it's much easier to answer your Q#6 than the others!

[Fungus]

Question #5:
MSO5074 is totally hackable via software. Extremely unlikely the loopholes will ever be closed. The FW may not yet be perfect but it's the best HW for the buck, in it's price range. And will most probably benefit from Rigol's investment in the new line of MSO8000 FW developments.

Yep. For $1000 and a few minutes of hacking the Rigol MSO5074 totally destroys the competition in terms of raw numbers (bandwidth, samples/sec, memory, etc). People are reporting that only minor bugs remain now and the next firmware is expected any time now.

[pipe2null]

Great feedback, thanks everybody!

I was unaware there were people here willing to give feedback on pcb layout, that's awesome, thanks for the tip nctnico!  I think that tidbit by itself completely takes care of my Question #3, so I don't need to worry about USB for this scope.

BUT: I had a head smacking moment a few hours after I started this topic.  I was so excited about finally having the excuse to buy a shiny new toy and trying to find one that would fulfill the most use cases that I didn't dig deep enough into the reason I need a scope in the first place...

So, the entire point of one of the projects I'm currently working on is to determine the max "usable" speed of a particular serial connection, and then, well, use that maximum speed.  The theoretical max is 250MHz.  The chip I'm communicating with has a return clock that catches up to my input clock once it has finished it's work each input clock edge, so I am expecting the periods of the 2 clock signals to be the same, but the half cycles of the return clock might be asymmetrical.  Since I do not "automatically" care about the waveform itself (nor can I afford a high enough BW scope to see it) I can just connect a probe across the 2 clocks, using the return clock as ground, and trigger on zero crossings to get my timing measurements.  At least I think that'll work...  It's been many, many years since I've done any EE.

Since I'm not MacGyver and my current "Lab" consists of a multimeter, a Radio Shack soldering iron, and miscellaneous bits 'o yarn, I have not come up with some hacky work around to measure the numbers.  I could try and do it with software, but I'd still have to use a scope to validate the software...  Yea.  Head smack.

So for that specific measurement with not caring about the wave forms, I'm thinking I might get away with 300MHz BW, or maybe even a 200MHz scope might work well enough to just get the timing.  But even 2GSa/s is not good enough, 4GSa might work, but 8GSa/s is probably the most reasonable target.  Please correct me if I am off.  And a hacked MSO5000 is the only scope I am aware of that could do that within my budget, assuming there aren't any bugs in that specific part of the feature set...  Arg.  The last thing I want to do is buy an unreliable piece of....... test equipment.

Using Dartboard Statistical Modelling, here is my revised basic requirement list (Available for comment):
Use case % : Appx Min Spec
0.01% : >=1GHz BW, 8GSa/s
80%-95% : 100MHz BW, 1GSa/s/ch or 500MSa/s/ch when using all 4
Rem. % : 200MHz BW, 2GSa/s, or maybe 300MHz BW for a little head room

Decoding ain't the same as "scoping" but maybe the OP can clear up what exactly he wants to do.

Well, aside from my 0.01% use case, "scoping" is obviously desired, but I'll have to settle for decoding when freq is high enough I cannot afford the 5th or even the 3rd harmonic.  It's a bizarre version of "pay per view", the harmonics are still there, but you have to go to your neighbors house to see them.

Unfortunately, the "once in a lifetime of a scope / 0.01%" use case is actually almost the very first measurement I need, and it's probably the only time I'll ever need anything remotely close to that high a spec...  That's another head smack.  I really like the suggestions to pick up an older used unit with higher specs, but I want my first scope to be new since I don't know the first or second thing about buying used equipment of this type.  Well, not yet at least.  Second "new" toy, I will likely buy used and would not hesitate exploratory hardware hacking.   ...  Great  ...  I haven't even bought my first scope and I'm already talking about my second.  That is not a good sign.

Well, the jury is still out on my final purchase decision, I need to find out if any/which MSO5000 features work reliably, but that is a little off topic for this thread.  As soon as I realized the digital probes came standard in the box (crazy talk!), I added the GW Instek MSO-2204E to the potential options list.  The R&S RTB2004 is a bit too much a stretch out of my price range, or at least the prices I was able to find.  I get the impression that R&S is basically the gold standard for scopes these days?


Updated tentative scope purchase decision tree:
TBD: Possibly my only choice, due to the 0.01% use case: hacked Rigol MSO5000

Backup decision tree, if the MSO5000 doesn't make it to the cart, or otherwise suffers a horrible batting accident after use:
Dead heat for first choice:

• Mythical SDS2354X-E - Since it doesn't exist, it's free!
• Discounted SDS2304X - For BW, 2GSa/s, mem depth, general quality for $ range
• GW Instek MSO-2204E - If my primary use is digital and digital decode, possible 125MHz SPI decoding...

Second choice:

• SDS1104X-E with hacked BW - If I only care about 80-ish% of use cases

Question #6:  The answer we were looking for is "African or European swallow?"  So congratulations for those of you who...  Wait a moment, the judges are deliberating...  Apparently, since this is an engineering forum, the judges are also accepting 24mph and awarding an extra bonus point to tv84 out of fear of what gene spliced monsters may be lurking in their basement...

[tautech]

................
Second choice:

• SDS1104X-E with hacked BW - If I only care about 80-ish% of use cases

With just a $500 outlay don't you think it's worth a try ?

Actually for you now you're a member here you can get a discount from Saelig with a members discount code.
Ask for it in their thread.

[Performa01]

Backup decision tree, if the MSO5000 doesn't make it to the cart, or otherwise suffers a horrible batting accident after use:
Dead heat for first choice:

• Mythical SDS2354X-E - Since it doesn't exist, it's free!
• Discounted SDS2304X - For BW, 2GSa/s, mem depth, general quality for $ range
• GW Instek MSO-2204E - If my primary use is digital and digital decode, possible 125MHz SPI decoding...

Second choice:

• SDS1104X-E with hacked BW - If I only care about 80-ish% of use cases

If you're willing to wait for a couple of months, you might find your 'mythical' first choice materializing in an unexpected way that exceeds your current imagination…

[nctnico]

So, the entire point of one of the projects I'm currently working on is to determine the max "usable" speed of a particular serial connection, and then, well, use that maximum speed.  The theoretical max is 250MHz.  The chip I'm communicating with has a return clock that catches up to my input clock once it has finished it's work each input clock edge, so I am expecting the periods of the 2 clock signals to be the same, but the half cycles of the return clock might be asymmetrical.  Since I do not "automatically" care about the waveform itself (nor can I afford a high enough BW scope to see it) I can just connect a probe across the 2 clocks, using the return clock as ground, and trigger on zero crossings to get my timing measurements.  At least I think that'll work...  It's been many, many years since I've done any EE.

For that you'll need a high speed differential probe (shameless plug: I sell a relatively low cost one) because you can't connect the ground of a regular probe to random points in a circuit. The ground of a regular probe is connected to the mains ground and the flying lead also has a lot of self inductance. But for looking at 250MHz clocks you'll need a scope with a bandwidth of at least 1GHz to see something slightly meaningfull.

However if you are only after timing then a logic analyser may be a better tool. Logic analysers are pretty cheap nowadays. A Tektronix TLA715 + TLA7AA4 acquisition module gets you a solution with 125ps time (8GHz sampling frequency) resolution. There is a forum member who has made / sells (?) adapter boards for the TLA7AA4 probes to have flying leads.

[Fungus]

Decoding ain't the same as "scoping" but maybe the OP can clear up what exactly he wants to do.

Well, aside from my 0.01% use case, "scoping" is obviously desired, but I'll have to settle for decoding when freq is high enough I cannot afford the 5th or even the 3rd harmonic.  It's a bizarre version of "pay per view", the harmonics are still there, but you have to go to your neighbors house to see them.
...
Well, the jury is still out on my final purchase decision, I need to find out if any/which MSO5000 features work reliably, but that is a little off topic for this thread.
...

Updated tentative scope purchase decision tree:
TBD: Possibly my only choice, due to the 0.01% use case: hacked Rigol MSO5000

Some people in the other thread have said that signals are visible until nearly 1Ghz on a Rigol MSO5000. The "350MHz" version will have the 1st+3rd harmonics of a 125MHz square wave, a lot of the 5th, some of the 7th.

It's mostly going to be down to your probing though. Poking at a PCB trace with a standard probe isn't going to give a clean signal at those frequencies on any 'scope. That's a whole thread in itself.

I'm sure there's plenty of people here who can tell you which features are working on the MSO5000. Last I read it was almost all of them, it's a long way from the deathtrap that Dave was driving when he made that video.

[pipe2null]

So, the entire point of one of the projects I'm currently working on is to determine the max "usable" speed of a particular serial connection, and then, well, use that maximum speed.  The theoretical max is 250MHz.  The chip I'm communicating with has a return clock that catches up to my input clock once it has finished it's work each input clock edge, so I am expecting the periods of the 2 clock signals to be the same, but the half cycles of the return clock might be asymmetrical.  Since I do not "automatically" care about the waveform itself (nor can I afford a high enough BW scope to see it) I can just connect a probe across the 2 clocks, using the return clock as ground, and trigger on zero crossings to get my timing measurements.  At least I think that'll work...  It's been many, many years since I've done any EE.

For that you'll need a high speed differential probe (shameless plug: I sell a relatively low cost one) because you can't connect the ground of a regular probe to random points in a circuit. The ground of a regular probe is connected to the mains ground and the flying lead also has a lot of self inductance. But for looking at 250MHz clocks you'll need a scope with a bandwidth of at least 1GHz to see something slightly meaningfull.

However if you are only after timing then a logic analyser may be a better tool. Logic analysers are pretty cheap nowadays. A Tektronix TLA715 + TLA7AA4 acquisition module gets you a solution with 125ps time (8GHz sampling frequency) resolution. There is a forum member who has made / sells (?) adapter boards for the TLA7AA4 probes to have flying leads.

Eeek...  I am really, really rusty.  And used to tools with a floating ground.  Thanks for pointing that out.  That's another well earned head smack.  I was trying get the measurement using only 1 channel so I could get 8GSa/s out of the Rigol, but 2+ channels at 8GSa/s each would be better.

Question:
Aside from being able to arbitrarily use a random voltage or signal as "ground", and ignoring high voltage scenarios, is there a reasonable rule of thumb on when you should be using a differential probe instead of a standard probe?  The probes I'm aware of cost as much as or more than some of the scopes I'm looking at, so knowing a reasonable line in the sand would be helpful.

A differential probe is the right way to get the measurement with one channel, no argument there.  Or use equipment with higher specs per channel, no argument there either.  I'll still need a scope but I am looking at different logic analyzers now that will hopefully cover my 0.01% use case within budget for both... 

Yes, I understand that I'm trying to do stuff that requires Ferrari performance with a junk yard budget.   

With just a $500 outlay don't you think it's worth a try ?

Actually for you now you're a member here you can get a discount from Saelig with a members discount code.
Ask for it in their thread.

Thanks for the Saelig discount tip!
Even without the BW hack I was seriously looking at the SDS1104X-E as a good starter scope until I started getting a better idea of what minimum specs I should be targeting for a few of my projects.  Since some of the required specs are cost prohibitive (like doing high speed USB without help from forum members), that Siglent model is still on my list until I... Well, until I make myself a little less ignorant and understand my actual, realistic needs well enough to make a decision.  I'm re/learning everything as I go, as my probe grounding related head smack demonstrates.

...

• Mythical SDS2354X-E - Since it doesn't exist, it's free!

If you're willing to wait for a couple of months, you might find your 'mythical' first choice materializing in an unexpected way that exceeds your current imagination…

I am intrigued...  If you know something that I haven't heard yet...

Some people in the other thread have said that signals are visible until nearly 1Ghz on a Rigol MSO5000. The "350MHz" version will have the 1st+3rd harmonics of a 125MHz square wave, a lot of the 5th, some of the 7th.

It's mostly going to be down to your probing though. Poking at a PCB trace with a standard probe isn't going to give a clean signal at those frequencies on any 'scope. That's a whole thread in itself.

I'm sure there's plenty of people here who can tell you which features are working on the MSO5000. Last I read it was almost all of them, it's a long way from the deathtrap that Dave was driving when he made that video.

I would hesitate much less if I already had a lower spec starter scope that was pretty reliable and/or other test equipment I could use to double check the numbers if I see something weird.  I am attempting to avoid the tradeoff of reliability in exchange for better specs at a given price, at least for my first scope.  But with my budget and my half-baked requirements, ...  The Rigol may not be my first or even second choice due to concerns about reliability (improved or not), but at this point it is the scope I'm most likely to actually buy.

Yea, I still have homework to finish before buying my new toy...  Patience is not something I'm good at.

[tautech]

With just a $500 outlay don't you think it's worth a try ?

Actually for you now you're a member here you can get a discount from Saelig with a members discount code.
Ask for it in their thread.

Thanks for the Saelig discount tip!
Even without the BW hack I was seriously looking at the SDS1104X-E as a good starter scope until I started getting a better idea of what minimum specs I should be targeting for a few of my projects.  Since some of the required specs are cost prohibitive (like doing high speed USB without help from forum members), that Siglent model is still on my list until I... Well, until I make myself a little less ignorant and understand my actual, realistic needs well enough to make a decision.  I'm re/learning everything as I go, as my probe grounding related head smack demonstrates.

A tip if I may, always remember a scope ground lead is the reference point for the signal itself and the fact that it is also mains ground referenced is an unfortunate safety coincidence unless we select an isolated input scope of which there are very few. With 'reference' firmly implanted in your mind and mains ground connected the chance of connecting a probe where you shouldn't are vastly reduced.
Differential probes are for safety and convenience instead of the now long frowned upon practice of floating a scope.

...

• Mythical SDS2354X-E - Since it doesn't exist, it's free!

If you're willing to wait for a couple of months, you might find your 'mythical' first choice materializing in an unexpected way that exceeds your current imagination…

I am intrigued...  If you know something that I haven't heard yet...

Performa01 gets to test pre market release models and his hint need be taken seriously.....something that I didn't know was coming as I've probably not asked the right ppl. 

[pipe2null]

Differential probes are for safety and convenience instead of the now long frowned upon practice of floating a scope.

This makes sense, and your tip about "reference AND mains ground" is slowly burrowing its way into my noggin.

My head scratching "but why?!?" question about differential probes was more from:

...
It's mostly going to be down to your probing though. Poking at a PCB trace with a standard probe isn't going to give a clean signal at those frequencies on any 'scope. That's a whole thread in itself.

The funny thing is, the moment I read this I had a vague recollection of one of my old professors explaining "everything you need to know about differential probes" in lab one day, but for the life of me I have no idea what he said, but I'm pretty sure the intro to his lecture included either "do" or "do not" float the scope.  I'll mark that as "do not".  I understand that a signal transmitted on a differential pair of wires (in this case a probe) has better signal integrity, the exact same reason it's used in USB, ethernet, etc.  As far as whether or not I need to buy a differential probe (line in the sand), is there a rule of thumb that says "if you are probing a signal over xMHz, use a differential probe"?  This might be an advanced topic for where I am at right now, only looking for a short answer if it exists.

Performa01 gets to test pre market release models and his hint need be taken seriously.....something that I didn't know was coming as I've probably not asked the right ppl. 

Man...  I am loving this forum more and more every day!  Heh.

[tautech]

Differential probes are for safety and convenience instead of the now long frowned upon practice of floating a scope.

This makes sense, and your tip about "reference AND mains ground" is slowly burrowing its way into my noggin.

My head scratching "but why?!?" question about differential probes was more from:

...
It's mostly going to be down to your probing though. Poking at a PCB trace with a standard probe isn't going to give a clean signal at those frequencies on any 'scope. That's a whole thread in itself.

The funny thing is, the moment I read this I had a vague recollection of one of my old professors explaining "everything you need to know about differential probes" in lab one day, but for the life of me I have no idea what he said, but I'm pretty sure the intro to his lecture included either "do" or "do not" float the scope.  I'll mark that as "do not".  I understand that a signal transmitted on a differential pair of wires (in this case a probe) has better signal integrity, the exact same reason it's used in USB, ethernet, etc.  As far as whether or not I need to buy a differential probe (line in the sand), is there a rule of thumb that says "if you are probing a signal over xMHz, use a differential probe"?  This might be an advanced topic for where I am at right now, only looking for a short answer if it exists.

Have a look here and the differences between traditional and differential probing become apparent.
http://www.pintek.com.tw/files/pintek/PROBE-COMPARISON-CHART.pdf
http://www.pintek.com.tw/files/pintek/Oscilloscope_e.pdf
Still it's a rabbit hole to venture into if you are needing to do any HF work as they get expensive quick.
If we also remember any connection to an active circuit for measurements in itself affects the measurement value so ideally we want probes that have no or little effect on the circuit and we judge what to use in a case by case basis. You already know all this so just engage the grey matter and dig it out from things long since forgotten.   

Performa01 gets to test pre market release models and his hint need be taken seriously.....something that I didn't know was coming as I've probably not asked the right ppl. 

Man...  I am loving this forum more and more every day!  Heh.

[pipe2null]

Still it's a rabbit hole ...

My grey matter is slooooowly waking up.  I think I've (re)achieved
Engineering 101 aka Rule #1:  Matter is squishy, everything is relative, and the only hard and fast rule is that there are no hard and fast rules.

Hopefully it won't take too long to get to
Engineering 202 aka Rule #2:  Everything is a probability curve, so start ignoring Rule #1.


Thanks for all the help everybody, I have some homework to do...