And presented in Dave's unique non-scripted overly enthusiastic style!
PC Based Digital Storage Oscilloscope Comparison
Gold! Anyone want to buy a Bitscope?
Just to send you all my support! There’s not enough of such initiatives on the web! Thank you! By the way: i’ll definitely go for a bench scope
Well said as always.
It’s strange that no one seems to make PC-based scopes with decent sample rate and memory. It’s not exactly rocket science after all, all you need is some off-the-shelf 250Msps converters to interleave, an FPGA and some DDR RAM and bingo! 1Gsps and 64M+ sample depth…
You can get them, but they cost big $$$$
Dave, thanks for the review. I’ve been using the SDS200A for several years as I mentioned in a prior comment and for the work I do, it’s been great. Of course, most of my work is low speed digital and that’s an area where the SDS200A shines – especially when you consider the complex triggering capabilities it has. (I have no affiliation with softDSP – just a satisfied customer, btw).
I’ll have to consider the Rigol in addition to the pc based scope after your review. I think it would be a good addition to the workbench.
Finally, I would recommend you get a laptop to add to your workbench. It’s great to be able to look up datasheets and replacement part availability right at the work bench!
Another difference between PC-based and standalone oscilloscopes: Ground Fault tolerance.
Example: Accidentally touch the ground clip of your probe onto a live wire, and the fault current through a standalone scope’s innards will generally have a quick path to mains ground. There’s usually little damage, except melted chunks missing from the ground clip and a faint air of embarrassment.
The same fault in a PC-based instrument will usually take out the scope AND the PC. Did you remember to back up your files?
Great stuff. I am a software/firmware guy with previous extensive hardware background. Your videos are terrific for explaining very important concepts.
I have always thought PC based scopes should have completely replaced bench scopes by now. System partitioning suggests ‘minimize coupling, maximize cohesion’. Seems like on a PC based scope, the hardware part could be focused on – very efficient/optimal properties for high speed high resolution signals. Then on a separate partition (the PC) deal with user interface, presentation, etc.
The problem with this is, just as you have stated in EEVBlog #13, the PC based oscopes have poorer front end performance than bench scopes. (go figure…)
If the PC based scope was engineered with this partitioning, the front end could be optimized without consideration for user interface etc. Then the user interface and presentation could be a continual software development process on the PC (without the necessity for analog expertize).
cheers …Jim Hughen
You can in fact get high performance PC based scopes, like the Agilent U2701A for example, but they aren’t exactly entry level prices.
Indeed, most top end bench scopes for probably the last decade are in fact just PC boxes running Windows or some other embedded OS, with an oscilloscope front end, they are very well partitioned designs.
The main problem is the usual one, it’s not easy to design a fast sampling, high performance, high update rate oscilloscope front end. So you basically can’t buy this from anyone but the big players. So the cheap PC based scopes from smaller companies will never achieve the front end performance of the big players.
What do you think if the USBee DX?
As far as its digital scope capability goes, it’s just like almost all the other USB scopes, with the same limitations.
2.4MHz single shot bandwidth is pretty paltry, and 100mV is the lowest range, so not very useful at all.
No ability to use standard scope probes, although at that sample rate that’s not really essential.
The Logic Analyser sample rate looks to be 24MS/s, that is ridiculously small by today’s standard.
Looks like it has some nice serial protocol decoding and control features etc, but US$1500!!! are they kidding??
Ridiculously overpriced IMO, by an order of magnitude!
Its not like the rest, the main difference is it lacks any logic, its a pure USB buffer/data pump, it just reads whats on the edge and dumps it into PC at real time, very clever and ingenious, real price saver (for the manufacturer that is)
Reverse engineered :
I’m loving your series so far – I managed to spend a couple of hours last night going through the first few episodes.
I would suggest you have a look at the picoscope range of PC oscilliscopes – I have the 5204 and use it every day for digital / analogue troubleshooting etc. The 5203 is cheaper and has less buffer memory. I know it’s more expensive than your rigol (I have one of those too – brilliant for the money!) but it does have really good performance.
The benefit in my mind of the PC oscilloscope is the large screen – I have 4x the horizontal screen space on the PC than I do on the bench scope. This is really useful for analysing captures with lots of data. I also love the fact that it can save a whole sequence of trigger events so you can look back through the last few triggers, compare them etc.
Keep up the good work!
The Picoscope 5200 is a high end PC based scope. These are a different beast as I’ve mentioned, as they have equivalent sample rates and other features to bench scopes, but you pay through the nose for it of course. Actual performance can be as good as a bench scope in these high end units.
Yes, large screens can be good, and you can have that with bench scopes too, many have SVGA outputs for direct monitor connection.
Regardless of how good the PC based scope is though, IMO they are still clunky to use for everyday use, and you can’t just move it from one bench to another, on the floor etc. Not always an important feature, but essential to many.
I think your comparison of PC scopes makes some good points but is lacking in nuance. I use a Cleverscope 328A for one reason: deep memory. I have the 8Ms version (2 chan @100 Ms/sec, 12 bit per sample, 8 digital inputs). I was willing to sacrifice some sample rate in order to get more depth in my application. If instantaneous sampling in a very small window is most important, then a bench scope or more expensive PC scope is a better choice.
It’s kind of sad that oscilloscopes have not come down in price. You can’t even get a 10-year-old used Agilent 54622D for less than $3000 US. This was about the level of scope I wanted (2 chan @200 Ms/sec, 4Ms memory, 16 digital inputs). In comparison, the Cleverscope was $1000.
My laptop + the Cleverscope is much smaller than lugging around the Agilent. I can use a giant external screen, easily interface with it at high speed via software (USB vs. GPIB), and have much more flexible triggers, and logic analyzer decoding.
More modern Tek scopes still have puny memory if you want to spend around $1000 US. I just don’t see how a bench scope is a reasonable choice if you can only spend $1000.
I’ve always said PC based scopes have their place, if one with a particular feature set suits your particular needs, then go for it. The cleverscope is a nice unit, if the features suit you for the price. But if you wanted more than 10MHz single shot bandwidth then it’s effectively useless.
Bench scopes haven’t come down in price?, they have plummeted in the last year or two! What about the Rigol bench scope with 1M of 500MS/s (15KB/1GS/s) memory for around $500?
The Cleverscope is twice that price (for 1/5th the single shot bandwidth), hardly entry level category.
The problem is people think that these cheap $200 USB scopes are good value for general use, and I wanted to clear that up. They are simply toys, and poor value compared to a modern bench scopes like the Rigol.
If a USB scope comes along that matches the value for money of the Rigol then I’ll happily promote it. IMO a decent real-time general purpose scope needs 50MHz analog bandwidth and at least 500MS/s. Only the expensive USB scopes can do that, so that’s what makes them poor value for money.
Hey thanks for the reviews..
I just bought a Rigol DS1102E, and a pure digital logic analyzer from Usbee/CWAVE. So I saved a bunch of money there.
They will be here tomorrow!
You could of saved even more by just building USBee clone. USBee SX is:
for USBee AX-Pro you have to add >24Msps ADC and some frontend opamps (frontend wasnt fully cloned yet)
Hi, I really like your blog, keep up the good job!!
In the end, you convinced me on the Rigol scope, but I need a logic analyzer too. Would you recommend the DS1102D with integrated logic analyzer or a normal DS1102E + a pc-based logic analyzer? Having a PC near the bench it’s not an issue for me. I think that the DS1102D is better because i can trigger from a digital pin, and also see the waveforms in perfect synch. I need to see BLDC motor waveforms, so the analog phase + digital position sensors, so the rigol is good, but I also need to anzlyze bus signals, so a pc-based would be better. What do you think? I highly value your opinion!
I haven’t played with the Rigol logic analyser version, but it actually looks to be poor value for money. Almost $1000 extra for the DS1102E over the DS1102D!
(the 100MHz DS1102E is now only $636, so seems to have dropped recently?)
I’ve heard the logic analyser aspect of the Rigol is not all that flexible too, like not being able to set individual level thresholds (I haven’t checked the manual to confirm, the Rigol site link looks Broken?).
Mixed signal scopes are very nice for the convenience and easy analog/digital triggering, but $1000 buys you a really good PC based based logic analyser with money left over.
Also, the Rigol screen is only 320×240, so it’s hard to fit 16 channels of digital logic display on there, let alone the analog as well. I’d play with one before you buy it.
Perhaps you’d be better of getting a better speced USB one with external trigger in/out capability so you can cross trigger with your scope?
Thank you very much, you told me exactly what I wanted to know! I’ll follow your advice!
The LA doesn’t look that bad, btw…
Hey I just got in the DS1102E, very nice…
Keep up the good work, cant wait to see what your going to put out next.
The USBee SX also came in, I haven’t gotten too much time to play with it yet. Its an 8 channel logic only USB based solution.
On a side note, I was kinda disappointed with the Rigol PC control software. I was hoping for live scope readout and more remote control functionality. But hey what was I going to do with it anyway?
For the people that want to buy a logic analyzer…
take a look here
2K sample space? plus I cant find a way to zoom on the graphs
Thank you for the reviews! Very helpful.
Have you tried the VS series from Rigol? Specifically, the VS5202/VS5202D?
I’m thinking hard about one of these.
My guess is they are based on the desktops but without display and control knobs/buttons. You have any thoughts on this?
Sorry, I haven’t tried the VS series and don’t have access to one, so can’t comment first-hand.
My guess would also be that the front end and acquisition engine are the same the bench scopes. So performance should be the same.
Really depends on how they have implemented the user interface as to how easy it will be to use.
Very Nice, the pc test instruments unit is just what i was looking for.
and the USBeeSX is a piece of junk!
Looks like I bought something I will never use!
BTW… The Rigol DS1102E is awesome..
Dave.. The E version is the cheaper one while the D appears to be the logic analyzer version?
* August 23rd, 2009 at 09:48
*For the people that want to buy a logic *analyzer
Bit confused, if I purchase a 50MHz BW USB scope with 200MHz sampling rate, will I not get 50MHz single shot bandwidth?
I must say, a small, portable scope powered from USB that I can slip in my laptop bag really appeals…
Thanks in advance for any feedback.
You will still get 50MHz single shot bandwidth, but you’ll only get 4 samples per cycle. If you are happy with 4 samples then you have a 50MHz single shot bandwidth.
But most people consider the minimum samples to be 10 per cycle, so that means you need 500MS/s.
Join 505 other subscribers