USB and Other Cheaper Oscilloscopes

[enorbet]

Firstly let me apologize if this is the wrong subdirectory. I didn't find one specific to just electronics and suppose the computing section may be too limited to apply or be seen by many. If it needs to be moved, no problem.

For quite some time I have been aware that I have spent rather a great deal of money on computers and especially in the graphics department - video cards and monitors. It seems to me there is no need to duplicate this expense and often the options are less powerful than what is in my PC. So I have wondered why more measuring devices that display an output haven't made more use of what most people already have.

In this thread I will concern myself with just Oscilloscopes. As of a year or two ago the vast majority if not all USB scopes and even those with moderate self-contained CPUs that only lack just the display section, have been written off as mere toys. With the advent of such devices as

https://www.picotech.com/oscilloscope/2200/picoscope-2200-portable-oscilloscopes

I'm wondering if that has changed or specifically if anyone here has had any experience, good or bad with such scopes. Is progress occurring?

 

[nsaspook]

The problem with most cheap USB scopes is the resolution, buffer depth and true real-time random data acquisition speed. This one is 8 bit raw with continuous sample rates of 1MS/s to 9.6MS/s on the high end and the noise spec is not very good. It's ok for a low end hobby unit but there are deals on used Oscilloscopes that blow it away.

 

[davenn]

I have asked for the thread to be moved to electrical engineering, where most electronics related stuff is discussed

cheers
Dave

 

[Jeff Rosenbury]

I've used picoscope products and they work fine for what they do.

As with any scope you need to understand their limitations. Don't buy one that is the minimum of what you need. If you need to measure 200 MHz, try to get a 1 Ghz scope for example, not a 200 MHz one. The 200 MHz one will work, but only if you're very careful about all the parameters. A digital scope is not an analog scope. With way too much oversampling, they may work the same, but not near the limits of the machines.

Read the manual and know your test equipment.

It's possible to overspend, but generally you get what you pay for. Good scopes are expensive. Good probes cost money.

 

[enorbet]

Thanks to all who have replied. I welcome any input but perhaps I should be clearer about my aims and experience. I have worked in the electronics field for 50+ years and got my first oscilloscope ( a Heathkit ) at age 17 so I'm not a newcomer looking for immediate purchasing advice. In fact I recently got a decent deal on eBay for a Tektronix 2430A which suits my needs nicely.

What I am interested in is the advancement of digital scopes since I have seen reviews of $1500.00 digital storage units that are rather dismal and others of $500.00 that are at least actually useful. Presently, as nsaspook has mentioned, used analog units costing well under $200 will "blow away" scopes costing 10 times as much.

It seems to be rather a minefield out there and newcomers need considerable research to avoid being severely disappointed. At some point used analog deals will become rare and digital will be the main viable option. I'm interested in it's progress especially for those of us that can benefit from portability.

Since mobile computing has made great leaps I would think manufacturers would cater more to devices that can employ their power, even if only the display since all metering requires some form of visual output to the user and in most cases the display electronics (chips, monitor, and the power to run them) are a large percentage of the cost of any such equipment. Also even the smallest notebooks or even tablets have screens substantially larger than most handheld test equipment and in the case of oscilloscopes pretty much all of them including large bench units.

So I think I'm asking for any feedback on test equipment, in this case centering on oscilloscopes, that might show if portable affordable tools are advancing. I was encouraged to look at the bigger picture when I saw this very low-priced but actually decent DMM http://www.bkprecision.com/products/multimeters/390A-test-bench-dmm-with-protective-rubberized-case-and-usb-interface.html .
https://www.youtube.com/watch?v=RmJ2pZloW18
On the flip side I was quite discouraged in viewing this https://www.youtube.com/watch?v=RmJ2pZloW18 review of the above mentioned $1500.00 digital scope. The review shows that this example has some very nice elements but also some truly severe drawbacks and problems. When will some manufacturer fine tune design and function so someone doesn't waste such a substantial investment in what amounts to a toy? So far I haven't seen much advancement in oscilloscopes and it seems such a natural "marriage" I'd like to keep up with it's progress.

Ideally we will hear from someone who has bought a unit with which they've been satisfied and pleased.

 

[nsaspook]

I don't want to be completely down on USB scopes, there are some fairly good units for specialized uses like Mixed Signal embedded design at a competitive price. The problem you have is will the 'drivers' for it still be good on the next OS release even if the hardware if still good. Will they be around for 20 years for needed PC updates that a standalone ds1054z setting on bench for 20 years won't need.

http://www.oscopes.info/market/2256-usb-oscilloscopes-product-overview-low-end

Low end standalone that has about the same functionally as good USB units.

 

[enorbet]

Thank you nsaspook that's a really useful bullet points comparison list you linked. Since you seem to enjoy our Aussie brother-in-arms too we should mention that he points out (as if we needed reminders, but then again sometimes we do :) ) manufacturers often engage in "specmanship" that is posting unrealistic or ideal ( equals "you're never going to see that ever in your life") situations so we should all read those specs with a nod an and a wink.

The Rigol has certainly set the bar but I'm betting they could offer one for up to $100 less sans display. Now wouldn't that be great, a semi-pro quality scope, brand new for $300 USD just add PC display? It might even be possible to simply use so-called smart TVs if an HDMI output was employed. The advantage of digital that is essentially impossible for analog is the Storage aspect where a snapshot can be displayed indefinitely for analysis. That would be interesting to see on my 32" smart TV. :)

Regarding drivers I'm unconcerned for base hardware drivers since any that would interest me would have native Linux drivers available so at the very least the function I started with would still be available 20+ years later. Unlike Windows, support isn't broken with new releases. However I did see some article about specialty drivers specific to an app like Matlab. That could possibly become a problem down the road, so good point.

 

[nsaspook]

Regarding drivers I'm unconcerned for base hardware drivers since any that would interest me would have native Linux drivers available so at the very least the function I started with would still be available 20+ years later. Unlike Windows, support isn't broken with new releases. However I did see some article about specialty drivers specific to an app like Matlab. That could possibly become a problem down the road, so good point.

Don't think that Linux can't have bitrot for drivers even at the open source level. I've written several in tree device drivers for Linux, almost every major kernel revision breaks something as there is no userland API that remains completely stable. Another problem is that Linux/Windows IS NOT A REALTIME OS so it's very tricky to get consistant data flow and displays at high speeds without lots of buffering so you lose that 'at the moment' response you get with a standalone unit running a special stripped down RT-Linux version.

I've got a hobby project that needs a 'high' speed DAQ driver for the Pi board so I'm writting a simple custom kernel driver for it. The interface for the ADC/DAC is SPI at 1mhz but the problems are basically the same as USB. A completely userland driver is slow because of the non-realtime nature of the OS and context switches so you must run as close to the hardware without much abstraction if possible at ring 0 to have complete control of timing and data-flow. I currently have to maintain two verions of the driver because of kernel level API changes from the 3.18+ to 4.+ kernels along with the hardware verisions in the code. There is just no way it will be usable with future changes in hardware and software without someone to maintain the code.
Linux Driver exampe

 

[enorbet]

Thanks again nsaspook...thought provoking and I much agree with you. In examining my habits I see that I have gravitated to creating fairly dedicated PCs for exactly this reason. I work in audio recording and editing rather a lot so I compile custom realtime, low-latency kernels but you're correct in that it is tiresome to deal with updates so I tend to lock such boxes down, in effect turning them into something like an appliance, or an embedded device. This is not a problem for machines that do jobs that don't require direct internet connectivity. If it worked yesterday it will work tomorrow just as it did if one doesn't cave to "new and improved" and does not need any sort of updates including security updates.

Specifically, my main general purpose box runs Slackware 14.0 but my DAW (which never connects to the internet) runs 12.2 and will for the foreseeable future since it does what I want handsomely and any possible benefit (well... not exactly "any"... they just require considerable scrutiny and planning before I would risk what took so long to put together and tweak) is generally outweighed by the risk. At some point "good enough" and amassing a capital of experience of familiarity and expertise outweigh "new and improved". Yes, I buy stick shift cars :)

 

[nsaspook]

Some images from a date code from the 90's 2465A DM (350 MHz 4 channels with 4 1/2 digit multimeter) that I'm using to test and calibrate my Raspberry PI driver. These things are a steal for advanced hardware hacking if you can find a good used one for under $200.

The next in line 2465B was one of the best analog scopes ever made and I think the last analog scope Tektronix made.
https://www.valuetronics.com/Manuals/TEK-2465A.pdf
https://www.amazon.com/dp/B0064SF27S/?tag=pfamazon01-20

SPI signal tracing:
http://forum.allaboutcircuits.com/threads/raspberry-pi-daq-system.75543/#post-870092

 

[enorbet]

ping nsaspook - That surely is a beauty. I definitely have a prejudice in favor of analog in the case of Oscilloscopes and part of the reason for my search and this thread was to try to rise above my comfort zone. My conclusion is that the USB type digitals are getting quite good but are not quite "there" yet, still straddling the line between serious toy and entry level pro.

I would have loved to have been so fortunate to score a 2465A for even more than $200 USD but my needs time wise began to push me so I grabbed a Tektronix 2430 off eBay for $190 USD. Not a bad buy not a great one but at least it is in excellent working order and though it was a bit of a search (apparently the 2430A was much more popular) I got full manuals. It serves the dual purpose of giving me a decent working scope right away and also a good learning tool to get comfortable with digital OScopes. One oddity that sometimes accompanies rapid new growth is the apparent disconnect or disparity between USB scopes and full scopes, having their own graphics and monitor. It is possible (though hardly recommended :P ) to buy a usb scope which although adds some features some will find useful is still lacking in the specs department for up to $1500 USD while at the same time Tektronix, Agilent, and Rigol offer what I would call mini scopes since they have their own dedicated graphics and screen real estate (quite small but actually very usable) with very serious specs where it counts most and these can be had from between $400 - $600.

This discrepancy to me seems just ready to pop for USB types and the emergence of devices like the Raspberry Pi and it's clones looks like very fertile ground just waiting to "get some seeds".

Thank everyone for your responses, I hope this thread helps some or provides some food for thought and/or motivation. It helped me and more than just the common effect of having to articulate a nebulae of thought to others, forcing more critical and focused thinking. There are some very skilled people on these forums. Nice place.

 

[nsaspook]

The 2465A (that includes the CTT/WR/GPIB/DM options) was a dropped equipment score from work. It still works perfectly but a few knobs and the case was crunched so it was scrapped by me when I ran the cal dept. and found a new home. Your scope is a great unit and when Tek says 100mhz you can usually see 200mhz without a problem as they were very conservative with bandwidth specs. Enjoy...