PicoScope 2000

[MrW0lf]

Were you able to install both PS5 and PS6 on 1 pc?

Yes but something is little weird with some PS6 menus now. However I have windows GUI scaling on and just installed some Win patches also so dunno. Functionally both seem to work as far as messed with so far. In itself PS5 is of course very crude compared to PS6. I point attention to this mainly so people would notice and perhaps some good stuff would be resurrected due to public demand.

[_Wim_]

Were you able to install both PS5 and PS6 on 1 pc?

I point attention to this mainly so people would notice and perhaps some good stuff would be resurrected due to public demand.

Good idea, I was very supprized to hear they use to have a "normal" persistance mode, this is one thing I find much better on my Rigol than on the Pico. The current implementation I find useless for my kind of work.

[MrW0lf]

PS6 lacks wfm - wfm timing measurements so must invent own, heres unsigned delta(ChX,ChY) to avoid hassle with cursors:

ChB=1MHz sine, ChC=1MHz square, ~45° phase. Measurement matches cursors. Operating in good conditions (no noise).


ChB=1MHz sine, ChC=1MHz sine, ~15° phase. Additional function plots delay graph. Graph little corrupted by noise. DC average on non-plotting function is more accurate.


ChB=1MHz sine, ChC=1.01MHz sine. Graph gives general picture.


To avoid noise issues it may make sense to lower sampling rate, apply resolution enhancement, filter etc. I just demonstrated default situation.

Math for non graphing (PWM):
((((atan(1/tan(pi*(B/10000)))/pi)+(B/10000))*-((atan(1/tan(pi*(C/10000)))/pi)+(C/10000)))+0.25)*(1/freq(C))

Math for graphing version:
duty((((atan(1/tan(pi*(B/10000)))/pi)+(B/10000))*-((atan(1/tan(pi*(C/10000)))/pi)+(C/10000))))*(1/200)*(1/freq(C))

It works by finding zero crossings which results in PWM which can be further processed and graphed using duty(x).

[jasonRF]

I know this thread is old, but it seems like the best place to post this. 

This is a short review of the Picoscope 2204a, which I think is a reasonable, inexpensive entry-level USB oscilloscope for folks that don't require larger bandwidths for their signals of interest.  It currenly sells for $165 (US), although when I purchased mine in the 2015/2016 time-frame it was $139 (US).  Mine is made in the UK, so that probably does add a little to the price over what it might cost if it were produced in a less expensive country.  The picotech.com web site has all the specs, but the short-form is

2-channel, 8-bit, 100 MS/s (down to 50 MS/s with both channels running), 10 MHz nominal bandwidth, 35 ns rise-time, 8 ksample total memory.  Includes built-in signal generator that operates up to 100 kHz for sinusoids, and 4V peak-to-peak into high-impedance loads.

I have only used the Picoscope software under Windows, so all comments are in that context.   

Features I like
- Analog front-end operates like a real scope - when vertical setting is changed in Picoscope software the front-end does change gain/attenuation.  No digital zoom!  My unit has around 220 levels full-scale at each setting (about 7.8 bits).   
- actual hardware bandwidth higher than spec - my unit has at least 25 MHz bandwidth based on both rise/fall time measurements and direct bandwidth measurements.
- Low-budget scope uses high-budget software - since Pico uses same software for all models it is polished and feature-rich.  It only disables settings that the attached hardware doesn't support; nothing is disabled for marketing reasons.  So the 2204a has tons of features that most budget usb scopes lack.  Some are below.
- Lots of fancy math included - such as user-defined high-pass/low-pass/band-pass/band-stop digital filters, etc.  Other posts on this thread show quite complicated math...
- serial decoding but shallow memory can limit usability
- high-resolution mode to get up to 12 effective bits
- equivalent-time sampling up to 2 GS/s for periodic signals
- built-in signal generator provides frequency response analysis (aka Bode plots) up to 100 kHz.  You do need to download the free FRA4picoscope app for this.   
- user-defined keyboard shortcuts to make operation efficient.  For example I use left/right arrow keys to adjust time-base, so isn't any harder than using a knob.
- small physical size - so easy to shove in a drawer when not in use.  This is the primary reason I purchased it instead of a used full-size scope.

The 2204a is a low-budget device with many limitations that are obvious from the spec sheet, and some that are not spelled-out (although some come to light if you read this entire thread). 
- Signal generator actually isn't very good.  Sure, it is fine for sinusoids up to 100 kHz so works for Bode plots, and is good for square waves  up to a few kHz.  However, it has some flaky jittery performance for square-waves at 10-ish kHz that are much worse than expected based simply on the loss of Fourier components.
- Lacks automated measurements of time/phase delay between channels - this is a serious oversight in my opinion.  It is just software, and it is weird that Pico neglects it in the current version.
- FFT is less useful than you might think - because it is limited to 4k samples, and also won't display any frequencies beyond the nominal 10 MHz bandwidth of hte unit.
- No external trigger input, so if you need one you must use one of the channels, and hence this becomes a one-channel scope. 
- minimum vertical setting is 10 mV/div - so if you look at small signals all the time this is not the scope for you.  At least this is a real setting, though, as some manufacturers seem happy to advertise small vertical settings that require digital zoom.
- No vertical offsets
- channel B has a maximum of 50 MS/s sample rate, even if it is the only channel enabled.  This doesn't actually limit performance since channel A can just be used when need the full sample rate, but it can be inconvenient.
Was just playing with the 2204A and this is wrong.  With just B enabled I get 100 MS/s.  I've had this thing for years and somehow had this wrong impression for awhile 
- No trigger output

For similar or less money than the 2204a you can certainly get more capable hardware (more bandwidth, higher sample rates, external trigger, more memory, usb isolation, etc), but seldom will get a signal generator with associated Bode plots, and in all cases I'm aware of you will get less polished and/or capable software.  That is the trade, and of course there is no one-size-fits-all 'best' option. 

The primary reason I will eventually upgrade is because I want at least 100 MHz bandwidth.  Unfortunately, 100 MHz Picoscopes start at $809 (US), which is too rich for me.

jason

[jasonRF]

As a follow-up, here are a few measurements of the 2204a so support a couple of the statements I made in the last post. 

The first image is a rise-time estimate of the analog front-end.  A BK Precision 4013DDS function generator is the source, using a BNC coaxial cable and a 50-Ohm feed-through adapter at the scope input.  Rise time is shorter than 11 ns, so front-end bandwidth is at least about 32 MHz.  EDIT: this example is the best-case I found, but 12 ns is a safe upper-bound so perhaps 29 MHz is a better number.  In any case, >25 MHz seems like a safe statement given these and other measurements I've made. 

The second image is a plot showing the issue with the square-wave of the built-in signal generator.  Here I am using persistence mode, so the jitter is clearly visible. 

EDIT: decided to add a third image to show the free FRA4Picoscope app at work. This example is a passive first-order low-pass filter.  I find this app to be very useful at times, although the 600-Ohm output impedance of the signal generator is a little high for some circuits.  When that matters I add a low-output-impedance buffer between the signal generator and the device under test. 


jason

[midix]

Thanks for the review, jasonRF.

I just tried Picoscope 7 software and the demo mode felt very intuitive for a beginner, unlike AD2 software which has its specifics to keep in mind. Considering I'm debugging mostly some simple stuff (audio, video, Arduinos) and I already have a cheap logic analyzer, 2204A might be enough for my needs? Or would I be overpaying and there are better choices with software that does everything I would ever need?

I currently have OSC482 (which seems to be the same thing as SainSmart DS482) and the software is clumsy, zooming is awkward, and you have to stop the device after capturing a signal with a trigger and then use scrollbars to select previews from PC cache, and only then you can zoom with your mouse (and scroll wheel works opposite way to all the other software).

But if 2204A is not any better hardware-wise than OSC482 (comparing not only specs on paper - OSC482 looks good there, but actual behavior and quality), then I'll have to live with OSC482 limitations. I might go for 2205A or Analog Discovery 2 sometime later.

[jasonRF]

Hi Midix,

I am not familiar with the scope you have, and a quick google search landed me on pages with pretty sparse descriptions.  It is of course difficult for us to know what your use-cases will be, but if I were you I would be thinking about what measurements I want to do and how convenient I want them to be.  I would also be reluctant to spend real money on new equipment that is only a small improvement over what I already own - I'm not sure if the 2204a falls in that category or not for you.   But for basic audio stuff, at least, the 2204a is fine as long as you aren't dealing with the high voltages required for most tube circuits.  The actual analog bandwidth of mine is about 25 MHz, so is pretty good for a lot (but not all) of the stability concerns with audio circuits.  When I had the 2204a, I purchased a reasonable external audio interface to do high resolution distortion measurements before bothering to look for a better oscilloscope. 

Anyway, if you are willing to purchase used equipment, then you can sometimes pick up an Analog Discovery 1 on ebay for less than the price of a new 2204a.  The used Picoscope 2204a models don't seem to sell for that much of a discount it seems, but the higher (and older) model Picoscopes do.  You just need to be patient and may need to keep your eye out for a number of months to find the really good deals. 

jason

[mtk]

Is it possible to make PicoScope 2204A to work with 2205A firmware, the boards seems the same ?

[jasonRF]

Is it possible to make PicoScope 2204A to work with 2205A firmware, the boards seems the same ?

Interesting question.  Before now I did not realize that the boards looked the same.  Photo of the 2204a
https://sigrok.org/wiki/Pico_Technology_PicoScope_2204A
and the 2205a
https://community.element14.com/products/roadtest/rv/roadtest_reviews/1272/picoscope_2205a_osci_2
The board shown on the sigrok site even has a sticker that seems to indicate it is for both.

My understanding is that the Picoscope software takes care of firmware updates automatically, but I could be wrong about that.  In any case I haven't found any useful information about updating firmware on picoscopes. 

jason

[_Wim_]

Is it possible to make PicoScope 2204A to work with 2205A firmware, the boards seems the same ?

I was thinking exactly the same for my 5000 series scope: https://www.eevblog.com/forum/testgear/picoscope-hack/