About Picos' Sampling Rate

[balnazzar]

Hi, I was look at the Pico 5000 series datasheet. These scopes do have the "usual" 1 GSa/s resolution, but here is something that caught my attention:

Maximum sampling rate (continuous USB streaming into PC memory):

USB 3, using PicoScope software: 15 to 20 MS/s
USB 3, using PicoSDK: 125 MS/s (8-bit) or 62.5 MS/s (12 to 16 bit modes)

I'd like someone knowleadgeable about Picos to explain this.. I mean, what good will the 1 GSa/s do, if the signal is processed and displayed by the software, which has access to just 15-20 MSa/s??

Of course it can't function in such a way, so I basically don't understand how a USB scope actually work.

Thanks.

[voltsandjolts]

Question for you:
How much memory is in the picoscope 5000 and what is it for?

[balnazzar]

Question for you:
How much memory is in the picoscope 5000 and what is it for?

128-512M, as per datasheet. It's used for holding the stuff it acquires, and it's obviously related to the sampling rate.. Since the software resides on a PC with virtually unlimited memory, I'd expect the sampling rate of a USB scope to be the same or higher (if the ADC allows) with respect to a standalone scope.

[balnazzar]

Here is another noob-ish observation: more memory depth means that the scope can be slow under certain conditions if it hasn't enough processing power. But again, on a usb scope much of the processing is done by the pc,  that has much more processing power than a standalone scope.

[jasonRF]

Hi, I was look at the Pico 5000 series datasheet. These scopes do have the "usual" 1 GSa/s resolution, but here is something that caught my attention:

Maximum sampling rate (continuous USB streaming into PC memory):

USB 3, using PicoScope software: 15 to 20 MS/s
USB 3, using PicoSDK: 125 MS/s (8-bit) or 62.5 MS/s (12 to 16 bit modes)

I'd like someone knowleadgeable about Picos to explain this.. I mean, what good will the 1 GSa/s do, if the signal is processed and displayed by the software, which has access to just 15-20 MSa/s??

Of course it can't function in such a way, so I basically don't understand how a USB scope actually work.

Thanks.

You can run the picoscopes in two modes.  One is the traditional block mode where the on-board memory is used to store the captures, so you are limited to that memory depth.   This mode allows for the highest sample rates.  But there will be time gaps between the sampled waveforms.  I believe most scopes work like this.

The second mode is streaming mode, which continuously streams samples from the acquisition system to the pc.  But you cannot do this at the highest sample rates.  If you write your own code using the picoSDK API, you get the 125 MS/s and can stream for hours on end if you know what you are doing when you code it up.  But you can also stream using Picoscope6 software (not sure it is in picoscope7 yet), and this is at the lower rate od 15-20 MS/s.  But the picoscope6 software only has a 100 MSample buffer to handle streaming. 

So it makes no sense to use streaming mode inside the picoscope software with models that have more than 100 MSamples of memory. 

I have used streaming inside picoscope6 with my 2204a - it gives me 100 MSamples at 1 MS/s - so can be useful for catching weird behavior in audio curcuits.  But unless I write my own code in picoSDK I will never use streaming mode with my 5244b.

Jason

[balnazzar]

You can run the picoscopes in two modes. 

I didn't know that. Thanks Jason, it's an important feature for my needs.

Are you aware of other oscilloscopes that have streaming mode?

But there will be time gaps between the sampled waveforms.  I believe most scopes work like this.

That's exactly what I found to be annoying in some usage scenarios I already made experience of.

continuously streams samples from the acquisition system to the pc.  But you cannot do this at the highest sample rates.  If you write your own code using the picoSDK API, you get the 125 MS/s and can stream for hours on end if you know what you are doing when you code it up.  But you can also stream using Picoscope6 software (not sure it is in picoscope7 yet), and this is at the lower rate od 15-20 MS/s.  But the picoscope6 software only has a 100 MSample buffer to handle streaming. 

But why just 125 MSa/a for the API and only 100 MSa buffer size for the software? It's a very, very low figure when compared to the average pc ram availability (not to mention swapping on superfast NVMe drives...).

I have used streaming inside picoscope6 with my 2204a - it gives me 100 MSamples at 1 MS/s - so can be useful for catching weird behavior in audio curcuits.  But unless I write my own code in picoSDK I will never use streaming mode with my 5244b.

I'm planning to buy a 2205A-D2 mainly for evaluation purposes and to get the very good software. Allow me to ask another question, so that I'd avoid creating another thread that I think wouldn't be useful to many.

The integrated AWG is "a bit" limited (100 KHz), so the Pico won't be a great bode plotting device (even the beefier B models only get to 1 MHz, probably due to heat dissipation issues).

But then I found this: https://www.eevblog.com/forum/testgear/picoscope-2000/25/

Look at post #39. Along with the usual app, he mention producing a Bode plot with an external generator. Note that this one has the phase plot too.. So it's not the usual "sweep and get the gain". I have not managed to exactly understand how this bloke achieved that. Can you explain? Thanks!

[jasonRF]

You can run the picoscopes in two modes. 

I didn't know that. Thanks Jason, it's an important feature for my needs.

Are you aware of other oscilloscopes that have streaming mode?

But there will be time gaps between the sampled waveforms.  I believe most scopes work like this.

That's exactly what I found to be annoying in some usage scenarios I already made experience of.

continuously streams samples from the acquisition system to the pc.  But you cannot do this at the highest sample rates.  If you write your own code using the picoSDK API, you get the 125 MS/s and can stream for hours on end if you know what you are doing when you code it up.  But you can also stream using Picoscope6 software (not sure it is in picoscope7 yet), and this is at the lower rate od 15-20 MS/s.  But the picoscope6 software only has a 100 MSample buffer to handle streaming. 

But why just 125 MSa/a for the API and only 100 MSa buffer size for the software? It's a very, very low figure when compared to the average pc ram availability (not to mention swapping on superfast NVMe drives...).

I have used streaming inside picoscope6 with my 2204a - it gives me 100 MSamples at 1 MS/s - so can be useful for catching weird behavior in audio curcuits.  But unless I write my own code in picoSDK I will never use streaming mode with my 5244b.

I'm planning to buy a 2205A-D2 mainly for evaluation purposes and to get the very good software. Allow me to ask another question, so that I'd avoid creating another thread that I think wouldn't be useful to many.

The integrated AWG is "a bit" limited (100 KHz), so the Pico won't be a great bode plotting device (even the beefier B models only get to 1 MHz, probably due to heat dissipation issues).

But then I found this: https://www.eevblog.com/forum/testgear/picoscope-2000/25/

Look at post #39. Along with the usual app, he mention producing a Bode plot with an external generator. Note that this one has the phase plot too.. So it's not the usual "sweep and get the gain". I have not managed to exactly understand how this bloke achieved that. Can you explain? Thanks!

They are using a math channel to compute the phase offset - although it is unsigned.   I have played with this a little in the distant past, but I was using single tones, not sweeps.  One thing about this approach is that the number of decades will be limited by your sample memory. 

I am currently working on a simple project that uses the python wrappers for picoSDK to control my 2204a and a simple function generator to make Bode plots from 1Hz-5Mz.  The function generator is yet another arduino-controlled device based on the AD9833 breakout boards you can find everywhere.  The generator is currently on a solderless breadboard and needs a rework because I made a dumb choice in placing the amplitude attenuator prior to the amplifier - as expected it creates a noise problem.   My python script uses the simple single-frequency-at-a-time approach to building up the plot.  The 2204a and 2205a both share the same code in picoSDK, so the same script should work.


Have you checked the used market where you live?    I have seen used 2206B sell for the same price as a new 2205a, and used 2205a sell for the price of a new 2204a.  I picked up a used 5244b a few months ago for $457 delivered, and while it is a discontinued model with only usb2 it was not much more than a new 2206b.

Jason

[balnazzar]

You can run the picoscopes in two modes. 

I didn't know that. Thanks Jason, it's an important feature for my needs.

Are you aware of other oscilloscopes that have streaming mode?

But there will be time gaps between the sampled waveforms.  I believe most scopes work like this.

That's exactly what I found to be annoying in some usage scenarios I already made experience of.

continuously streams samples from the acquisition system to the pc.  But you cannot do this at the highest sample rates.  If you write your own code using the picoSDK API, you get the 125 MS/s and can stream for hours on end if you know what you are doing when you code it up.  But you can also stream using Picoscope6 software (not sure it is in picoscope7 yet), and this is at the lower rate od 15-20 MS/s.  But the picoscope6 software only has a 100 MSample buffer to handle streaming. 

But why just 125 MSa/a for the API and only 100 MSa buffer size for the software? It's a very, very low figure when compared to the average pc ram availability (not to mention swapping on superfast NVMe drives...).

I have used streaming inside picoscope6 with my 2204a - it gives me 100 MSamples at 1 MS/s - so can be useful for catching weird behavior in audio curcuits.  But unless I write my own code in picoSDK I will never use streaming mode with my 5244b.

I'm planning to buy a 2205A-D2 mainly for evaluation purposes and to get the very good software. Allow me to ask another question, so that I'd avoid creating another thread that I think wouldn't be useful to many.

The integrated AWG is "a bit" limited (100 KHz), so the Pico won't be a great bode plotting device (even the beefier B models only get to 1 MHz, probably due to heat dissipation issues).

But then I found this: https://www.eevblog.com/forum/testgear/picoscope-2000/25/

Look at post #39. Along with the usual app, he mention producing a Bode plot with an external generator. Note that this one has the phase plot too.. So it's not the usual "sweep and get the gain". I have not managed to exactly understand how this bloke achieved that. Can you explain? Thanks!

They are using a math channel to compute the phase offset - although it is unsigned.   I have played with this a little in the distant past, but I was using single tones, not sweeps.  One thing about this approach is that the number of decades will be limited by your sample memory. 

I am currently working on a simple project that uses the python wrappers for picoSDK to control my 2204a and a simple function generator to make Bode plots from 1Hz-5Mz.  The function generator is yet another arduino-controlled device based on the AD9833 breakout boards you can find everywhere.  The generator is currently on a solderless breadboard and needs a rework because I made a dumb choice in placing the amplitude attenuator prior to the amplifier - as expected it creates a noise problem.   My python script uses the simple single-frequency-at-a-time approach to building up the plot.  The 2204a and 2205a both share the same code in picoSDK, so the same script should work.


Have you checked the used market where you live?    I have seen used 2206B sell for the same price as a new 2205a, and used 2205a sell for the price of a new 2204a.  I picked up a used 5244b a few months ago for $457 delivered, and while it is a discontinued model with only usb2 it was not much more than a new 2206b.

Jason

Understood, thanks.. About the phase difference, I was thinking about doing the same point-by-point by wrapping SCPI commands into python, so that you are not limited by the available sample memory. If you plan to share your work in future, please give me a whistle

Used market.. I'm checking ebay almost daily, but EU-available used Picos are either extremely expensive (they just slash 50 bucks w.r.t. the new ones) or in a very bad shape, or a bit cheaper but from UK (and you pay a sh*tload of customs fees since Brexit, plus pesky bureaucracy).

I'd be interested in used 4ch 5000 series Picos, sub-1000 euros. Even 60 MHz BW would do. I love these flexible ADC. But they are nowhere to be found for 1000 euros.

[jasonRF]

If you are holding out for 4 channels then you will probably need to wait a long time, or be willing to spend a lot of money.  Are you keeping all of these devices you are buying, or returning some?   Because sds1104x-e + AD2 + 2205a = a hefty sum of money.   

By the way, I do not think Picoscopes support SCPI. 

I will post my project when it is in a little better shape.  I need to redo the hardware design and clean up the script.   It will probably take 4+ weeks, since my job is taking up most of my time at the moment.  I am making it truly bare-bones, for the purpose of sharing something simple that actually works.  The script uses standard Fourier analysis to extract just the frequency of interest, and also adjusts the vertical setting for each frequency to maximize dynamic range. Others that have more microcontroller and python experience can add bells and whistles (lcd front-panel on generator, GUI for software, etc).   This is my first project using either, and I am not actually doing anything with the arduino beyond using a few commands from a library I downloaded.    I may eventually pick up a cheap SCPI-compatible scope (maybe vds1022i, or fanless benchtop owon or rebranded owon, or used tek tbs1000 if get a good deal) to see if I can make that work as well.  Just for fun.   

Jason

[balnazzar]

If you are holding out for 4 channels then you will probably need to wait a long time, or be willing to spend a lot of money.  Are you keeping all of these devices you are buying, or returning some?   Because sds1104x-e + AD2 + 2205a = a hefty sum of money.   

By the way, I do not think Picoscopes support SCPI. 

I will post my project when it is in a little better shape.  I need to redo the hardware design and clean up the script.   It will probably take 4+ weeks, since my job is taking up most of my time at the moment.  I am making it truly bare-bones, for the purpose of sharing something simple that actually works.  The script uses standard Fourier analysis to extract just the frequency of interest, and also adjusts the vertical setting for each frequency to maximize dynamic range. Others that have more microcontroller and python experience can add bells and whistles (lcd front-panel on generator, GUI for software, etc).   This is my first project using either, and I am not actually doing anything with the arduino beyond using a few commands from a library I downloaded.    I may eventually pick up a cheap SCPI-compatible scope (maybe vds1022i) to see if I can make that work as well.  Just for fun.   

Jason

I'm returning both the Siggy and the AD2 (both of them = 1000 bucks). My budget is limited to that sum.
I'm returning the Siglent because I like the scope very much, but I just cannot concentrate with its acoustical noise. And I'm returning the AD2 because I like the software but the instrument itself is quite limited. Great for some specific stuff, not so for other tasks. So that leaves me with the VDS1022i that is surprisingly decent for the price (and has much more sensitivity than the ad2) but cannot bode plot.. Once I saw the post about bode plotting with an external generator, I thought to buy the 2205A for 200$ so to have a pro tempore cheap bode plot device until I find a decent scope for 1 grand maximum.  In your opinion, will its amount of memory be sufficient for doing FRA with an external generator?

Wasn't aware that Picos don't support SCPI  I'll do some experiments with the 1022i and let you know, though.

Thanks!

[bdunham7]

I'd like someone knowleadgeable about Picos to explain this.. I mean, what good will the 1 GSa/s do, if the signal is processed and displayed by the software, which has access to just 15-20 MSa/s??

Of course it can't function in such a way, so I basically don't understand how a USB scope actually work.

You may not understand how a DSO in general actually works, at least in practice.  The scope can acquire and store samples at its advertised rate up to the stated memory depth and that typically takes a very short time, memory depth divided by sample rate.  So for 10Mpoints and 1GSa/s, your acquisition time is 10ms.  After the acquisition, the scope gets to work processing and displaying the samples and depending on the mode that it is in, it typically won't do another acquisition until all that is done.  The time that it is busy like this is called 'blind time', and this is often very significant.  Lets say it take the scope 90ms to do all the processing, giving you a blind time of 90%.  You can see that the total throughput of samples is actually only 100MSa/s in the long run. 

So in the case of the PicoScope, it normally works the same way.  It takes a capture up to its internal memory depth then gets to work processing it.  In this case, the sample memory is transmitted over the USB cable to the PC for processing and the time it takes to do that is just added to the blind time.  Of course the faster processing in the PC and the ability of the PicoScope to retrigger right after the data has been sent may offset that  somewhat.

The continuous sample rate is much, much lower and this also has an analogue with regular DSOs.  They typically have a continuous sampling (roll mode) ability that will display a moving chart, but at a much lower maximum sample rate than the advertised maximum.  Some may also have a digitizing or data recording function that allows you to continuously export, but again always at a much lower sample rate.

The key here is real-time memory access.   In order for a scope to do 1GSa/s, it has to write at least 1Gb/s to memory in real time.  That just isn't going to happen over a USB cable or even internally in a PC without a hardware buffer.

[jasonRF]

If you are holding out for 4 channels then you will probably need to wait a long time, or be willing to spend a lot of money.  Are you keeping all of these devices you are buying, or returning some?   Because sds1104x-e + AD2 + 2205a = a hefty sum of money.   

By the way, I do not think Picoscopes support SCPI. 

I will post my project when it is in a little better shape.  I need to redo the hardware design and clean up the script.   It will probably take 4+ weeks, since my job is taking up most of my time at the moment.  I am making it truly bare-bones, for the purpose of sharing something simple that actually works.  The script uses standard Fourier analysis to extract just the frequency of interest, and also adjusts the vertical setting for each frequency to maximize dynamic range. Others that have more microcontroller and python experience can add bells and whistles (lcd front-panel on generator, GUI for software, etc).   This is my first project using either, and I am not actually doing anything with the arduino beyond using a few commands from a library I downloaded.    I may eventually pick up a cheap SCPI-compatible scope (maybe vds1022i) to see if I can make that work as well.  Just for fun.   

Jason

I'm returning both the Siggy and the AD2 (both of them = 1000 bucks). My budget is limited to that sum.
I'm returning the Siglent because I like the scope very much, but I just cannot concentrate with its acoustical noise. And I'm returning the AD2 because I like the software but the instrument itself is quite limited. Great for some specific stuff, not so for other tasks. So that leaves me with the VDS1022i that is surprisingly decent for the price (and has much more sensitivity than the ad2) but cannot bode plot.. Once I saw the post about bode plotting with an external generator, I thought to buy the 2205A for 200$ so to have a pro tempore cheap bode plot device until I find a decent scope for 1 grand maximum.  In your opinion, will its amount of memory be sufficient for doing FRA with an external generator?

Wasn't aware that Picos don't support SCPI  I'll do some experiments with the 1022i and let you know, though.

Thanks!

One thing that would be interesting for you to do is measure the channel-to-channel crosstalk on the vds1022i.  The specs are pretty poor at 40:1, compared to 200:1 for the Pico.  If it really is just 40:1 then that will limit the ability of the owon to do Bode plotting.  In practice I get much better than 200:1 on the 2204a although I have never measured it directly.  The owon might similarly exceed the spec. 

If the vds1022 has sufficiently low crosstalk, then if I were you I would try to get Bode plotting working on the vds1022i before buying a 2205a.   Especially since the scope hardware on the two models has similar capability in some ways.  The vds1022i has a little better sensitivity and the 2205a has a little more memory.   Of course the software is the big difference.   

I can post a draft of my script and the hardware by next weekend if you want to see if you can modify it to work with the vds1022i.   A quick and dirty approach might be to use the nice python API that florentbr wrote for the vds1022i. 

Jason

[Anthocyanina]

I'm returning both the Siggy and the AD2 (both of them = 1000 bucks)... And I'm returning the AD2 because I like the software but the instrument itself is quite limited. Great for some specific stuff, not so for other tasks...

I'm curious about this, what have you found you aren't able to do with the AD2? not from testing the limits of it but from daily use? I have a rigol 1104z, a keysight 1102g, an owon 1022i, and the AD2, and unless I need more bandwidth or more channels, I go to the AD2 for pretty much everything else

[balnazzar]

I'm returning both the Siggy and the AD2 (both of them = 1000 bucks)... And I'm returning the AD2 because I like the software but the instrument itself is quite limited. Great for some specific stuff, not so for other tasks...

I'm curious about this, what have you found you aren't able to do with the AD2? not from testing the limits of it but from daily use? I have a rigol 1104z, a keysight 1102g, an owon 1022i, and the AD2, and unless I need more bandwidth or more channels, I go to the AD2 for pretty much everything else

See this: https://www.eevblog.com/forum/testgear/is-a-rigol-mso5000-overkill-for-a-hobbyist/300/

from post #311 onwards. Maybe a bit before that in order to get context.

Basically I'm unhappy with its two levels of sensitivity and doing everything else as digital zoom.

Im interested in your opinions & experiences about the use of AD2 with (relatively) low level signals.

[balnazzar]

If you are holding out for 4 channels then you will probably need to wait a long time, or be willing to spend a lot of money.  Are you keeping all of these devices you are buying, or returning some?   Because sds1104x-e + AD2 + 2205a = a hefty sum of money.   

By the way, I do not think Picoscopes support SCPI. 

I will post my project when it is in a little better shape.  I need to redo the hardware design and clean up the script.   It will probably take 4+ weeks, since my job is taking up most of my time at the moment.  I am making it truly bare-bones, for the purpose of sharing something simple that actually works.  The script uses standard Fourier analysis to extract just the frequency of interest, and also adjusts the vertical setting for each frequency to maximize dynamic range. Others that have more microcontroller and python experience can add bells and whistles (lcd front-panel on generator, GUI for software, etc).   This is my first project using either, and I am not actually doing anything with the arduino beyond using a few commands from a library I downloaded.    I may eventually pick up a cheap SCPI-compatible scope (maybe vds1022i) to see if I can make that work as well.  Just for fun.   

Jason

I'm returning both the Siggy and the AD2 (both of them = 1000 bucks). My budget is limited to that sum.
I'm returning the Siglent because I like the scope very much, but I just cannot concentrate with its acoustical noise. And I'm returning the AD2 because I like the software but the instrument itself is quite limited. Great for some specific stuff, not so for other tasks. So that leaves me with the VDS1022i that is surprisingly decent for the price (and has much more sensitivity than the ad2) but cannot bode plot.. Once I saw the post about bode plotting with an external generator, I thought to buy the 2205A for 200$ so to have a pro tempore cheap bode plot device until I find a decent scope for 1 grand maximum.  In your opinion, will its amount of memory be sufficient for doing FRA with an external generator?

Wasn't aware that Picos don't support SCPI  I'll do some experiments with the 1022i and let you know, though.

Thanks!

One thing that would be interesting for you to do is measure the channel-to-channel crosstalk on the vds1022i.  The specs are pretty poor at 40:1, compared to 200:1 for the Pico.  If it really is just 40:1 then that will limit the ability of the owon to do Bode plotting.  In practice I get much better than 200:1 on the 2204a although I have never measured it directly.  The owon might similarly exceed the spec. 

If the vds1022 has sufficiently low crosstalk, then if I were you I would try to get Bode plotting working on the vds1022i before buying a 2205a.   Especially since the scope hardware on the two models has similar capability in some ways.  The vds1022i has a little better sensitivity and the 2205a has a little more memory.   Of course the software is the big difference.   

I can post a draft of my script and the hardware by next weekend if you want to see if you can modify it to work with the vds1022i.   A quick and dirty approach might be to use the nice python API that florentbr wrote for the vds1022i. 

Jason

yes, if you can post a draft, it would be great. Even if it's still rough, just to get the idea! Thanks.

Also, where can I find this florentbr's repository?

As a starting point (not specifically for FRA, but in general for Python/SCPI with Owon hardware), I thought about this: https://github.com/CircuitAnalysis/OWON

Made by this guy: for the 6000 series.

[balnazzar]

Also, I can get a rough idea about the interchannel crosstalk visually, but how to measure it rigorously (indulge me, I know it's a noob question).

[tautech]

Also, I can get a rough idea about the interchannel crosstalk visually, but how to measure it rigorously (indulge me, I know it's a noob question).

If Crosstalk is NOT defined in the datasheet look elsewhere !

[Anthocyanina]

I'm returning both the Siggy and the AD2 (both of them = 1000 bucks)... And I'm returning the AD2 because I like the software but the instrument itself is quite limited. Great for some specific stuff, not so for other tasks...

I'm curious about this, what have you found you aren't able to do with the AD2? not from testing the limits of it but from daily use? I have a rigol 1104z, a keysight 1102g, an owon 1022i, and the AD2, and unless I need more bandwidth or more channels, I go to the AD2 for pretty much everything else

See this: https://www.eevblog.com/forum/testgear/is-a-rigol-mso5000-overkill-for-a-hobbyist/300/

from post #311 onwards. Maybe a bit before that in order to get context.

Basically I'm unhappy with its two levels of sensitivity and doing everything else as digital zoom.

Im interested in your opinions & experiences about the use of AD2 with (relatively) low level signals.

Yeah, I've read some of that, and it's undeniable that the AD2 is not good for low level signals, but it's still a really good general purpose tool. It could be that you very often need to look at signals under 200~500mVpp and in that case, absolutely, the AD2 will not be great for that. I need to look at very small signals just ocasionally and that would be another use case for one of my other oscilloscopes, but more often than not I'm working(or playing ) with signals above 500mVpp and under 20MHz, and for those cases, the AD2 feels great to use, not to mention that with it being not only an oscilloscope, i get even more non-oscilloscope stuff done with it, like the pattern generator and logic analyzer at the same time to work on diy digital circuits, which you can also power from the AD2 supplies, that would require a lot more space with separate instruments, so there's also that convenience. I guess i'm more curious about what you do that will require really good performance at low mV/ settings so often that the AD2 becomes a bad purchase

[balnazzar]

Also, I can get a rough idea about the interchannel crosstalk visually, but how to measure it rigorously (indulge me, I know it's a noob question).

If Crosstalk is NOT defined in the datasheet look elsewhere !

It's defined, but these cheap scopes differ by a considerable margin from what's declared. Sometimes they best the spec, sometimes they are well below.

[jasonRF]

florentbr’s vds1022i software is at

https://github.com/florentbr/OWON-VDS1022

I have not used it, since I do not have one of those scopes.   But the python example he gives looks very convenient to use!   

Jason

[balnazzar]

Yeah, I've read some of that, and it's undeniable that the AD2 is not good for low level signals, but it's still a really good general purpose tool. It could be that you very often need to look at signals under 200~500mVpp and in that case, absolutely, the AD2 will not be great for that. I need to look at very small signals just ocasionally and that would be another use case for one of my other oscilloscopes, but more often than not I'm working(or playing ) with signals above 500mVpp and under 20MHz, and for those cases, the AD2 feels great to use, not to mention that with it being not only an oscilloscope, i get even more non-oscilloscope stuff done with it, like the pattern generator and logic analyzer at the same time to work on diy digital circuits, which you can also power from the AD2 supplies, that would require a lot more space with separate instruments, so there's also that convenience. I guess i'm more curious about what you do that will require really good performance at low mV/ settings so often that the AD2 becomes a bad purchase

That's true. The problem is that with the BNC adapter it's close to 500 eur. For a device with all those limitations, it's a bit too much. It's more than 1054Z, and on par with 1104x-e. I hear that before covid it could be had for like 199. At such a price, I would have never returned it.
On the positive side, it's silent and portable.

As for the specific application that left me dissatisfied, I was looking at a signal outputted by a high frequency accelerometer. TBH usually I look at it with a siglent sds5000 at the university, but I was triggered by the fact that the day after having had the discussion I linked above, I brought the AD2 with me at the lab and indeed it proved to be useless as soon as the signal degraded below ~20-30 mVpp.

[adam4521]

For the Picoscope, consider the 2205A MSO. They come up reasonably often on EBay and you should be able to get one for less than £200. A bit more memory and performance than 2205A, plus the extra digital channels. I have one and it’s great for decoding. Some idiosyncrasies though, a weird one is that there is no trigger hold off, which apparently people have been asking for years. So not perfect, but a good, convenient tool.

[balnazzar]

For the Picoscope, consider the 2205A MSO. They come up reasonably often on EBay and you should be able to get one for less than £200. A bit more memory and performance than 2205A, plus the extra digital channels. I have one and it’s great for decoding. Some idiosyncrasies though, a weird one is that there is no trigger hold off, which apparently people have been asking for years. So not perfect, but a good, convenient tool.

If you buy it new, it costs more than twice the analog-only 2205A (if you can find it)... I'll see if I can find it on ebay, but given the scarcity, I don't have much hope...

[balnazzar]

florentbr’s vds1022i software is at

https://github.com/florentbr/OWON-VDS1022

I have not used it, since I do not have one of those scopes.   But the python example he gives looks very convenient to use!   

Jason

The drop-in software is by itself better than the original one. Now I'm try to delve into the python side..

Thanks for having mentioned it

[nigelwright7557]

The scope takes in a fast burst of data from the A2D to RAM.
Then sends it to PC via USB (speed at whatever standard is used 2.0 etc)
Then scope takes in another sample etc.
So however slow USB is it will still work.
Obviously too slow will delay display on screen.