Picoscope 3203D sampling rate / sample interval

[markone]

Few days ago I purchased a Picotech 2204A with the aim to make an assessment of Picoscope platform (I have a customer that could be interested in this kind of instruments).

It took me little time to spot some evident limitations, some were expected due to price category of the device (which I would define "for educational purpose") but other not.

For example, applicable sample rates are constrained in a power of two rule like (from 2000 programming guide) :

"A PicoScope 2000 Series oscilloscope can sample at a number of different rates that correspond to the maximum sampling rate divided by 1, 2, 4, 8 and so on."

It follows that it is not possible to select simple sample interval like (i.e.) 10us, where the nearest obtainable value is 10.24us .

So I was  wondering if an higher model, like the 3203D, is capable to do that  but reading the related programming guide, again, it seem not to be the case (see attachment).

Does anyone with a similar model confirm or deny that ?

[voltsandjolts]

3000

5000

[markone]

3000

5000

I get that you missed my previous post attachment, where you can see an extract from the very first document that you linked; 5000 series is out of scope.

I need a boolean answer : is the 3203D capable to block capture with (i.e.) with 10us sample interval ?

From the formula inside the extract it seems not the case, but of course i could be wrong.

[voltsandjolts]

The answer is in your own OP and backed up by the documentation I linked.

The range of timebase values is divided into two subranges, with the subrange 0 to 2
specifying a power of 2, and the subrange greater than 2 specifying a number divided
by 125,000,000. The maximum value is 23^2–1.

It can't be any clearer.

e.g. n=1252   =>  (1252 - 2) / 125,000,000 = 10us exactly

[markone]

The answer is in your own OP and backed up by the documentation I linked.

The range of timebase values is divided into two subranges, with the subrange 0 to 2
specifying a power of 2, and the subrange greater than 2 specifying a number divided
by 125,000,000. The maximum value is 23^2–1.

It can't be any clearer.

e.g. n=1252   =>  (1252 - 2) / 125,000,000 = 10us exactly

Sorry, I completely misinterpreted the two subranges table structure 

So the boolean answer is : Yes, it can

Thank you for your help, now it's matter to find an 3203D.

[jasonRF]

On paper is looks like the sample rates for the 3000D series are the same as the 5000 series in 8-bit mode.   There is no guarantee that the software works the same with these different series, but I thought I would let you know that there is a quirk with the software when it comes to my 5244b in 8-bit mode. 

I find that I can realize any sample interval listed in the documentation inside the Picoscope6 software by playing with the number of samples I am requesting, but sometimes the software makes weird decisions when it 'automatically' picks things.  For example with 10 kSample memory selected I get
at 10 ms/div: 100 kS/s (10 us interval) with exactly 10k samples
at 5 ms/div: 200 kS/s (5 us interval) with exactly 10 k samples
at 2 ms/div: 498 kS/s (2.008 us interval) with 9960 samples

If I then set the number of desired samples to 10.01 kS, then at 2 ms/div I get 500 kS/s (2 us interval) with exactly 10k samples.  I never care if I have exactly 500 kS/s or 498 kS/s, so I never bother to fiddle to get nice 'clean' numbers.  But if you do care it could be annoying.  If it is a giant deal to you, I recommend contacting Pico to see if the 3000D series have the same issue. 

Picoscope 7 also has this issue, but it does not let you use the keyboard to select the exact number of desired samples.  I just now reported the bug to Pico, so hopefully they will fix it. 

EDIT: in picoscope7 you can also ask for sample rate.  When I pick 2 ms/div and 500 kS/s, I get the same 498 kS/s.  But if I ask for 501 kS/s I get 500 kS/s.  So you can work around the software bug sometimes. 

jason

[markone]

On paper is looks like the sample rates for the 3000D series are the same as the 5000 series in 8-bit mode.   There is no guarantee that the software works the same with these different series, but I thought I would let you know that there is a quirk with the software when it comes to my 5244b in 8-bit mode. 

I find that I can realize any sample interval listed in the documentation inside the Picoscope6 software by playing with the number of samples I am requesting, but sometimes the software makes weird decisions when it 'automatically' picks things.  For example with 10 kSample memory selected I get
at 10 ms/div: 100 kS/s (10 us interval) with exactly 10k samples
at 5 ms/div: 200 kS/s (5 us interval) with exactly 10 k samples
at 2 ms/div: 498 kS/s (2.008 us interval) with 9960 samples

If I then set the number of desired samples to 10.01 kS, then at 2 ms/div I get 500 kS/s (2 us interval) with exactly 10k samples.  I never care if I have exactly 500 kS/s or 498 kS/s, so I never bother to fiddle to get nice 'clean' numbers.  But if you do care it could be annoying.  If it is a giant deal to you, I recommend contacting Pico to see if the 3000D series have the same issue. 

Picoscope 7 also has this issue, but it does not let you use the keyboard to select the exact number of desired samples.  I just now reported the bug to Pico, so hopefully they will fix it. 

EDIT: in picoscope7 you can also ask for sample rate.  When I pick 2 ms/div and 500 kS/s, I get the same 498 kS/s.  But if I ask for 501 kS/s I get 500 kS/s.  So you can work around the software bug sometimes. 

jason

Hi Jason, thank you very much for having point out this "bug", it appears to be generated by wrong data rounding in formula / algorithm code that computes sampling parameters.

The need to get certain sample interval is not a whim, the customer has lot of scripts that post process sampling files at specific sample rate, resample treatment it is not a welcome option, especially if you consider the price.