Picoscope vs TiePie software comparison ?

[nurbster]

Hi EEVblog forum.  I'm interested in either the Picoscope 4444 or the TiePie HS6 DIFF PC-based differential oscilloscope.  The hardware seems similar enough and is certainly adequate for my needs.  The software is what I'm really interested as I think it a more important than hardware ( short of hardware quality issues that affect software ).  As they offer demo version of both, I've installed them on my computer and played with them a bit.  But as beginner in using an oscilloscope, I wonder if more seasoned pc-based oscilloscope users can chime in with their opinions and experiences with these companies and their software.  Please, no use a "real" ground-referenced single-ended oscilloscope.  My needs are ultra-portable, pc-powered and very well suited to differential oscilloscope.

Thank you, Sven

[MrW0lf]

What is your analog bandwidth requirement?

As for Picos - use them for 7 years now so know thing or two. TiePie - first time heard, but looks interesting on paper.

Installing Demo software failed (some missing files), but installed "real" software on top of that - then it finally worked. First impressions:
If Pico software is somewhat "designed by engineers" then this is step further down that road... maybe step too far for everyday use... 
Math channels non existent compared to Pico (for me math channels is the main thing going on for them).
However FFT on TiePie is max 32Mpts...  and some other surprising stuff going on... Tried 32Mpts FFT right away - seems to work but makes GUI almost non-responsive.

Maybe if you describe what functionality you need I can look at it in some organized manner, as far as demo mode allows.

[egonotto]

Hi,

I think the two are very different.

Example:

Bandwith: TiePie 250 MHz ... PicoScope 4444 20 MHz

Ranges   TiePie ±200 mV to ±80 V full scale ... PicoScope ±10 mV to ±50 V full scale

USB streaming from  PicoScope 4444 is 10 MS/s (is this only a true USB 2 scope?)

The Analog Discovery has too 2 differential inputs.

What want you do with the scope?

Best regard
egonotto


 

[MrW0lf]

It even more complex...

4444 w/o probes: €1085, probes cost extra and pick them according to need. Assumed VAT included on their website.
20MHz, 400MSa/s in 12bit mode, 50MSa/s in 14bit mode (1ch).
Max streaming 10MSa/s (in 14bit mode?), 100MS buffer (in software). However 50MSa/s with SDK, 14bit mode, no buffer limit.

HS6-DIFF-50XM-W5 with probes: €1227.6 (assume 20% VAT)
This is HS6D entry model: 100MHz, 50MSa/s in 8bit mode, 5MSa/s in 14bit mode (1ch).
Max streaming 10MSa/s in 8bit mode, 5MSa/s in 14bit mode (1ch).

So what can see here? Pico has handicapped their new products with somewhat outdated software. Dunno why silly 10MSa/s & 100MSa buffer limit there. Maybe to support old models? Software sure does support my old 2205 from 2010. Suggest hire some good programmers, Cambridge not far... But hardware adequate to support 20MHz 14bit incl full bw streaming with SDK (1ch?).

Now with TiePie entry model can see that costs about the same, but cannot support analog bandwidth at all due to insufficient sample  rates. In effect 20MHz 8bit and 2MHz 14bit.

Now look at TiePie top model:
HS6-DIFF-1000XM-W5 with probes: €2295.6 (assumed 20% VAT)
250MHz, 1GSa/s in 8bit mode, 100MSa/s in 14bit mode (1ch).
Max streaming 200MSa/s in 8bit mode, 100MSa/s in 14bit mode (1ch).
Can see it is 40MHz in 14bit mode (Nyquist limit!).

So if talking 14bit MHz/price ratio these 2 seem to scale pretty well, no free lunch. However TiePie entry model is joke, better buy AD2... only 2ch but at least 14bit up to 10MHz no problem and little further if pay attention.

Software wise... purely from GUI implementation standpoint (not looking features). AD2 is from this decade, Pico from last and TiePie another +10.

Since hardware wise each scope has some unique stuff that others do not it indeed comes down to exact needs: AD2 has 2x signal gen / pattern gen. Pico is only one having high wfm/s in special modes (up to 500kwfm/s in segmented mode). And TiePie top model has most brute force and streaming capability, but costs accordingly.

Tech data collected from:

https://www.picotech.com/oscilloscope/4444/picoscope-4444-overview
https://www.picotech.com/oscilloscope/4444/picoscope-4444-specifications

http://www.tiepie.com/en/products/Oscilloscopes/Handyscope_HS6-DIFF
http://www.tiepie.com/en/sales

[Fungus]

USB streaming from  PicoScope 4444 is 10 MS/s (is this only a true USB 2 scope?)

As I understand it: In "real-time" mode the low-end PicoScopes only send what you see on screen, not the full set of sample data. 10MS/s is enough for that.

Wolfie is the resident picoscope person, he can confirm/deny this and maybe tell us the ins and outs of real-time vs. 'streaming' on PicoScopes.

[MrW0lf]

As I understand it: In "real-time" mode the low-end PicoScopes only send what you see on screen, not the full set of sample data. 10MS/s is enough for that.

First look last post, tried to get to bottom of things... but otherwise  this is quite complex subject. Cannot say I fully understand it but try based on what know:
- pure streaming mode, indeed tops at ~10MS/s in software (not tried SDK, which should be much better).
- "normal" or "analysis" mode as I call it, relatively low wfm/s, full math/analysis/FFT - it tries to optimize data transfer based on zoom etc (downloads only zoom window in detail etc)
- "persistence" or "glitch hunt" mode, high wfm/s (tops at 110kwfm/s on my 2408B), no analysis - data rate unknown
- rapid trigger or segmented mode, ultra high wfm/s, ~1.3Mfms/s on my 2408B, full analysis after acquire - data rate unknown

Wolfie is the resident picoscope person

Just to clarify - I have no official connection to them - so can judge stuff only from non-professional end-user point of view. Decided to keep them on picture as a hobby because it's interesting product, has served me well and deserves it's place. Think I'll do this until there is learning opportunity for me - have learned more in last 6 months than just using them quietly in 6 years before that

[Fungus]

Wolfie is the resident picoscope person

Just to clarify - I have no official connection to them

But I think you're the only regular poster who owns one. 

I think there's big differences in how the different models of PicoScopes work though. Owning one model doesn't tell you much about the others.

[MrW0lf]

I think there's big differences in how the different models of PicoScopes work though. Owning one model doesn't tell you much about the others.

At least old 2000 and new 2000 are quite similar. However with 4444 there are big differences in hardware indeed. So dunno. If in doubt can only suggest contact their support, better on their forum because others can read also:
https://www.picotech.com/support/

[egonotto]

Hi,

However 50MSa/s with SDK, 14bit mode, no buffer limit.

where do you find this info.
I cannot even download a SDK for the 4444

Best regards
egonotto

[MrW0lf]

where do you find this info.
I cannot even download a SDK for the 4444

SDK info is on specs page:

SDK/API*
Supplied drivers   32- and 64-bit drivers for Windows 7, 8 and 10
Linux drivers
Mac OS X drivers
Example code   C, C#, Excel VBA, VB.NET, LabVIEW, MATLAB
Maximum sampling rate (USB streaming)   50 MS/s
Capture memory (USB streaming)   Up to available PC memory
Segmented memory buffers   > 1 million

But indeed, no download link. Maybe try link for 4824 (which is similar) or ask support? If its in the specs they should deliver it.

[kaos]

My TiePie from a few years ago  does not have a true trigger input, but accepts a trigger on a digital input with +-8 samples of the analog sampling. For averaging multiple traces the jitter this causes is unacceptable. When querried, TiePie has no solution. So there hardware and specs look great, but mine is unused. About to get a Pico 6000.

[nurbster]

What is your analog bandwidth requirement?
...
Maybe if you describe what functionality you need I can look at it in some organized manner, as far as demo mode allows.

Thanks for informative replies.  I think 20MHz would be adequate.  Intended applications:

1-General inspection of waveforms from inverters, SMPS, mains quality testing in remote locations.  Some three-phase stuff.

2-Serial troubleshooting of things like CAN and lots of UART as well as undocumented serial protocols.  Recently watched this video and although I only grasp some of these features seems very powerful.

3-Assorted automotive including common-rail fuel systems, noise vibration and harshness ( NVH )

Other things I think about: Fast Ethernet (100Base-TX) which I think is 37.5Mhz, but I think it easier to use a dedicated device to troubleshoot.  TDR as I'm often dealing with compromised wiring and it would be helpful to know where, but I think once again dedicated instrument easier ?  Red Pitaya has TDR application with 80cm resolution so perhaps AD2 can do this too as hardware seems similar.

[nurbster]

Hi,

where do you find this info.
I cannot even download a SDK for the 4444

Best regards
egonotto

From the Picoscope forum:

We are currently in the process of updating the examples for the ps4000a driver and there should be a driver package installer available soon.

[ebclr]

https://www.picotech.com/downloads

[nurbster]

I received the following information from Picoscope that seems relevant to this post:

Our USB 3.0 products can sample at up to around 20 MS/s with the PicoScope 6.12 software when used with a USB 3.0 connection, and the PicoScope 4444 can sample at up to around 16.67 MS/s.

The reason for the limitation is historically down to preserving PC resources, but as technology moves on this may well be reconsidered. I will notify the Development Team of your comments.

[ADT123]

I received the following information from Picoscope that seems relevant to this post:

Our USB 3.0 products can sample at up to around 20 MS/s with the PicoScope 6.12 software when used with a USB 3.0 connection, and the PicoScope 4444 can sample at up to around 16.67 MS/s.

The reason for the limitation is historically down to preserving PC resources, but as technology moves on this may well be reconsidered. I will notify the Development Team of your comments.

Just to clarify (as the above could be taken out of context) that the above sampling rates refer to USB streaming mode in PicoScope.  This is used at longer time periods of over 2 seconds across the screen where the trace is drawn live (roll mode).  The USB streaming rate using the drivers can be much faster (>150MS/s) if writing your own software.

The max sampling rate of the scope in PicoScope software is the max sampling rate of the scope - eg 1GS/s on the PicoScope 3000 series, 5GS/s on the 6000 series.