Data Logger with Scope Inputs

Which product is right for your exact requirements
Post Reply
d431hood
Newbie
Posts: 0
Joined: Wed Sep 16, 2015 1:13 pm

Data Logger with Scope Inputs

Post by d431hood »

Dear all, Forgive me if this has been asked before.

I need a basic datalogger with very high voltage inputs (differential). To log four to six channels.

However I only usually need to gather data at about 1000Hz (i.e. 4000 to 6000 samples per second), but 2000Hz or 5000Hz might be useful. It needs to run continuously to a display and trigger on defined events. Synchronisation is only loosely important... i.e. it doesn't need to be fantastically accurately coordinated.

I have a number of 200x isolators with probe type outputs that work with standard oscilloscopes (raw voltages are typically -100-0-100v, 0-1000v, and 500-1200v), and these can be reused for the logger.

As this is the real world the budget for this is very modest.....

Could I use the 2204 to record 2 channels continuously to a PC at 4000samples per second?
Could I add a second or third 2204 to record four six channels to the same file at 6000samples per second over three separate USB 2 channels to one PC? If not is it possible to record three separate files on one PC and then stitch records together to align the data (to 0.1 sec or ideally 0.01 sec) create one file?

Recordings might need to be 24hrs (ideally 36-48hrs) but will typically be 4-6hrs long.

All help appreciated.

Newbie

Hitesh

Re: Data Logger with Scope Inputs

Post by Hitesh »

Hi d431hood,

Although it is possible to connect more than one PicoScope 2204A to a PC and then use the PicoLog data logger software to collect data from all the devices (3 in this case), the smallest sampling interval would be 1ms so you would not be able to achieve 4000 samples per seconds.

Are you expecting several trigger events during the course of the data capture? Are you looking to capture all of the data over that time interval or just data around the trigger events?

If you only need to capture 4 - 6 hours at a time, then you might be able to use the PicoScope 4824 to capture data on 6 channels using the PicoScope 6 oscilloscope software. A timebase of 2000s/div will give you around 5.5 hours.

How do you intend to post-process the data as this will determine whether the above is suitable for your application.

Regards,

d431hood
Newbie
Posts: 0
Joined: Wed Sep 16, 2015 1:13 pm

Re: Data Logger with Scope Inputs

Post by d431hood »

Thank you for replying Hitesh.

In practice we will probably spend most of the time gathering data continuously (for 4-6 hours) and ideally store the data in separate files, each approx. one hour long. If we can do 4 channels at 250Hz (1000samples per second) as a maximum that might be a little limiting but not completely unusable. Is the 1ms sample interval a fixed value or dependant on number of connected devices/channels?

Trigger events when used will be simple: Eg volts on one channel are above (or below) value X, and then the recordings would be a lot shorter (in time). Higher resolution may be useful under these circumstances but we do have 4 channel DSOs we could use for this.

Post processing the data maybe be just visual inspection or simple-ish mathematics that appear to be supported by the Picolog software.

Unfortunately the 4824 is well beyond the budget available to me, and I doubt I can get enough justification to get the budget raised. However we probably could afford a 3403D: would I be right in saying that its 64MS buffer could store 4 channels at 4000S/s for 4 hours? And if so what time (typically) would be taken to do the buffer transfer to the PC (USB3)... Would I be right to assume the recording stops to let the buffer empty even it the (alarm?) macro then kicks off recording again afterwards?

Hitesh

Re: Data Logger with Scope Inputs

Post by Hitesh »

Hi d431hood,

The 1ms sample interval limit is set by the PicoLog software. A relatively faster sampling rate can be set when using the Software Development Kit (SDK) in order to create your own application.

The PicoScope 3403D will only work with the PicoScope 6 software or with an application created using the SDK but not with the PicoLog software.

Collecting data at 4kS/s with 4 channels enabled for 4 hours would result in just under 220MS so it would be better to collect data in streaming mode (refer to our Library A - Z of PC Oscilloscopes page for further information).

Having checked with the PicoScope 6 software, the maximum sampling rate with a 2000s/div timebase is 1.25kS/s with all 4 channels enabled on a PicoScope 3404D (this uses the same driver as the 3403D) so you would have to collect data over a shorter period of time, or use the SDK. There is a maximum record length of 100MS shared between all channels when collecting data in streaming mode using the PicoScope 6 software.

The software will stop data collection while it saves data to disk before restarting capture (assuming that this has been set in the Alarms dialog).

Hope this helps.

d431hood
Newbie
Posts: 0
Joined: Wed Sep 16, 2015 1:13 pm

Re: Data Logger with Scope Inputs

Post by d431hood »

Thank you Hitesh....

Either we are talking at cross purposes or I have basic mis-understanding of the terminology.

Based on my past datalogger experience (Turboview - Overpower Monitor): To my mind 4000S/s is number of channels independent, but sample frequency is not. So to my mind 4000S/s is 4 channels at 1000Hz. Then 4hrs of recording across 4 channels at 1000Hz would be 4x3600x1000x4=57.6MS and should fit in the advertised buffer of the 3403D (i.e. 64MS).

Also can you give me any clue about the buffer down load time, I know it is going to be PC and USB type dependant, but are we talking: a few seconds, 10s of seconds, a minute or two, or longer?

Thanks

Hitesh

Re: Data Logger with Scope Inputs

Post by Hitesh »

Hi d431hood,

Looks like there has been a misunderstanding - I thought you were referring to sampling rate per channel which is what we use for our the specifications on our PicoScope products.

If you do not need to see the data as it is collected then you can adjust the Slow Sampling Transition point in the PicoScope 6 oscilloscope software's Preferences dialog, in order to collect the data in block mode. Here the data is collected into the device memory and then transferred to the PC on completion of the data collection. Note that the memory is shared between all channels.

If you need to see the data as it is collected then use streaming mode data collection (the transition point is set by default in the software to a timebase of 200ms/div or above).

A 1kS/s sample rate per channel is possible with 4 channels enabled with a timebase setting of 2000s/div giving you 5h33m20.0s across the screen.

In terms of transfer rates after a capture in block mode, we quote 3.5s in the Programmer's Guide to retrieve the full 128MS from a PicoScope 3206B (now obsolete), so it should be less than this over USB 3.0. If you are capturing data in streaming mode, the data will be transferred to the PC as it is collected in chunks.

Regards,

bennog
Advanced User
Advanced User
Posts: 206
Joined: Mon Nov 26, 2012 9:16 am
Location: Netherlands

Re: Data Logger with Scope Inputs

Post by bennog »

In streaming mode you can also directly see the signal on screen while it is captured.
So you don't have to wait 5 hours to see the data you have captured.
At the speeds you suggested streaming mode is no problem and you will not be limited by scope memory but by the limitation of the PicoScope software of 100Msamples total.

Benno

Post Reply