DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Having problems ? let us know the details here
Post Reply
CaliGeek
User
User
Posts: 4
Joined: Fri May 07, 2021 9:03 pm

DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Post by CaliGeek »

I have a DrDAQ running with PicoLog 6 on a Windows machine. My primary intent is to use this setup to capture a count rate on two inputs along with other analog data and switch inputs. The desired count rate will generally be between 0 and 2000 counts per minute but will vary randomly. It appears that PicoLog is setup to measure frequency and not specifically counts per unit time. I can get the software to do the math to convert to counts per minute and that works such as it is, but I have a question and also a problem that prevents this setup from being usable.

1. Question: At a 1 minute sample rate, will this setup accurately measure the number of pulses during that minute? I'm worried that the frequency measurement scheme may not accurately reflect the number of counts (because I don't know how it is being measured and computed in this setup and the issue below has me concerned).

2. Problem: When plotting a constant pulse rate (from a signal generator, 5V, 1ms pulses) in PicoLog, there are large spikes in the frequency. I could see this being a noisy input, but some of the spikes are negative, so I don't think a negative frequency can be caused by noise. That sounds like some sort of overflow in hardware or software. I've tried frequencies between 10 and 2000 Hz, and 1 second and 1 minute sample periods and get the same issues. At 1 minute sample period the spikes get averaged somewhat, but are still massive compared to the actual signal (10Hz input but 600Hz spikes). At a 1 second sample period the spikes reach several 10's of kHz. In case it matters, I am collecting data from the built in light, sound, and temperature sensors as part of my testing.

Thanks in advance for any help offered.
Attachments
Freq Spikes.jpg

CaliGeek
User
User
Posts: 4
Joined: Fri May 07, 2021 9:03 pm

Re: DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Post by CaliGeek »

Perhaps this is a clue? I noticed that the spiking in the frequency data was tracking, at least roughly, a ratio with the value returned by the light sensor. I removed the light sensor from the data collection and then the spiking started tracking roughly with the temperature sensor. I tried different frequencies and pulse widths and it didn't seem to make a significant difference, but that was casual testing. To be clear, I have no reason to suspect there are actual changes in frequency (especially given the "negative" frequency spikes), this really looks like a data overflow or addressing issue.
Attachments
PicoLog Spikes Tracking Temperature.png

Martyn
Site Admin
Site Admin
Posts: 3992
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Post by Martyn »

Can you post a picolog file with this data
Martyn
Technical Support Manager

CaliGeek
User
User
Posts: 4
Joined: Fri May 07, 2021 9:03 pm

Re: DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Post by CaliGeek »

Here's the full picolog file. It's roughly 2 days worth of data.

Note that I was making changes during this run to observe the impact on the data. In addition, at some point the board stopped communicating with the software, but I don't remember if that was before or after I saved the attached picolog file. If you need a clean file, I'll do another run.
Attachments
May 9, 2021 1-54-08 PM.picolog
(481.5 KiB) Downloaded 22 times

Martyn
Site Admin
Site Admin
Posts: 3992
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Post by Martyn »

Thank you for the file, we need to discuss this with the development team to understand what is happening.
Martyn
Technical Support Manager

CaliGeek
User
User
Posts: 4
Joined: Fri May 07, 2021 9:03 pm

Re: DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Post by CaliGeek »

I've done some more testing and it appears that the frequency measurement spikes only occur when an analog channel is used along with a digital channel in frequency mode. Any combination of digital channels works fine, but as soon as I add any analog channel (light, sound (both "waveform" and dB), temperature, resistance, "external", or scope) the spikes show up. Interestingly, it does not show up when I enable the ph channel and I wonder if that's because the input is flagged as out of range and perhaps doesn't try to read it or log it (I don't have a ph probe).

Note this last batch of tests were done with digital input 1 firmly pulled low via a 1k resistor (and no input signal) so there was no chance that input noise was a factor.

Gerry
PICO STAFF
PICO STAFF
Posts: 947
Joined: Mon Aug 11, 2014 11:14 am

Re: DrDAQ + PicoLog6 Large spikes in Freq. Measurements (positive and negative)

Post by Gerry »

Hi CaliGeek,

I have modified this response as we are looking into this further and will be responding shortly.

Regards,

Gerry
Gerry
Technical Specialist

Post Reply