How to achieve USB cable isolation/grounding to reduce hum noise?

[Kouichi Nakamura]

Hi there,

We are trying to record readout from a piezoelectric attached to a spout for water delivery to monitor licking of a mouse while recording neuronal activities in its brain with Open Ephys.

Currently, we're having a massive noise issue in the piezoelectric data, which we suspect is aliased 50 Hz hum noise, and the connection to the Open Ephys acquisition box makes it really worse. While trying many ways of grounding and shielding, we found that inserting an isolator (Nobsound USB to USB Isolator Module Audio Noise Eliminator Industrial Isolator Protection, ADuM4160) between the Open Ephys acquisition box and the PC greatly reduced the noise in the piezoelectric data.

However, it turned out that, with this isolator inserted, Open Ephys GUI cannot recognize the input from the USB and nothing showed up in the Rhythm FPGA module after clicking RESCAN or relaunching the GUI. This seems odd because when we plug or unplug the USB, the PC makes a noise. It appears that while PC can recognize the USB, Open Ephys GUI cannot. Is this expected? Does anyone know a workaround for this problem?

On the same issue, this page says:

Line Noise caused by USB ground
Sometimes, the USB cable that connects the acquisition board to the PC (which is grounded to the building ground via the power supply) isn't good enough -  grounding the external shield of the USB cable that connects the FPGA to the computer can help.

Can someone elaborate more about HOW grounding of the external shield of the USB cable should be done?

Thanks.

Best regards,

Kouichi

Kouichi C. Nakamura, PhD
Senior Postdoctoral Neuroscientist
MRC Brain Network Dynamics Unit
University of Oxford
Mansfield Road, Oxford OX1 3TH, United Kingdom
+44-1865-271582 (office)
http://www.mrcbndu.ox.ac.uk/people/dr-kouichi-c-nakamura
kouichi.c...@gmail.com
kouichi....@pharm.ox.ac.uk

[Jon Newman]

1. One option would be to run your entire acquisition computer off an isolation transformer :).

2. With respect to below

On Thu, Oct 31, 2019 at 8:48 AM Kouichi Nakamura <kouichi.c...@gmail.com> wrote:

Hi there,

We are trying to record readout from a piezoelectric attached to a spout for water delivery to monitor licking of a mouse while recording neuronal activities in its brain with Open Ephys.

Currently, we're having a massive noise issue in the piezoelectric data, which we suspect is aliased 50 Hz hum noise, and the connection to the Open Ephys acquisition box makes it really worse. While trying many ways of grounding and shielding, we found that inserting an isolator (Nobsound USB to USB Isolator Module Audio Noise Eliminatmains or Industrial Isolator Protection, ADuM4160) between the Open Ephys acquisition box and the PC greatly reduced the noise in the piezoelectric data.

However, it turned out that, with this isolator inserted, Open Ephys GUI cannot recognize the input from the USB and nothing showed up in the Rhythm FPGA module after clicking RESCAN or relaunching the GUI. This seems odd because when we plug or unplug the USB, the PC makes a noise. It appears that while PC can recognize the USB, Open Ephys GUI cannot. Is this expected? Does anyone know a workaround for this problem?

Thoughts:

1. That isolator looks very weird. it has a 3rd connection for the DC path isolation and that seems be connected to an anemic looking isolated DC/DC converter. Not sure if the 5V from the USB is used anywhere on the circuit, but if it is, and is being stripped off by the isolator and/or DC/DC converter is overloaded, its not going to work. Here is one that also looks pretty crappy, but at least has the DC/DC converter in series with the In/out usbs as far as i can tell

https://www.amazon.co.uk/Ganquer-Convenient-Isolator-Protection-Isolation/dp/B07Q856GF3/ref=sr_1_9?keywords=usb+isolator&qid=1572621597&sr=8-9

2. How is the Open Ephys box powered?

3. Is there anywhere on the animal size of the isolator that ground is somehow still reach the mains earth connection? Easiest way to test is with a multimeter: stick one prong in the 3 connection of a socket in the room and the other on what supposed to be isolated ground on the headstage.

 

On the same issue, this page says:

Line Noise caused by USB ground
Sometimes, the USB cable that connects the acquisition board to the PC (which is grounded to the building ground via the power supply) isn't good enough -  grounding the external shield of the USB cable that connects the FPGA to the computer can help.

Can someone elaborate more about HOW grounding of the external shield of the USB cable should be done? 

Thanks.

Best regards,

Kouichi

Kouichi C. Nakamura, PhD
Senior Postdoctoral Neuroscientist
MRC Brain Network Dynamics Unit
University of Oxford
Mansfield Road, Oxford OX1 3TH, United Kingdom
+44-1865-271582 (office)
http://www.mrcbndu.ox.ac.uk/people/dr-kouichi-c-nakamura
kouichi.c...@gmail.com
kouichi....@pharm.ox.ac.uk

--
You received this message because you are subscribed to the Google Groups "Open Ephys" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-ephys+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/4f43425d-3779-48d4-98b7-d4b295829103%40googlegroups.com.

--

Jonathan Newman

Postdoctoral Fellow, MIT

http://www.mit.edu/~jpnewman/

[Kouichi C. Nakamura]

Thank you, Jon, for your thoughts and advices.

1. One option would be to run your entire acquisition computer off an isolation transformer :).

We'll consider this.

1. That isolator looks very weird.

We've ordered the one you suggested. Thanks.

2. How is the Open Ephys box powered?

At the moment, from the mains with a normal cable. But we've bought a replacement to make it better (sorry, I don't have specifics right now).

3. Is there anywhere on the animal size of the isolator that ground is somehow still reach the mains earth connection? Easiest way to test is with a multimeter: stick one prong in the 3 connection of a socket in the room and the other on what supposed to be isolated ground on the headstage.

At the moment, the Faraday cage is not grounded to the earth. But we have tested it for a short while, and didn't see much improvement. Nevertheless, we're planning to ground everything as much as possible to the earth.

Best regards,

Kouichi