Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Inconsistent BNC out
#1
Hi Josh and users,
I am using Bpod r2 on windows 10.
I set BNC out channel to be high in a certain state, however it does so only in part of the trials but not always. 
I checked the BNC cable and its fine. 
I don't see any indication of BNC high in the Bpod console, but record it in my recording device that receives that input. 
Any ideas?

Thanks,
Noa
Reply
#2
Hi Noa,


We had another user describe roughly the issue you're having, while triggering a TDT audio processor (RZ6).
We thought the issue was the BNC output isolator IC so they shipped it back to us, but on our end it seemed to work fine.
We're still trying to solve the issue.

Please let us know the following:

-If you connect the output channel to an oscilloscope and use the manual override button from the GUI, do you see a 5V signal when the override button is on, and 0V when it's off?
-Which device are you triggering?
-Does this affect both BNC output channels?

Thanks,
-Josh
Reply
#3
Hi Josh,
Thanks... 
for your questions:
I don't have an oscilloscope but measured the output voltage with a voltmeter while manually controlling the BNC out channel, I do see 0 v when low and 5 v when high. 
I deliver the output signal to TDT RZ5P (similar to the other user you mentioned), I see the problem in both BNC channels, during experiment and also when manually controlling the output. 
When I connect it the a digital input BNC in the TDT the problem is more severe, but occasionally occurred when I connect it to a analog to digital TDT input channel. 

I contacted TDT for this issue as well, and will let you know if we detect a problem there.
Reply
#4
Hi Noa,

I'm also going to contact TDT.

The BNC output channels are driven by a logic isolator IC. It's a fairly unique chip - it contains a tiny (albeit inefficient) isolated DC/DC converter in addition to the standard logic isolation circuitry, to provide power on the far side of the galvanic isolation barrier. When I speak with their engineers, I'll try to find out why it has trouble driving whatever is on their end.

While we're resolving the issue, one thing I'd try is triggering via the 3.3V LED channel on a free behavior port. You'd connect it to TDT via the LED + and - terminals on a port interface board and wire to BNC adapter. Then, in your state matrix (assuming behavior port 4) you'd use: {'LED4', 1} and {'LED4', 0} to make it high (3.3V) and low (0V).

If it turns out neither option works, we'll find out what it takes to trigger their devices and push out a new Bpod module.

To help the TDT engineers understand what we're trying to do, can you please post the exact channel you're connecting to on your RZ5P device?

Thanks, and sorry for all the trouble!

-Josh
Reply
#5
(09-05-2018, 12:44 PM)Josh Wrote: Hi Noa,

I'm also going to contact TDT.

The BNC output channels are driven by a logic isolator IC. It's a fairly unique chip - it contains a tiny (albeit inefficient) isolated DC/DC converter in addition to the standard logic isolation circuitry, to provide power on the far side of the galvanic isolation barrier. When I speak with their engineers, I'll try to find out why it has trouble driving whatever is on their end.

While we're resolving the issue, one thing I'd try is triggering via the 3.3V LED channel on a free behavior port. You'd connect it to TDT via the LED + and - terminals on a port interface board and wire to BNC adapter. Then, in your state matrix (assuming behavior port 4) you'd use: {'LED4', 1} and {'LED4', 0} to make it high (3.3V) and low (0V).

If it turns out neither option works, we'll find out what it takes to trigger their devices and push out a new Bpod module.

To help the TDT engineers understand what we're trying to do, can you please post the exact channel you're connecting to on your RZ5P device?

Thanks, and sorry for all the trouble!

-Josh

Hi Josh,
My colleague solved the issue,
here is the solution if someone else encounter the same problem. 

I had my sound processor (RP2) connected to my main processor (RZ5) using a DB25 digital connector.
This included a pin which overlapped with the physical connection on the RZ5 to which I connected the Bpod output.

As the DB25 connections are bi-directional, this resulted in noise in the channel receiving analog input from the RP2, as well as the unreliable registration of the Bpod BNC input.

I set it up in a way that allows control over individual pins, so once I disconnected the pin connecting the RP2 and the RZ5, the channel was free and everything works.
If it was set up with a single 25 pin cable, I guess I could have proken off the relevant pin to acheive the same effect.

TL;DR - make sure there are no overlapping connections on the relevant channels, which could happen with global connectors, like a DB25 pin connection.

Thanks for the support!
Reply
#6
Here's a reply from another lab that was having trouble triggering TDT (quoted with permission!):

"Our issue was that when the Bpod was hooked up to the Patch Panel "PP24" with a RZ CONN link to the Digital I/O of the RZ6, we could not pull a 5V reading when manually triggering the Bpod with the console. In this case, we measured these conditions:
 - Bpod alone would correctly send out a 5V signal
 - PatchPanel alone would send out a 2/2.5V signal when RZ6 was on
 - When Bpod was connected to Patch Panel, a reading of 2/2.5V was measured. Manually triggering the Bpod resulted in no change in this measured Voltage

I talked with a member of TDT support. He was explaining how the PP24 we have contains a 3.3V pull up resistor that floats the outputs to 2.0-3.3V when the system doesn't have a known voltage reference. So when we originally plugged in the Bpod, I can only assume the patch panel wasn't registering the known reference of the Bpod. This would result in the Bpod sending signals to the PatchPanel, but nothing happened.

Since this seemed to be a problem with the PatchPanel and not the Bpod itself, we drew up a RPvdsEx circuit to send a TTL signal to the Patch Panel out through the C0 port (set to output). This would overcome the pull-up resistor load and ensure the PP24 was working correctly. After this fix, we measured these conditions:
 - Bpod alone would correctly send out a 5V signal
 - PatchPanel alone would  send out a 2/2.5V signal when RZ6 was on
 - Bpod connected to PatchPanel: the reading of 2/2.5V would dissapear as soon as the Bpod was connected and we would measure 0. Once the Bpod was manually triggered, we would measure 5V."
Reply


Forum Jump: