I had an O2 signal on ch A, then I was about to look at injector voltage on ch B and set it up with an attenuator (3000srs Pico). Ch B "zero-line" was jumping up and down with the O2 signal. (weird lambda injector psdata)

Changed the setup to a 1:1 probe, and did even unplug the lead from Ch B, still looked like this: (weird psdata file)

Altering the timebase changed it to normal. Why did this happen? Ground was good, checked it many times over. Grounding Ch B didn't change a thing.

---

That's very interesting.  What software build were you using?

Can you reproduce it by going back to the same time base again?

What model scope?  3223?

- 6.7.21.2
- err... 3423? 3000-series 4ch automotive

I still have this car here, I will try later today to reproduce, and also try different combinations of channels.. Yesterday I was just leaving the garage when this came up.

PicoFin wrote on May 29^th, 2013 at 8:45pm:

3423? 3000-series 4ch automotive

Oh duh.  It's there in your file.


PicoFin wrote on May 29^th, 2013 at 8:45pm:

I will try later today to reproduce, and also try different combinations of channels..

I tried to duplicate here using your file as a setup with a 3423 and 6.7.21.2, and was not able to.  If it was a one off event and you cannot duplicate it then it was likely a PC psychotic episode.   [smiley=justcrazy.gif]

Ok. I had the very same car in my garage, so I did the exact same measurement earlier today and was not able to duplicate the issue, no matter what.

Later today, I saw some strange activity when I was looking at my dad's cars seat belt tensioners as I know one is faulty. I noticed I have activity on both active channels A and B, even if I only actually measured with one of them.

I used the belt tensioner circuit for these attached captures, in both captures there is one full screen. At the beginning, nothing is attached to the scope. Then, I sequentally plug the very same BNC to each channel from A to D. These captures are made the same way, only the scope settings have changed.

It has previously been in to Pico for fried ch A repair, so I will contact them with this issue. This is here just to be looked at.

---

PicoFin wrote on May 30^th, 2013 at 11:55am:

I will contact them with this issue. This is here just to be looked at.

OK.  Just refer them here for the story.  I don't think I have seen anything like this before.  Still cannot duplicate that here so I do not think it is a software bug.

Pocket magnets?

I keep catching myself with that one...

Good point, James.  Not the effect I usually see, but magnets are kryptonyte to Pico.  Never have your Pico box anywhere near any kind of magnet.

Ok, I will have to check that, then. What would be a safe distance to any normal magnet in a shop? Pocket magnets, magnetic trays, set of magnetic sockets..

The changes I have recently made, is a self fabricated steel tube boom, and the USB-cable is a non-Pico 3m/10ft one.

I just can't imagine a USB cable could cause only one channel to act like this, rather it would terminate the connection between the PC and scope, right?

PicoFin wrote on May 30^th, 2013 at 8:37pm:

What would be a safe distance to any normal magnet in a shop?

That depends on the strength of the magnet.  If you have Pico set up on a cart, have no magnets on or around that cart.  I would say at least two feet away for any magnetic tray or tool holder.


PicoFin wrote on May 30^th, 2013 at 8:37pm:

rather it would terminate the connection between the PC and scope, right?

No.  It makes odd things happen.  The problem relates to poorly manufactured USB ports on some PCs where the contact areas are reduced and the port is sloppy.  The Pico cable is designed to compensate for these cheap USB ports with more contact area in the cable itself.  This issue affects the 4000 series scopes.  Pico makes a long version of the special blue USB cable.  These do solve this problem.

These issues are common and important issues so we are sharing them here for any Pico users.