How not to blow my PC based USB oscilloscope

[abhayb_e]

Hi,
   I am planning to buy an PC based USB oscilloscope from picotech. And i am concerned about how to prevent from blowing all three (1) My Laptop(to which i will connect my oscilloscope through USB) - Dell Inspiron N4110
         (2) USB oscilloscope
         (3) My board >Pandaboard which i aim to probe. I will power pandaboard with AC to DC adapter(5V-2A output, having 2 pin, no earth pin).
I saw Dave's video "EEVblog #279 - How NOT To Blow Up Your Oscilloscope!". Also read the post "Test Equipment / Interface Isolation on Entry Oscilloscopes? (+USB Scopes, A Hard Lesson Learned)". I tried to add in same post but it said its >120 days old so i made a new post.
So my question as given below
"Suppose i remove my laptop charger. Now according to me laptop is powered by battery and its ground is not connected to earth ground. Then i connect usb oscilloscope to PC. Pandaboard is connected to adapter as mentioned above. Now i start probing the board. If by mistake i touch probe ground to a point on board which is not ground & has some voltage level, but circuit will not complete through earth ground because my laptop is not connected to earth ground, and nothing will blow up".
So what i could conclude that disconnecting the laptop charger should prevent the devices from blowing because low impedance circuit through earth ground is not complete.
I request experts to review this conclusion since i am a newbie.
Also in the above post("Interface Isolation on Entry Oscilloscopes?"), there is a mention of differential probes. Does differential probe prevent the devices from blowing ? Again i need expert advice.
- Abhay

[BradC]

So what i could conclude that disconnecting the laptop charger should prevent the devices from blowing because low impedance circuit through earth ground is not complete.

Correct. I thought the same thing and used the same process to debug a mains power meter. Unfortunately there is always that one time you forget to unplug the charger and reduce what was a very expensive laptop to a pile of smoking parts.
It's quite astounding what goes up in smoke and the astonishingly loud noise it makes when you poke 240v into a USB port. 

[Jeroen3]

With a bit of luck your scope uses an isolated USB interface. Check the manuals.
Otherwise you can safely use your laptop without any external powered peripherals connected.
Although your laptop supply uses galvanic isolation, it will introduce significant noise.

You can also isolate your dut, using a lab power supply. The wall adapter you have is probably isolated as well, but with different quality circuit than a lab supply. Non-isolated separate power supplies are rare. Most of them are found inside inseparable assemblies, such as LED light bulbs, variacs excluded.

This is a widely discussed topic on the forum, I'm sure you can find a lot more info.

[mzzj]

So what i could conclude that disconnecting the laptop charger should prevent the devices from blowing because low impedance circuit through earth ground is not complete.

Correct. I thought the same thing and used the same process to debug a mains power meter. Unfortunately there is always that one time you forget to unplug the charger and reduce what was a very expensive laptop to a pile of smoking parts.
It's quite astounding what goes up in smoke and the astonishingly loud noise it makes when you poke 240v into a USB port. 

Disconnecting laptop charger is equally bad option: If you connect your scope ground clip to live 240v then every metal part of your scope and laptop is waiting for a change to kill you.
Having said that I have done it also myself, but paying EXTRA attention not to touch the scope. And I would not use it on a daily basis.

[David Hess]

Besides having the laptop disconnected from the AC line, if you want it isolated, you also have to make sure the laptop is not grounded through any attached devices.

A better option is to use the laptop normally if only for safety and make sure the USB interface is galvanically isolated.  If the Picoscope does not provide USB isolation, and as far as I can tell they do not (*), then an external USB isolator should be used.  Beware that most USB isolators are limited to 12 Mb/s even if they support USB 2.0.

Also in the above post("Interface Isolation on Entry Oscilloscopes?"), there is a mention of differential probes. Does differential probe prevent the devices from blowing ? Again i need expert advice.

Given the cost of a high voltage differential probe, I would also plan on buying a USB isolator and using both.

(*) Picoscope's official replies on this subject by themselves would lead me to avoid their products.  They have a casual disregard for safety in their responses about this issue, in their documentation, and in their designs.  Of course most other manufacturers of USB instruments are no better but that is why I have avoided USB based instruments in general.

[abhayb_e]

Pandaboard power supply is 5V DC, on the board voltage is mostly 1.8V or 3.3V or max 5V( as given in http://pandaboard.org/sites/default/files/board_reference/pandaboard-ea1/panda-ea1-manual.pdf) which is way below dangerous level i think so it is safe to unplug the laptop charger.
As mentioned by David & also mentioned at http://www.picotech.com/support/kb/Oscilloscopes-kb190. Also i could not find in specs the mention for supporting galvanic isolation.
As suggested i am planning to buy USB isolator and also Lab power supply(isolated).

[David Hess]

Pandaboard power supply is 5V DC, on the board voltage is mostly 1.8V or 3.3V or max 5V( as given in http://pandaboard.org/sites/default/files/board_reference/pandaboard-ea1/panda-ea1-manual.pdf) which is way below dangerous level i think so it is safe to unplug the laptop charger.

I agree it will be safe enough.  If a ground loop is created through a non-isolated USB port, a trace may still be blown out on something important like the motherboard even at low voltages if the current is not limited.  The high current output from an ATX power supply for instance can cause considerable damage.

[cybertronicify]

I did that with my ACER ASPIRE laptop and i forgot to unplug the AC charger and it blew one of my usb ports out. I made a USB isolator, 5v coming from an isolated supply, and the data is passed through optocouplers and isolated. Sounds like a lot of work but its well worth it. Luckily my laptop survived the blow, but one of the usb ports is still out. I took my laptop apart and theres a nice burn mark around the TVS diode, interestingly i still get power from the usb port.

[David Hess]

I learned this lesson long ago with RS-232 ports but still shy away from USB test instrumentation because of the potential for damage. 

At least isolated USB hubs are available now but most of them are limited to 12 Mb/s.  A better solution if isolation is desired is Ethernet and many test instruments support it.

I do have a funny story involving RS-232 and a blown ground trace.  One day on my workstation my external modem which I used for my internet connection started displaying this strange behavior where it would pick up, dial out, negotiate a connection, connect, and then immediately disconnect.  What happened is that I had inadvertently blown the ground trace off of the RS-232 interface board with a 12 volt supply but the modem's data carrier detect line was holding the RS-232 voltage levels in place until carrier was detected at which point the modem saw data terminal ready change state so it would hang up.

[Rick Law]

If you want to "float" your laptop ground, don't overlook these:

1. Hard-wired network
2. KVM
3. External video/audio

These needs to be disconnected as well.

[AG6QR]

I hope everyone on this thread sees the following as obvious, but I'll throw it out anyway, because it has been known to cause issues...

Even if you get a completely isolated scope, whether by way of isolating the laptop from mains, or by way of isolating the USB scope from the laptop, don't forget that you haven't isolated the channels from one another.  Your ground leads are all connected together through the scope.  If you plug two ground leads from two different channels to two different points on the DUT, you have just shorted out those two points through your scope.  Failure to think about this before you hook up your ground leads may result in various things letting smoke out or behaving dangerously or violently.

[Jeroen3]

If you need isolated channels you will need to buy a special battery powered scope.

Ethernet is isolated upto 1500 Vac, except when you're using RJ45 plugs with shielded connectors. They are expensive and thus not that common.

[abhayb_e]

Thanks all, for sharing information on how to protect my devices from burning.Earlier i did not know that its so easy to destroy my expensive devices.
Before i connect my scope to laptop, As a precaution i will disconnect all the grounded connections to laptop, charger/ethernet wire/external speaker. Also i will use lab power supply(isolated).
There is one more point mentioned by AG6QR. There are two channels in my picoscope, i should be careful with that also. I will first check with multimeter(i have Brymen, Vici) point in DUT which i want to probe. But by mistake i might touch some point on dut unintentionally, so i need to be very careful. I will contact picotech on this & update here.

[David Hess]

Before i connect my scope to laptop, As a precaution i will disconnect all the grounded connections to laptop, charger/ethernet wire/external speaker.

Ethernet using UTP (unshielded twisted pair) is safe to leave connected.  It includes transformer based galvanic isolation at each interface.