USB Oscilloscope for undergrad student

[MrW0lf]

Testing for wfm/s quite simple with signalgen that has burst mode (see if all edges get caught to segmented mem) or rising+falling trigger pwm technique. Maybe Owon/AD2 owners try look at this deeper. I'm currently on the run.

[Ronny]

Okay folks,

After reading through this thread I notice 2 kinds of comments:
1 - Yeah go and buy it
2 - Use university lab / build your own lab

People on number 2 forgot one detail, I am an exchange student, to be specific, from Brazil! So nah, building a lab in my home would cost me the equivalent of one semester of tuition. The university has 2 kinds of labs, the one you can just go and use, it is full of bad scopes, and the nice ones, which are a pain to get in to.

So for me the choice was obvious, just ordered the AD2 + BNC board. I also happened to see a pickit 3 there, it was the cheapest place to buy it, I was going to buy it too so I threw it in there making a total (with shipping) of $250 
Whatever, it is done now haha.

So, I followed your advice and did not order the probes from Digilent. Are any of those here good?
https://www.amazon.com/RioRand-Oscilloscope-Clip-Probes-Accessory/dp/B0030KWM30/ref=sr_1_1?ie=UTF8&qid=1492252259&sr=8-1&keywords=oscilloscope+probes

https://www.amazon.com/Hantek-200MHz-Oscilloscope-Switchable-Accessory/dp/B005O0V01E/ref=sr_1_4?ie=UTF8&qid=1492252259&sr=8-4&keywords=oscilloscope+probes


And lastly, I am starting to wonder if USB isolation is really needed, I will use it with a laptop running of batteries. If you still think it is needed, can you recommend me one from US Amazon? I don't understand anything about USB isolators.
Thank you very much guys.

[rstofer]

Re: USB isolators.

I plug my AD into a powered USB hub such that the power comes from a wall wart, not the PC power supply.  On the AD1, the power supply outputs come from the PC via the USB cable.  On the AD2, you can use an external wall wart and you might want to look into that.  It not only saves on the PC, it increases the current capability of the power supplies.

The reason for isolation is the off chance that you make a mistake in grounding and the PC ground gets used as a path for some really high currents.  This simply shouldn't happen.

First, know what you are probing.  Second, keep your probe on 10x - always.  Yes, that limits the very low end of voltage measurements but it improves frequency response (reduced circuit loading).  In the last 13 years of messing around I can't remember EVER using 1x.  Third, pay attention to the voltage limits of the device.  There is excellent documentation at Digilent.

If you simply have to get involved in high voltage (higher than 12V in my view) and high current (more than a couple of amps), use the school's lab.  Why do you suppose the scopes don't work?  High voltage, high current, low thinking.

Treat it gently and I think it will last a very long time.  You might check that the digital IO is all 5V tolerant.  Otherwise, the AD2 would prefer to talk to 3.3V logic.  Some Arduinos are 3.3V and some are 5V.  Make sure you know what you are probing!

I would also suggest that you disconnect the AD from stuff that is powered when it isn't.  In fact, put it back in the box when its not in use.  Treat it with respect...

I guarantee you will be impressed!

ETA:  Don't loan it out!  By now you probably understand why the school's scopes are junk!  Same guy...

[Ronny]

Re: USB isolators.

I plug my AD into a powered USB hub such that the power comes from a wall wart, not the PC power supply.  On the AD1, the power supply outputs come from the PC via the USB cable.  On the AD2, you can use an external wall wart and you might want to look into that.  It not only saves on the PC, it increases the current capability of the power supplies.

The reason for isolation is the off chance that you make a mistake in grounding and the PC ground gets used as a path for some really high currents.  This simply shouldn't happen.

First, know what you are probing.  Second, keep your probe on 10x - always.  Yes, that limits the very low end of voltage measurements but it improves frequency response (reduced circuit loading).  In the last 13 years of messing around I can't remember EVER using 1x.  Third, pay attention to the voltage limits of the device.  There is excellent documentation at Digilent.

If you simply have to get involved in high voltage (higher than 12V in my view) and high current (more than a couple of amps), use the school's lab.  Why do you suppose the scopes don't work?  High voltage, high current, low thinking.

Treat it gently and I think it will last a very long time.  You might check that the digital IO is all 5V tolerant.  Otherwise, the AD2 would prefer to talk to 3.3V logic.  Some Arduinos are 3.3V and some are 5V.  Make sure you know what you are probing!

I would also suggest that you disconnect the AD from stuff that is powered when it isn't.  In fact, put it back in the box when its not in use.  Treat it with respect...

I guarantee you will be impressed!

ETA:  Don't loan it out!  By now you probably understand why the school's scopes are junk!  Same guy...

One of those?
https://www.amazon.com/Protronix-Port-USB-Power-Adapter/dp/B00REX6DRK/ref=sr_1_7?s=pc&ie=UTF8&qid=1492267504&sr=1-7&keywords=powered+usb+hub

If that protects my laptop I am going to buy it 

[rstofer]

The USB hub looks like a much better device than mine - notably, it does USB 3.0

That helps cover the problem of overcurrent but it won't do anything to prevent a gigantic ground current travelling through the USB cable into the PC chassis and out the power cord to ground.  That's why a lot of people run their PC on battery - it removes the ground path. 

I haven't done that.  Maybe I should but, frankly, I have never used my AD on anything over a few volts.  It just hasn't come up!  My main power supply has current limiting (and the AD has severe current limiting, especially when powered only through the USB port) so I don't expect to see a lot of current coming down the ground wire.

Do you expect to do a lot of mains work or perhaps mess around with switching mode power supplies?  I wouldn't do that with an AD even though it is definitely workable.  You just need to watch Dave's video on how not to destroy a scope.

Pay attention to this:  When you run the AD with flyleads, the scope channels are floating.  You need to hook up both leads for each channel although you may often ground the striped wire (you'll see what I'm talking about), or you can take a differential measurement across a resistor, for example.  Nevertheless, there is no ground reference even though one of the black wires probably needs to be connect to the DUT ground.  RTFM re: ground reference.

However:  When you use the BNC adapter, the shell of all 4 BNC connectors ARE connected to ground - just like the ground lead on a real scope or the BNC on a signal generator.  All of a sudden, you have lost the differential signaling capability that was provided by the fly leads.  The fly leads had a common mode limit (voltage above ground) but the inputs were truly differential.

I encourage you to read everything you can at the Digilent site.  Maybe watch the videos...  The AD is a fantastic tool but it's usage may not be self-evident.

But truly differential inputs, how cool is that?  Sometimes you want to measure the voltage drop across a component that isn't grounded.  Yes, you will have to remove the BNC adapter and use the flyleads but, when you really need to do this, the AD is a nice way to get it done.

[Ronny]

The USB hub looks like a much better device than mine - notably, it does USB 3.0

That helps cover the problem of overcurrent but it won't do anything to prevent a gigantic ground current travelling through the USB cable into the PC chassis and out the power cord to ground.  That's why a lot of people run their PC on battery - it removes the ground path. 

I haven't done that.  Maybe I should but, frankly, I have never used my AD on anything over a few volts.  It just hasn't come up!  My main power supply has current limiting (and the AD has severe current limiting, especially when powered only through the USB port) so I don't expect to see a lot of current coming down the ground wire.

Do you expect to do a lot of mains work or perhaps mess around with switching mode power supplies?  I wouldn't do that with an AD even though it is definitely workable.  You just need to watch Dave's video on how not to destroy a scope.

Pay attention to this:  When you run the AD with flyleads, the scope channels are floating.  You need to hook up both leads for each channel although you may often ground the striped wire (you'll see what I'm talking about), or you can take a differential measurement across a resistor, for example.  Nevertheless, there is no ground reference even though one of the black wires probably needs to be connect to the DUT ground.  RTFM re: ground reference.

However:  When you use the BNC adapter, the shell of all 4 BNC connectors ARE connected to ground - just like the ground lead on a real scope or the BNC on a signal generator.  All of a sudden, you have lost the differential signaling capability that was provided by the fly leads.  The fly leads had a common mode limit (voltage above ground) but the inputs were truly differential.

I encourage you to read everything you can at the Digilent site.  Maybe watch the videos...  The AD is a fantastic tool but it's usage may not be self-evident.

But truly differential inputs, how cool is that?  Sometimes you want to measure the voltage drop across a component that isn't grounded.  Yes, you will have to remove the BNC adapter and use the flyleads but, when you really need to do this, the AD is a nice way to get it done.

I was about to order the HUB then I notice something. I intend to used it with a laptop running out of batteries, this way I can't close the ground loop. When a powered hub is connected, my AD2 will have a path to ground in it, closing the ground loop.
It seems that it is much safer to use it without the powered hub, perhaps just a normal hub.
I started to look into USB isolators, the only one in Amazon with a reasonable price is this one:
https://www.amazon.com/SMAKN-Isolator-Digital-Isolation-Industrial/dp/B00XXPO4UG/ref=sr_1_1?ie=UTF8&qid=1492571592&sr=8-1&keywords=usb+isolator
But it seems to be extremely slow. The fast ones are extremely expensive.
It really seems that the best option for me is to run it off a laptop running out of batteries and have a simple, non-powered,  HUB as a "spike shield".

What do you guys think?

ps. My AD2 has arrived, looking good, I am still waiting for the probes tough. Is there any way to test it without any lab equipment? I only have the stuff in my dorm, otherwise I will need to get access to the lab (I don't know how to do this yet, since I am an exchange student). I don't wanna discover a problem in it once I am back in Brazil, where, most likely, there will be no warranty.

Thanks!

[mtdoc]

I was about to order the HUB then I notice something. I intend to used it with a laptop running out of batteries, this way I can't close the ground loop. When a powered hub is connected, my AD2 will have a path to ground in it, closing the ground loop.
It seems that it is much safer to use it without the powered hub, perhaps just a normal hub.

Keeping in mind the issues pointed out by rstofer, in almost all circumstances a hub is unneeded and you're right, the nice thing about using the AD with a laptop running off of battery is that you have a floating scope. A grounded usb hub would defeat that.

ps. My AD2 has arrived, looking good, I am still waiting for the probes tough. Is there any way to test it without any lab equipmen

You can loop the wavegen output into the scope and play with those and I believe you can loop the "patterns" output into the logic analyzer and play with those.

[rstofer]

ps. My AD2 has arrived, looking good, I am still waiting for the probes tough. Is there any way to test it without any lab equipmen

You can loop the wavegen output into the scope and play with those and I believe you can loop the "patterns" output into the logic analyzer and play with those.

Yes, you can do both of these and it can be good practice using the UI.

Grab a resistor and capacitor (1k 0.1 ufd?  3 dB down at 1.5 kHz) and build a low pass filter.  Here's where the floating nature of the AD inputs comes into practice:  Use the signal generator to send a 1V p-p input, connect one scope channel across the resistor (floating measurement of current) and the other channel across the capacitor to measure voltage.  You can run through a bunch of frequencies and watch the output.  You could also take Ch 1 off of the resistor and put it across the input.  Be aware:  You can't float the scope inputs when you use the BNC adapter!  The shields are grounded!

Better yet: Use the Network gadget to sweep a range of frequencies and create a Bode' plot of phase and magnitude.  There's not an EE student alive who wouldn't appreciate the convenience of this tool!