Author Topic: Analog Discovery differential input  (Read 3555 times)

0 Members and 1 Guest are viewing this topic.

Offline not1xor1Topic starter

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Analog Discovery differential input
« on: July 24, 2017, 04:38:23 pm »
Hi, ladies and gentlemen  :)

I just got an Analog Discovery 1 rev.C from ebay.

It is second hand and it is just the basic kit (less the male jumpers).

I noticed that the BNC adapter board doesn't make use of the differential input connecting the negative input to ground.
I was planning to make use of the differential input feature (yes I know the circuits should share a GND connection) for measuring power supplies ripple, etc., but I read (and have just noticed how bad a 5MHz square wave looks on the scope tab) that without a proper adapter board the usable bandwidth is reduced to few MHzs, while with the BNC adapter board it gets around 30MHz (-3dB).

So I wonder if:
- somebody else made a dual differential input adapter board
- there is something which I do not know that make such a thing likely to be impracticable
- what should I care about if I have to devise that myself (apart tracks as short as possible and of the same length and parasitic capacitance)
- should I move this to a different section (i.e. Beginners, or Project, Designs, etc.) ?

thanks

 

Online rstofer

  • Super Contributor
  • ***
  • Posts: 9886
  • Country: us
Re: Analog Discovery differential input
« Reply #1 on: July 24, 2017, 05:04:20 pm »
The BNC adapter makes the AD scope operate the same way other scopes operate; one side of the input connector is ground.

In this case, you need to buy a differential probe and that will set you back a bunch of money.  Dave sells one:
https://www.eevblog.com/product/hvp70/

So, how in the world have I been able to work in electronics for decades without a differential probe?  Beats me!  I guess I just work with circuits that have one side grounded.

The advantage to the BNC adapter is that it allows 10x probes.  Sure, this helps with bandwidth but it also helps prevent over-volting the ADC.

There are times when a differential input is nice.  Consider an RC integrator circuit where you want to watch the capacitor voltage on one channel (one side of capacitor is ground) and also the capacitor current by measure the voltage drop across the resistor where there is no ground reference.  Without a differential input, you just measure the capacitor voltage and realize that the resistor is dropping the difference from the input voltage and you can calculate the current.  You don't get a nice 2 channel trace but so be it.

I use my AD2 with and without the BNC adapter.  If the voltages are low, I don't have to worry about destroying the AD.

Think about what you are measuring and see if there isn't some way to do it without a differential input.
 

Offline alm

  • Super Contributor
  • ***
  • Posts: 2840
  • Country: 00
Re: Analog Discovery differential input
« Reply #2 on: July 24, 2017, 05:11:47 pm »
Another application for differential inputs is to avoid ground loops. This can be helpful in low level measurements, like power supply noise.

I think the bandwidth limitation might be mostly about the flying leads from the AD that lack any shielding or defined impedance. Does the adapter board contain any termination resistors? If not, then you may be able to make a cable that connects the differential signals to a twisted pair or twinax cable (Ethernet and USB cables come to mind). Make the part that is not twisted or shielded as short as possible (millimeters rather than centimeters), and see what bandwidth you get. Capacitance may be an issue if you make the cable too long, so try to keep it as short as possible.

Offline RGK

  • Contributor
  • Posts: 37
Re: Analog Discovery differential input
« Reply #3 on: July 25, 2017, 04:01:18 am »
 

Offline not1xor1Topic starter

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Analog Discovery differential input
« Reply #4 on: July 26, 2017, 04:46:04 am »
The BNC adapter makes the AD scope operate the same way other scopes operate; one side of the input connector is ground.

In this case, you need to buy a differential probe and that will set you back a bunch of money.  Dave sells one:
https://www.eevblog.com/product/hvp70/

Thanks for your reply, but I already know that the BNC adapter doesn't allow to use the differential input.

Analog discovery already has its own differential amplifier, so why should I buy another one rather than making a custom adapter board?  :-//

The purpose of my post was just to check if there are any problems I'm not aware of.

I would like to make a board with differential inputs for both channels, 1/10 attenuator switches and AC/DC coupling switches.  :-/O
 

Offline not1xor1Topic starter

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Analog Discovery differential input
« Reply #5 on: July 26, 2017, 04:53:43 am »
I think the bandwidth limitation might be mostly about the flying leads from the AD that lack any shielding or defined impedance. Does the adapter board contain any termination resistors? If not, then you may be able to make a cable that connects the differential signals to a twisted pair or twinax cable (Ethernet and USB cables come to mind). Make the part that is not twisted or shielded as short as possible (millimeters rather than centimeters), and see what bandwidth you get. Capacitance may be an issue if you make the cable too long, so try to keep it as short as possible.

Digilent released full data about the board, but I would like to make my own board. I plan to mount it like a panel perpendicular to the Analog Discovery case with differential inputs (2 BNCs per channel) and switches for 1/10 attenuation and DC/AC coupling.
 

Offline not1xor1Topic starter

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: Analog Discovery differential input
« Reply #6 on: July 26, 2017, 04:57:51 am »
How about this?

http://softone.a.la9.jp/english/FrontBox/FrontBox1.htm

BK

Thanks  :clap:  :clap:  :clap:

That is exactly what I meant.
So there is no real problem.
I'll read carefully the linked article and modify the board according to my needs.

Thanks again  :)
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf