T
Tareq Matar
Guest
what is the difference between these configurations on an ADC ?
Navigation
Install the app
How to install the app on iOS
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
More options
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
R
RobertMacy
Guest
On Wed, 10 Sep 2014 07:38:41 -0700, Tareq Matar <tareqmatarr@gmail.com>
wrote:
uh, ...throw away half the capability?
wrote:
what is the difference between these configurations on an ADC ?
uh, ...throw away half the capability?
R
RobertMacy
Guest
On Wed, 10 Sep 2014 07:38:41 -0700, Tareq Matar <tareqmatarr@gmail.com>
wrote:
> what is the difference between these configurations on an ADC ?
First answer was flippant, so....
Balanced has some attributes that can really improve dynamic range and
performance. Plus, many parts support 'balanced' operation. The main one
for me is that a balanced system will almost ALWAYS have a lower noise
floor. Envision two conductors in a shield that each go a mirrored voltage
away from each other. Difficult to inject noise there. Now envision a
receiver that looks at two incoming conductors again with balanced
voltages on them, again difficult to misunderstand the signal.
In other words, the signal going down a blanaced connection is 'self
contained' and does NOT rely on anything else.
Single ended, well. you have now included a 'common' connection that has a
lot of other signals using it!
Does that explain some of the advantage?
Also, in single ended as you try to use that common connection, as you go
up in frequency, parasitic components [all those components you usually
ignore like series inductance and leakage cpacitances] start to rear their
ugly heads and you have little hope of everything working well, so
balanced is the ONLY way to go.
wrote:
> what is the difference between these configurations on an ADC ?
First answer was flippant, so....
Balanced has some attributes that can really improve dynamic range and
performance. Plus, many parts support 'balanced' operation. The main one
for me is that a balanced system will almost ALWAYS have a lower noise
floor. Envision two conductors in a shield that each go a mirrored voltage
away from each other. Difficult to inject noise there. Now envision a
receiver that looks at two incoming conductors again with balanced
voltages on them, again difficult to misunderstand the signal.
In other words, the signal going down a blanaced connection is 'self
contained' and does NOT rely on anything else.
Single ended, well. you have now included a 'common' connection that has a
lot of other signals using it!
Does that explain some of the advantage?
Also, in single ended as you try to use that common connection, as you go
up in frequency, parasitic components [all those components you usually
ignore like series inductance and leakage cpacitances] start to rear their
ugly heads and you have little hope of everything working well, so
balanced is the ONLY way to go.
J
John Larkin
Guest
On Wed, 10 Sep 2014 07:38:41 -0700 (PDT), Tareq Matar
<tareqmatarr@gmail.com> wrote:
>what is the difference between these configurations on an ADC ?
Two input wires versus one!
For slow ADCs, like delta-sigmas, diff can eliminate ground-loop
errors.
Fast pipeline ADCs seem to have less distortion if they are used in
differential-input mode; has something to do with the internal
charge-based digitizer stages.
--
John Larkin Highland Technology, Inc
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
<tareqmatarr@gmail.com> wrote:
>what is the difference between these configurations on an ADC ?
Two input wires versus one!
For slow ADCs, like delta-sigmas, diff can eliminate ground-loop
errors.
Fast pipeline ADCs seem to have less distortion if they are used in
differential-input mode; has something to do with the internal
charge-based digitizer stages.
--
John Larkin Highland Technology, Inc
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
Guest
Thing is, you don't really need it on the ADC. tha tis local, with local commons. Where differential really helps in in the transmission of the signals. Like LVDS. That is how the signal gets to your TV in some cases.
It seems actually easier to achieve a high CMRR with lower levels, so they chose the levels for like HDMI cables and all that because of that. You don't need forty volts of video.
Even wiht LVDS they usually shield the transmission cables becae you CAN swamp theinput amps with noise in common mode, and there is a limit. But when you get to a CMRR of lik 200:1, you're doing pretty good. shiled that nad it is gojng to be pretty much noise free.
One of the reasons for low voltage in a LVDS system also is that the wires have capacitance. Too much voltage swing requires better slew rates in the amps both feeding and recieving the signal. This costs money. So this is a compromise now, use a lower signal level and you don't need to pour as much money in the amps for a given bandwidth. with a lower bandwidth you might choose a higher signal level and be sloppier with the physical wiring, maybe even do without any shieldiong as long as you don't bother the FCC.
It seems actually easier to achieve a high CMRR with lower levels, so they chose the levels for like HDMI cables and all that because of that. You don't need forty volts of video.
Even wiht LVDS they usually shield the transmission cables becae you CAN swamp theinput amps with noise in common mode, and there is a limit. But when you get to a CMRR of lik 200:1, you're doing pretty good. shiled that nad it is gojng to be pretty much noise free.
One of the reasons for low voltage in a LVDS system also is that the wires have capacitance. Too much voltage swing requires better slew rates in the amps both feeding and recieving the signal. This costs money. So this is a compromise now, use a lower signal level and you don't need to pour as much money in the amps for a given bandwidth. with a lower bandwidth you might choose a higher signal level and be sloppier with the physical wiring, maybe even do without any shieldiong as long as you don't bother the FCC.
R
RobertMacy
Guest
On Wed, 10 Sep 2014 19:25:47 -0700, <jurb6006@gmail.com> wrote:
Actually, balanced helps a great deal on the ADC. Input voltage noise
limits the ADC's noise floor. Single eneded means the signal can go from
+rail through GND to -rail. a 2 rail range. Now consider a balanced system
where one side goes from +rail through GND to -rail and the other side
goes opposite from -rail through GND to +rail, signal range of 4 rails!
Thus you gain 6dB.
And sorry I snipped too much, but *if* you do balance correctly, you'll
breeze through FCC Compliance requirements WITHOUT shielding your cables.
Actually run 100Mbs over a carefully made twisted pairof wires, no
shielding. However, single ended can be a strain at high frequencies,
where a shielded coax can 'leak' like a sieve and fail compliance.
Consider single ended stripline in a PCB structure. Even in a six, or 8,
layer PCB using 1/2 oz copper the shielding you get between striplines can
be irritatingly low, [from memory] 106dB down which is usually ok, but not
great if it's important.
Done properly, the same PCB stackup using balanced can get better than
138dB shielding, [again from memory] but I do remember better than 30dB
improvement adjacent and over 60dB improvement in further out layers.
Thing is, you don't really need it on the ADC. tha tis local, with local
commons. Where differential really helps in in the transmission of the
signals. Like LVDS. That is how the signal gets to your TV in some cases.
...snip...
Actually, balanced helps a great deal on the ADC. Input voltage noise
limits the ADC's noise floor. Single eneded means the signal can go from
+rail through GND to -rail. a 2 rail range. Now consider a balanced system
where one side goes from +rail through GND to -rail and the other side
goes opposite from -rail through GND to +rail, signal range of 4 rails!
Thus you gain 6dB.
And sorry I snipped too much, but *if* you do balance correctly, you'll
breeze through FCC Compliance requirements WITHOUT shielding your cables.
Actually run 100Mbs over a carefully made twisted pairof wires, no
shielding. However, single ended can be a strain at high frequencies,
where a shielded coax can 'leak' like a sieve and fail compliance.
Consider single ended stripline in a PCB structure. Even in a six, or 8,
layer PCB using 1/2 oz copper the shielding you get between striplines can
be irritatingly low, [from memory] 106dB down which is usually ok, but not
great if it's important.
Done properly, the same PCB stackup using balanced can get better than
138dB shielding, [again from memory] but I do remember better than 30dB
improvement adjacent and over 60dB improvement in further out layers.
J
John Fields
Guest
On Wed, 10 Sep 2014 07:38:41 -0700 (PDT), Tareq Matar
<tareqmatarr@gmail.com> wrote:
>what is the difference between these configurations on an ADC ?
---
A single-ended input will connect one of the inputs to ground, while
a differential input will let both inputs float and evaluate the
difference between them with respect to a previously established
reference.
<tareqmatarr@gmail.com> wrote:
>what is the difference between these configurations on an ADC ?
---
A single-ended input will connect one of the inputs to ground, while
a differential input will let both inputs float and evaluate the
difference between them with respect to a previously established
reference.