Hi all.
I'm making line receiver circuit using INA1650 and signal relay, which can receive both differential and single-ended line inputs.
This circuit's main goal is converting differential/single-ended line input to single-ended line output, suppressing common mode noise and forms suitable signal to use amplifier input.
Using relays gives ability to the circuit to selecting input source by using 1 INA1650 chip, whatever it's differential or SE(SE signal also treated as differential using 4 signal path).
Making this circuit can be possible to me thanks to Ti's INA165x datasheets, which kindly suggests both differential and SE input scenario at Typical Application(Figure 58, 59, 61).
But I got a problem.
Seeing the Figure 58 and 59, those application input source should only be connected to the chassis ground, which shares same ground plane between every input sources whatever it is selected to use or not.
The Ti's datasheets also describes the Figure 59 as 'Floating Single-Ended Input', that means the chassis ground and signal ground must be isolated. Am I right?
Seems nothing problem here.
But... The main problem is Figure 61.
In Figure 61, in TRS audio interfaces, the input source's sleeve should be connected to chassis ground and also SIGNAL ground. Wait... what?
At first I mentioned, I will gonna mix those above three circuits in one.
Figure 58 and 59 is okay. Those inputs should be connected to only chassis ground and can be isolated from the signal ground.
But if I also following Figure 61 and connecting both chassis ground and signal ground, Isolated chassis and signal ground at Fig. 58 and 59 gonna be useless!
I'm really concerned about sharing chassis ground and signal ground can cause any potential danger, just like electrical shocking or audible noise or harm to any other chip, etc.
So I'm considering about NO connecting signal ground at Figure 61's TRS input source.
If I do not connect signal ground at TRS input, what will happen?
No sound will play when selecting TRS channel? or more huge problem can be happens?
Thanks for reading, and I will hope for any suggestion.
I'm making line receiver circuit using INA1650 and signal relay, which can receive both differential and single-ended line inputs.
This circuit's main goal is converting differential/single-ended line input to single-ended line output, suppressing common mode noise and forms suitable signal to use amplifier input.
Using relays gives ability to the circuit to selecting input source by using 1 INA1650 chip, whatever it's differential or SE(SE signal also treated as differential using 4 signal path).
Making this circuit can be possible to me thanks to Ti's INA165x datasheets, which kindly suggests both differential and SE input scenario at Typical Application(Figure 58, 59, 61).
But I got a problem.
Seeing the Figure 58 and 59, those application input source should only be connected to the chassis ground, which shares same ground plane between every input sources whatever it is selected to use or not.
The Ti's datasheets also describes the Figure 59 as 'Floating Single-Ended Input', that means the chassis ground and signal ground must be isolated. Am I right?
Seems nothing problem here.
But... The main problem is Figure 61.
In Figure 61, in TRS audio interfaces, the input source's sleeve should be connected to chassis ground and also SIGNAL ground. Wait... what?
At first I mentioned, I will gonna mix those above three circuits in one.
Figure 58 and 59 is okay. Those inputs should be connected to only chassis ground and can be isolated from the signal ground.
But if I also following Figure 61 and connecting both chassis ground and signal ground, Isolated chassis and signal ground at Fig. 58 and 59 gonna be useless!
I'm really concerned about sharing chassis ground and signal ground can cause any potential danger, just like electrical shocking or audible noise or harm to any other chip, etc.
So I'm considering about NO connecting signal ground at Figure 61's TRS input source.
If I do not connect signal ground at TRS input, what will happen?
No sound will play when selecting TRS channel? or more huge problem can be happens?
Thanks for reading, and I will hope for any suggestion.
Figure 61 refers to TRS jacks, which are uninsulated.
Any exposed metal parts that the user can touch MUST be connected to chassis earth and mains earth. This is non-negotiable. That is the reason for that connection, it's a legal requirement.
Any exposed metal parts that the user can touch MUST be connected to chassis earth and mains earth. This is non-negotiable. That is the reason for that connection, it's a legal requirement.
Hi, Sangram.
Thanks for your reply.
But if I share signal ground and chassis ground, every line interfaces include differential and RCA terminal's input also be connected to the signal ground.
That means those line inputs should be isolated from signal ground goes no more 'floating'.
Then what effects can be caused by connecting both grounds? Got a noise problems?
Thanks for your reply.
But if I share signal ground and chassis ground, every line interfaces include differential and RCA terminal's input also be connected to the signal ground.
That means those line inputs should be isolated from signal ground goes no more 'floating'.
Then what effects can be caused by connecting both grounds? Got a noise problems?
You may. Or you may not.
But you cannot defeat that connection.
This TRS application is not professional balanced TRS, but more the stereo phone jack type input connector, like the aux single 3.5mm input in a cheap amplifier. The tip and the sleeve are the two channels, hence you can't use the 'ground isolation' method they've shown for RCA (which may actually not be legal in some parts of the world).
But you cannot defeat that connection.
This TRS application is not professional balanced TRS, but more the stereo phone jack type input connector, like the aux single 3.5mm input in a cheap amplifier. The tip and the sleeve are the two channels, hence you can't use the 'ground isolation' method they've shown for RCA (which may actually not be legal in some parts of the world).
Now I got it.
When I make anything electrically conductive and contactable by human, that must be connected to chassis ground and mains earth. Right?
Then what about the module mounted inside of the enclosure that cannot be touched? (just like bluetooth module)
If possible, without connecting differential line input to any ground and use as pure source can be helpful to improving audio quality?
When I make anything electrically conductive and contactable by human, that must be connected to chassis ground and mains earth. Right?
Then what about the module mounted inside of the enclosure that cannot be touched? (just like bluetooth module)
If possible, without connecting differential line input to any ground and use as pure source can be helpful to improving audio quality?
Yes.When I make anything electrically conductive and contactable by human, that must be connected to chassis ground and mains earth. Right?
(just like bluetooth module)
Bluetooth is extremely compressed and lossy. If you consider that to be sufficient for your use case it's OK, but I wouldn't use the word 'Pure' with Bluetooth. Unless it was followed by a slightly different word.
It is single supply so ground / virtual ground can add confusion.
with dual supply be different and to eliminate ground loops. input jack sleeves would not be
connected to chassis. Which creates a huge loop.
But yes signal ground does eventually go to ground. Its where the connection is made.
doesn't matterin this case , likely add more confusion anyways, it is single supply
single supply and the topology of TI design is connecting
input sleeve to chassis. Which leaves it susceptible to noise and a almost guarantee
ground loop.
This design has extra diodes, resistor and caps. shown in figure 61 called loop breaker.
And pretty much does just that. Help break the ground loop formed when input sleeves
go to chassis
with dual supply be different and to eliminate ground loops. input jack sleeves would not be
connected to chassis. Which creates a huge loop.
But yes signal ground does eventually go to ground. Its where the connection is made.
doesn't matterin this case , likely add more confusion anyways, it is single supply
single supply and the topology of TI design is connecting
input sleeve to chassis. Which leaves it susceptible to noise and a almost guarantee
ground loop.
This design has extra diodes, resistor and caps. shown in figure 61 called loop breaker.
And pretty much does just that. Help break the ground loop formed when input sleeves
go to chassis
Hi WhiteDragon. Thanks for your reply.
As I understand your word correctly, you're meaning following lines :
1. Single supply /w virtual ground circuit makes huge ground loop, can almost always create audible noise.
2. BUT in my cases, application circuit that Ti suggested can break ground loop and reduce noise.
3. So I don't have to worry about connecting sig.GND and chas.GND
is it right?
As I understand your word correctly, you're meaning following lines :
1. Single supply /w virtual ground circuit makes huge ground loop, can almost always create audible noise.
2. BUT in my cases, application circuit that Ti suggested can break ground loop and reduce noise.
3. So I don't have to worry about connecting sig.GND and chas.GND
is it right?
- Home
- Source & Line
- Analogue Source
- INA1650 : What happens if I do NOT connect signal ground to chassis ground?