I need help in understanding the attached schematic. It is a balanced to unbalanced converter. Can someone explain to me the signal path and the overall gain of the circuit. I don't understand what the purpose of R1 (20k) is. What is the gain of IC1a and IC1b respectively?
Thanx in advance.
Thanx in advance.
Attachments
It's a standard instrumentation amplifier https://en.wikipedia.org/wiki/Instrumentation_amplifier R1 sets the gain.
Yep, it's a standard instrumentation amplifier circuit. It converts a balanced interconnect signal to an unbalanced internal circuit. Many mic pre-amps have input circuits similar to this, but they also have phantom power and gain adjustment to deal with.
This drawing does have a error.
It shows XLR pin1 connected to the circuit common (aka ground). Pin 1 should only be connected to the chassis at the connector.
This drawing does have a error.
It shows XLR pin1 connected to the circuit common (aka ground). Pin 1 should only be connected to the chassis at the connector.
and the ground at LK3 should only see the ground terminal of the output jack (which needs to be isolated from the case) otherwise LK3 won't be the gnd lift it's supposed to be!
Or not. With out the ground the common mode voltage can be too much for the circuit. It's not a true isolated balanced input like a transformer with 1000volt common mode isolation.
If you connect pin 1 to chassis "only", then how does P48 work? Isn't "chassis ground" also connected to "signal ground"? And, why the need for two grounds?
I'm being a jerk, but still, chassis can't float relative to signal ground, so what you're probably talking about is how one would route "ground" in the context of a PCB and chassis. Still, Bill Whitlock has shown us that it's very important for pin 1 to be firmly connected to chassis so that RFI and fault currents can be routed through the chassis, and not various snaky PCB traces, but does not say that this chassis connection needs to be isolated from "signal ground".
On the contrary, I find that the chassis pin 1 connection is a great way to bring "ground" into a circuit board. The chassis ground is probably established by a good bit of metal, and already has all of the RF and junk on it - might as well make the RF and junk equal to ground so that it disappears… remember, all voltages are relative! ;-)
It's generally implied that chassis and signal ground connect somewhere.
This schematic apparently splits "ground" into power ground and signal ground, with power ground and chassis apparently used intercheangeably.
BTW, LK3 is indeed connected in the totally wrong place as mentioned (it makes precious little sense to short out V-, a ground lift seems more likely). It also is unclear whether the unbalanced output is wired up correctly. IC2a is referenced to signal ground (read Bruno Putzeys' "The G-Word" whitepaper), while the 100k drain resistor at the output goes to power ground.
BTW, the function of the circuit (IC1a/b) is best explained when assuming an out-of-phase balanced input signal (pure differential mode) and the middle of R1 grounded. In that case you see two standard noninverting amplifiers of gain (1+ 10k/10k) = 2.
In practice the middle of R1 is not grounded and will freely follow a common-mode signal component. For this, the circuit is just a follower.
So differential-mode gain is 2 and common-mode gain is unity.
I wish I could explain circuit simplification for differential mode and common mode better. I am sure that there are some tricks for CM as well.
This schematic apparently splits "ground" into power ground and signal ground, with power ground and chassis apparently used intercheangeably.
BTW, LK3 is indeed connected in the totally wrong place as mentioned (it makes precious little sense to short out V-, a ground lift seems more likely). It also is unclear whether the unbalanced output is wired up correctly. IC2a is referenced to signal ground (read Bruno Putzeys' "The G-Word" whitepaper), while the 100k drain resistor at the output goes to power ground.
BTW, the function of the circuit (IC1a/b) is best explained when assuming an out-of-phase balanced input signal (pure differential mode) and the middle of R1 grounded. In that case you see two standard noninverting amplifiers of gain (1+ 10k/10k) = 2.
In practice the middle of R1 is not grounded and will freely follow a common-mode signal component. For this, the circuit is just a follower.
So differential-mode gain is 2 and common-mode gain is unity.
I wish I could explain circuit simplification for differential mode and common mode better. I am sure that there are some tricks for CM as well.
I agree with Speedskater.
Connecting XLR pin 1 directly to signal ground is a well known inter-connection faux pas. The problem this creates is that it establishes a path for noise current to flow between the signal grounds of devices connected via the balanced cable shield conductor. Connecting pin 1 instead to chassis/safety ground maintains a low voltage differential between the chassis of connected devices without introducing a signal ground loop.
Yes, at some point in each device, signal ground should be tied to chassis/safety ground so to limit common-mode differential voltage between a circuit and it's chassis, as well as across inter-connected circuits, however, there are a number of ways to do that which also minimize the affect of signal ground contamination via inter-chassis signal ground loop noise.
Connecting XLR pin 1 directly to signal ground is a well known inter-connection faux pas. The problem this creates is that it establishes a path for noise current to flow between the signal grounds of devices connected via the balanced cable shield conductor. Connecting pin 1 instead to chassis/safety ground maintains a low voltage differential between the chassis of connected devices without introducing a signal ground loop.
Yes, at some point in each device, signal ground should be tied to chassis/safety ground so to limit common-mode differential voltage between a circuit and it's chassis, as well as across inter-connected circuits, however, there are a number of ways to do that which also minimize the affect of signal ground contamination via inter-chassis signal ground loop noise.
Last edited:
The easy bit about which no one should ever argue:
Pin1 goes direct to chassis at the socket.
This needs a Chassis symbol.
End of easy bit.
The more difficult bit:
The other two "grounds" are Signal and Power.
The 5 bar of equal length are connected to the 10k. These are signal ground and need to be connected together.
The local decoupling from the 100nF are connected to the 3bar unequal length. These are power grounds and must connect back to Power Ground/Zero Volts.
I think LK3 is there to convert to dual polarity supply or to single polarity supply. In that respect it is correctly labeled between V- and power ground. Note the whole circuit is AC coupled at INput and OUTput.
If it's intended to be a form of optional ground lift, then it is shown incorrectly.
The 100pF are interference attenuation. These should go to Chassis, not power ground.
The 100k at the input are signal. These should go to signal ground, not to Pin1 and not to power ground.
The 100k at the output gets it's current from the supply rails. It has to return it's current to the supply rails. On that basis I think the 100k has to go to power ground. It uses the correct power ground symbol.
Summarising
the two Signal grounds (on the 10k) are correct.
The four Power grounds to the right of IC1 are correct.
Pin1 is wrong
The two power grounds to the left of IC1 are wrong. The 100pF go to Chassis. And would be better located on the Input socket, or next to the 100k input resistors.
The 100k go to signal ground.
Pin1 goes direct to chassis at the socket.
This needs a Chassis symbol.
End of easy bit.
The more difficult bit:
The other two "grounds" are Signal and Power.
The 5 bar of equal length are connected to the 10k. These are signal ground and need to be connected together.
The local decoupling from the 100nF are connected to the 3bar unequal length. These are power grounds and must connect back to Power Ground/Zero Volts.
I think LK3 is there to convert to dual polarity supply or to single polarity supply. In that respect it is correctly labeled between V- and power ground. Note the whole circuit is AC coupled at INput and OUTput.
If it's intended to be a form of optional ground lift, then it is shown incorrectly.
The 100pF are interference attenuation. These should go to Chassis, not power ground.
The 100k at the input are signal. These should go to signal ground, not to Pin1 and not to power ground.
The 100k at the output gets it's current from the supply rails. It has to return it's current to the supply rails. On that basis I think the 100k has to go to power ground. It uses the correct power ground symbol.
Summarising
the two Signal grounds (on the 10k) are correct.
The four Power grounds to the right of IC1 are correct.
Pin1 is wrong
The two power grounds to the left of IC1 are wrong. The 100pF go to Chassis. And would be better located on the Input socket, or next to the 100k input resistors.
The 100k go to signal ground.
Last edited:
- Status
- Not open for further replies.