And More
.
Thanks for your comments, guys. I'm familiar with TI's TLE2426 "Rail Splitter" ( http://www.ti.com/litv/slos098d ). The thing is, I don't exactly see the point, at least not for audio. I suppose such precision might be good for digital, or maybe for the golden ear crowd. By the way, if I remember rightly the price has come down lately to under 2 bucks from USA suppliers--plus maybe 7 bucks shipping for an entire order. Some eBay sellers want 4 bucks including shipping.
Others make reference to the circuit below. This is discussed by Ron Eliot, among others, in his article on virtual grounds ( Virtual Ground Circuits ), and I've see tech bulletins on the subject by Burr Brown or somebody, which I can't find right now. I really gotta clean up my files someday.
This is actually the circuit I had in mind originally, although in my question I omitted the IC for the sake of clarity. Around here, I've found, it's unwise to complicate things by addressing more than one issue at a time.
This oddity with this circuit is that connecting to a second device immediately pulls the second device's ground to the +4.5 volts of the virtual ground. This doesn't shut down the entire North American power grid because: 1. The effect is swamped at the building's circuit breaker box, or actually probably at the wall receptacle. 2. The power company knows people do this stuff, and takes measures against it. 3. Before anything else voltage is a potential, not an event. That is, something potentially could happen, but it's not happening right now.
The "something" that could happen is current flow, motivated by that +4.5 volts. But there's no return path to the battery in the circuit, so that potential event never occurs. The voltage just sits there doing nothing, like a wall receptacle with nothing plugged in.
That is, maybe it does nothing. But I can imagine that +4.5 volts, applied to the chassis of the second device, and/or to all its grounds, working mischief just out of sheer boredom. Eddy currents wandering around, grounds not being grounds, and of course the dread 60/120 Hz hum. There seems to be opportunity for lots of that kind of thing. But I dunno.
I wish we shadetree mechanics didn't have to keep re-inventing the wheel. I wish the engineers at TI or someplace would take 45 minutes and come up with something. I wish they'd say, "OK, here's your standard preamp, here's your standard amp, and this is how they connect." But then, If wishes were horses, I'm given to understand, beggars would ride.
.
.
Thanks for your comments, guys. I'm familiar with TI's TLE2426 "Rail Splitter" ( http://www.ti.com/litv/slos098d ). The thing is, I don't exactly see the point, at least not for audio. I suppose such precision might be good for digital, or maybe for the golden ear crowd. By the way, if I remember rightly the price has come down lately to under 2 bucks from USA suppliers--plus maybe 7 bucks shipping for an entire order. Some eBay sellers want 4 bucks including shipping.
Others make reference to the circuit below. This is discussed by Ron Eliot, among others, in his article on virtual grounds ( Virtual Ground Circuits ), and I've see tech bulletins on the subject by Burr Brown or somebody, which I can't find right now. I really gotta clean up my files someday.
This is actually the circuit I had in mind originally, although in my question I omitted the IC for the sake of clarity. Around here, I've found, it's unwise to complicate things by addressing more than one issue at a time.
This oddity with this circuit is that connecting to a second device immediately pulls the second device's ground to the +4.5 volts of the virtual ground. This doesn't shut down the entire North American power grid because: 1. The effect is swamped at the building's circuit breaker box, or actually probably at the wall receptacle. 2. The power company knows people do this stuff, and takes measures against it. 3. Before anything else voltage is a potential, not an event. That is, something potentially could happen, but it's not happening right now.
The "something" that could happen is current flow, motivated by that +4.5 volts. But there's no return path to the battery in the circuit, so that potential event never occurs. The voltage just sits there doing nothing, like a wall receptacle with nothing plugged in.
That is, maybe it does nothing. But I can imagine that +4.5 volts, applied to the chassis of the second device, and/or to all its grounds, working mischief just out of sheer boredom. Eddy currents wandering around, grounds not being grounds, and of course the dread 60/120 Hz hum. There seems to be opportunity for lots of that kind of thing. But I dunno.
I wish we shadetree mechanics didn't have to keep re-inventing the wheel. I wish the engineers at TI or someplace would take 45 minutes and come up with something. I wish they'd say, "OK, here's your standard preamp, here's your standard amp, and this is how they connect." But then, If wishes were horses, I'm given to understand, beggars would ride.
.
Attachments
Last edited:
Funny you happen to ask. Please see post #21.
But NOT an LM386. It's biased exactly backward, its output is at 1/2 supply, not its ground. In this circuit the ground is biased at 1/2 supply.
No, the LM386 is biased the same as your op amp circuit. Are the two not more-or-less interchangeable in your circuit?
If I can find an IC model I'll give a simulation a try, though tbh I'm usually content with a few milliamps through a simple R voltage divider.
Nooo! The LM386 is not an operational amplifier at all, it's a low wattage amplifier made as an 8 pin chip. Apples and oranges.
In the LM386 the output audio signal rides on a DC voltage of 1/2 power supply voltage. This is why an output capacitor is mandatory, to remove the DC component.
In my circuit the output audio signal is centered on zero volts, so no output capacitor is needed. BUT the entire circuit's ground is at 1/2 power supply voltage.
The confusion comes from the world "ground," of course. Everybody considers that ground = ground = ground...
It would have better if I'd never used the word "ground" at all when referring to my power supply circuit. Much better to use the term "common," and then realize that common is at 1/2 power supply voltage.
Changing the subject entirely, I'm posting the LM386 model I use in LTspice. It's not mine, it's by Helmut, originally posted in Yahoo Groups, and I think now posted in Google Groups. But difficult to find in either place since they reorganized, so I'm taking the liberty of posting it here.
How to:
There are two files:
LM386.sub is the circuit, it goes into:
C:\Program Files\LTspice\lib\sub\LM386.sub
LM386.asy is the symbol, it goes into:
C:\Program Files\LTspice\lib\sym\Opamps\LM386.asy
Copy each of the two following posts into each of two text files, then rename the files from .txt to the appropriate .sub or .asy extension.
Drag/copy/move the files into the appropriate folders, and after LTspice is closed and then started again (if it was running) the LM386 will be available. Hit F2 to open the components dialog, and look in the Opamps subfolder.
.
Last edited:
Helmut's LM386.sub
* Model from Dave Dilatush
* lm386 subcircuit model follows:
* IC pins: 2 3 7 1 8 5 6 4
* | | | | | | | |
..subckt lm386 inn inp byp g1 g8 out vs gnd
* input emitter-follower buffers:
q1 gnd inn 10011 ddpnp
r1 inn gnd 50k
q2 gnd inp 10012 ddpnp
r2 inp gnd 50k
* differential input stage, gain-setting
* resistors, and internal feedback resistor:
q3 10013 10011 10008 ddpnp
q4 10014 10012 g1 ddpnp
r3 vs byp 15k
r4 byp 10008 15k
r5 10008 g8 150
r6 g8 g1 1.35k
r7 g1 out 15k
* input stage current mirror:
q5 10013 10013 gnd ddnpn
q6 10014 10013 gnd ddnpn
* voltage gain stage & rolloff cap:
q7 10017 10014 gnd ddnpn
c1 10014 10017 15pf
* current mirror source for gain stage:
i1 10002 vs dc 5m
q8 10004 10002 vs ddpnp
q9 10002 10002 vs ddpnp
* Sziklai-connected push-pull output stage:
q10 10018 10017 out ddpnp
q11 10004 10004 10009 ddnpn 100
q12 10009 10009 10017 ddnpn 100
q13 vs 10004 out ddnpn 100
q14 out 10018 gnd ddnpn 100
* generic transistor models generated
* with MicroSim's PARTs utility, using
* default parameters except Bf:
..model ddnpn NPN(Is=10f Xti=3 Eg=1.11 Vaf=100
+ Bf=400 Ise=0 Ne=1.5 Ikf=0 Nk=.5 Xtb=1.5 Var=100
+ Br=1 Isc=0 Nc=2 Ikr=0 Rc=0 Cjc=2p Mjc=.3333
+ Vjc=.75 Fc=.5 Cje=5p Mje=.3333 Vje=.75 Tr=10n
+ Tf=1n Itf=1 Xtf=0 Vtf=10)
..model ddpnp PNP(Is=10f Xti=3 Eg=1.11 Vaf=100
+ Bf=200 Ise=0 Ne=1.5 Ikf=0 Nk=.5 Xtb=1.5 Var=100
+ Br=1 Isc=0 Nc=2 Ikr=0 Rc=0 Cjc=2p Mjc=.3333
+ Vjc=.75 Fc=.5 Cje=5p Mje=.3333 Vje=.75 Tr=10n
+ Tf=1n Itf=1 Xtf=0 Vtf=10)
..ends
*----------end of subcircuit model-----------
* Model from Dave Dilatush
* lm386 subcircuit model follows:
* IC pins: 2 3 7 1 8 5 6 4
* | | | | | | | |
..subckt lm386 inn inp byp g1 g8 out vs gnd
* input emitter-follower buffers:
q1 gnd inn 10011 ddpnp
r1 inn gnd 50k
q2 gnd inp 10012 ddpnp
r2 inp gnd 50k
* differential input stage, gain-setting
* resistors, and internal feedback resistor:
q3 10013 10011 10008 ddpnp
q4 10014 10012 g1 ddpnp
r3 vs byp 15k
r4 byp 10008 15k
r5 10008 g8 150
r6 g8 g1 1.35k
r7 g1 out 15k
* input stage current mirror:
q5 10013 10013 gnd ddnpn
q6 10014 10013 gnd ddnpn
* voltage gain stage & rolloff cap:
q7 10017 10014 gnd ddnpn
c1 10014 10017 15pf
* current mirror source for gain stage:
i1 10002 vs dc 5m
q8 10004 10002 vs ddpnp
q9 10002 10002 vs ddpnp
* Sziklai-connected push-pull output stage:
q10 10018 10017 out ddpnp
q11 10004 10004 10009 ddnpn 100
q12 10009 10009 10017 ddnpn 100
q13 vs 10004 out ddnpn 100
q14 out 10018 gnd ddnpn 100
* generic transistor models generated
* with MicroSim's PARTs utility, using
* default parameters except Bf:
..model ddnpn NPN(Is=10f Xti=3 Eg=1.11 Vaf=100
+ Bf=400 Ise=0 Ne=1.5 Ikf=0 Nk=.5 Xtb=1.5 Var=100
+ Br=1 Isc=0 Nc=2 Ikr=0 Rc=0 Cjc=2p Mjc=.3333
+ Vjc=.75 Fc=.5 Cje=5p Mje=.3333 Vje=.75 Tr=10n
+ Tf=1n Itf=1 Xtf=0 Vtf=10)
..model ddpnp PNP(Is=10f Xti=3 Eg=1.11 Vaf=100
+ Bf=200 Ise=0 Ne=1.5 Ikf=0 Nk=.5 Xtb=1.5 Var=100
+ Br=1 Isc=0 Nc=2 Ikr=0 Rc=0 Cjc=2p Mjc=.3333
+ Vjc=.75 Fc=.5 Cje=5p Mje=.3333 Vje=.75 Tr=10n
+ Tf=1n Itf=1 Xtf=0 Vtf=10)
..ends
*----------end of subcircuit model-----------
Helmut's LM386.asy
Version 4
SymbolType CELL
LINE Normal -32 0 64 64
LINE Normal -32 128 64 64
LINE Normal -32 0 -32 128
LINE Normal -28 32 -20 32
LINE Normal -28 96 -20 96
LINE Normal -24 100 -24 92
LINE Normal 16 0 16 32
LINE Normal -16 128 -16 117
LINE Normal -21 19 -13 19
LINE Normal -17 15 -17 23
LINE Normal -20 110 -12 110
LINE Normal 48 0 48 53
LINE Normal -16 0 -16 11
LINE Normal 32 85 32 128
LINE Normal -32 32 -48 32
LINE Normal -32 96 -48 96
LINE Normal 64 64 80 64
TEXT -49 19 Left 0 2
TEXT -49 83 Left 0 3
TEXT -9 3 Left 0 6
TEXT -9 124 Left 0 4
TEXT 19 21 Left 0 1
TEXT 52 41 Left 0 8
TEXT 35 99 Left 0 7
TEXT 62 78 Left 0 5
WINDOW 0 64 16 Left 0
WINDOW 3 48 120 Left 0
SYMATTR Value LM386
SYMATTR Prefix X
SYMATTR SpiceModel LM386.sub
SYMATTR Value2 LM386
SYMATTR Description Audio Power Amplifier
PIN -48 32 NONE 0
PINATTR PinName In-
PINATTR SpiceOrder 1
PIN -48 96 NONE 0
PINATTR PinName In+
PINATTR SpiceOrder 2
PIN 32 128 NONE 8
PINATTR PinName B
PINATTR SpiceOrder 3
PIN 16 0 NONE 0
PINATTR PinName COMP2
PINATTR SpiceOrder 4
PIN 48 0 NONE 0
PINATTR PinName V+
PINATTR SpiceOrder 5
PIN 80 64 NONE 0
PINATTR PinName OUT
PINATTR SpiceOrder 6
PIN -16 0 NONE 0
PINATTR PinName COMP1
PINATTR SpiceOrder 7
PIN -16 128 NONE 0
PINATTR PinName V-
PINATTR SpiceOrder 8
Version 4
SymbolType CELL
LINE Normal -32 0 64 64
LINE Normal -32 128 64 64
LINE Normal -32 0 -32 128
LINE Normal -28 32 -20 32
LINE Normal -28 96 -20 96
LINE Normal -24 100 -24 92
LINE Normal 16 0 16 32
LINE Normal -16 128 -16 117
LINE Normal -21 19 -13 19
LINE Normal -17 15 -17 23
LINE Normal -20 110 -12 110
LINE Normal 48 0 48 53
LINE Normal -16 0 -16 11
LINE Normal 32 85 32 128
LINE Normal -32 32 -48 32
LINE Normal -32 96 -48 96
LINE Normal 64 64 80 64
TEXT -49 19 Left 0 2
TEXT -49 83 Left 0 3
TEXT -9 3 Left 0 6
TEXT -9 124 Left 0 4
TEXT 19 21 Left 0 1
TEXT 52 41 Left 0 8
TEXT 35 99 Left 0 7
TEXT 62 78 Left 0 5
WINDOW 0 64 16 Left 0
WINDOW 3 48 120 Left 0
SYMATTR Value LM386
SYMATTR Prefix X
SYMATTR SpiceModel LM386.sub
SYMATTR Value2 LM386
SYMATTR Description Audio Power Amplifier
PIN -48 32 NONE 0
PINATTR PinName In-
PINATTR SpiceOrder 1
PIN -48 96 NONE 0
PINATTR PinName In+
PINATTR SpiceOrder 2
PIN 32 128 NONE 8
PINATTR PinName B
PINATTR SpiceOrder 3
PIN 16 0 NONE 0
PINATTR PinName COMP2
PINATTR SpiceOrder 4
PIN 48 0 NONE 0
PINATTR PinName V+
PINATTR SpiceOrder 5
PIN 80 64 NONE 0
PINATTR PinName OUT
PINATTR SpiceOrder 6
PIN -16 0 NONE 0
PINATTR PinName COMP1
PINATTR SpiceOrder 7
PIN -16 128 NONE 0
PINATTR PinName V-
PINATTR SpiceOrder 8
I won't belabor this.
"1/2 power supply voltage" is what I'm proposing as ground.
And my LM386 is powered from +V & -V. Just like your op amp.
Ground is at pin 5. It's a "rail splitter".
I did find the SPICE model last night, and some NatSemi apps that may have been the original spark for the idea. I'll have some fun with it while waiting for Santa.
Obviously a typo, because the ground pin on an LM386 is pin 4. Beyond that, clearly we're looking at the same situation from different angles. Well, no harm in that, and some good might come of it.
The return path to the battery is through the power supply pins of the opamp. If the circuit is connected to a load that sends current to "ground", then the opamp needs to sink or source that current, and it does that through its own supply pins.
You might find this page useful: Virtual Ground Circuits
Oh no, none whatsoever. Only that's not the point of this particular exercise.
"Then what is the point?" you might ask. Well...actually...at the moment that question is not easy to answer.
.
Nothing is going to change the fact that in this circuit--and only in this particular circuit--what's taken as "ground" is at a potential of +4.5 volts. This circuit is a closed system, a universe unto itself, with a "ground" of its own devising, not related to any other "ground."
This is in opposition to any device with a 3-prong grounded power plug. In any such device "ground" is zero volts. Essentially at earth potential literally, because the third prong of the power plug is ultimately connected to a 10-foot metal rod driven into...the earth.
(Every electrical system, commercial or residential, has such a "grounding rod" driven into the earth somewhere near the power company's meter.)
My ponders had to do with what happens when/if this "+4.5 volt ground" is connected to equipment that has a true earth ground. The ground's the thing.
.
Well might you, because obviously I don't get it. I can only observe that, as far as I can see, grounding pin five of an LM386 is not likely to give a useful result. Or a happy one.
.
If the power source is floating then the power source voltages "float" until the virtual ground matches the ground of the connected equipment.
If the floating power source is a 9V battery and the virtual ground is at half voltage, then that is it, until you connect it to some other equipment.
The floating virtual ground gets referenced to 0 volts relative to "Earth".
The +9V terminal of the battery "appears" to be @ 4.5V and the negative terminal of the battery "appears" to be @ -4.5V
Similarly if ne uses an isolated 6Vac secondary winding to generate the 9Vdc the smoothing capacitor across the rectifier will "appear" to be at +-4.5V relative to "Earth".
This happens with ALL FLOATING supplies.
The isolation is what allows the supply to float to the reference.
If the floating power source is a 9V battery and the virtual ground is at half voltage, then that is it, until you connect it to some other equipment.
The floating virtual ground gets referenced to 0 volts relative to "Earth".
The +9V terminal of the battery "appears" to be @ 4.5V and the negative terminal of the battery "appears" to be @ -4.5V
Similarly if ne uses an isolated 6Vac secondary winding to generate the 9Vdc the smoothing capacitor across the rectifier will "appear" to be at +-4.5V relative to "Earth".
This happens with ALL FLOATING supplies.
The isolation is what allows the supply to float to the reference.
.
Time for some lab work. Spoiler alert: yep AndrewT, nail on the head.
Using a circuit near-identical to the one in Post #3 of this thread (which is the same as the circuit in Post #1, except the power is hooked up frontward and the IC is not smoking), this breadboaded circuit measured 9 volts across the battery, while the floating ground measured +4.5 volts to the battery negative terminal.
Connected first to a Sony BluRay player, then to a computer, and measuring with the Sony/computer's chassis as ground, sure enough the floating ground measured 0 volts, while the rails measure positive and negative 4.5 volts. In both cases this was with audio playing.
Same test conditions, but powered with a two-prong (ungrounded) wall wart, same result.
All of which tells me that the floating ground doesn't care where it floats, it would as soon be at true earth potential as anywhere.
(If there were such a thing as "true earth potential," which there's not, but that question is best referred to Mr. Tesla.)
So thanks for jumping in, AndrewT. You called it.
.
Time for some lab work. Spoiler alert: yep AndrewT, nail on the head.
Using a circuit near-identical to the one in Post #3 of this thread (which is the same as the circuit in Post #1, except the power is hooked up frontward and the IC is not smoking), this breadboaded circuit measured 9 volts across the battery, while the floating ground measured +4.5 volts to the battery negative terminal.
Connected first to a Sony BluRay player, then to a computer, and measuring with the Sony/computer's chassis as ground, sure enough the floating ground measured 0 volts, while the rails measure positive and negative 4.5 volts. In both cases this was with audio playing.
Same test conditions, but powered with a two-prong (ungrounded) wall wart, same result.
All of which tells me that the floating ground doesn't care where it floats, it would as soon be at true earth potential as anywhere.
(If there were such a thing as "true earth potential," which there's not, but that question is best referred to Mr. Tesla.)
So thanks for jumping in, AndrewT. You called it.
.
I can't guess why picbuck was unable to understand the idea about using an LM386. The only drawback might be the 4ma power consumption of the LM386.
A pair of resistors and an electrolytic cap will consume less and still provide a 4.5 volt 'ground' point for the circuit.
A pair of resistors and an electrolytic cap will consume less and still provide a 4.5 volt 'ground' point for the circuit.
We are all using an isolated mains transformer to power out mains projects.
The "isolated" part is what protects us from killing ourselves.
If we are at "earth" potential and grab hold of one end of a secondary, that secondary "floats" to match our potential.
All isolated PSUs do this. Not just batteries.
The "isolated" part is what protects us from killing ourselves.
If we are at "earth" potential and grab hold of one end of a secondary, that secondary "floats" to match our potential.
All isolated PSUs do this. Not just batteries.
- Status
- Not open for further replies.