## Is a TLE2426 enough to drive 9 opamps in this scenario

Expand Messages
• In the process of building something. I have an OPA134 in unity gain feeding the signal on to 4 x OPA2134 also in unity gain. Both amps in the 2134 are
Message 1 of 10 , Dec 16, 2012
In the process of building something. I have an OPA134 in unity gain feeding
the signal on to 4 x OPA2134 also in unity gain. Both amps in the 2134 are
utilised one for an in phase output and another for an inverted phase
output. It is a distribution buffer for guitar or 4dBU line level sources.

I see the quiescent current for these amps is 4mA each so I expect quiescent
draw to be in the region of 36mA. Regarding signal current draw the OPA2134
I presume will draw extremely minimal current from the input OPA134. The
outputs of all the other 8 opamps are feeding the signal out through a DC
decoupling 10uF NP connected to a 1K resistor and The highest load I expect
is when then outputs are muted when I will tie the output side of the
resistor to virtual ground

I intended running the opamps from a +9/0/-9V split rail supply generated by
a TLE2426. I see from the datasheet that it can sink/source 20mA.

I am thinking that if the input signal is say 4 volts peak to peak and gain
is unity everywhere the heaviest load is when the output is muted to virtual
ground via the 1K resistor. I am thinking this should result in a peak
current of 2/1000 =2mA, so I am not expecting the virtual ground output to
have to supply more than 10mA.

enough capacity. I've seen several posts that include the quiescent drive
current in the TLE2426 load calculation but I was under the impression that
that current is sourced directly from the power rails and has nothing to do
with the virtual ground current. Confused as to how to calculate the load to
determine whether the TLE2426 is up to the job without buffering or not. Can
anyone shed some light for me please?

Thanks,
John

[Non-text portions of this message have been removed]
• Instead of worrying about whether a TLE2426 has adequate capability to supply +/- current for your design you may want to use one of the circuits in the
Message 2 of 10 , Dec 16, 2012
supply +/- current for your design you may want to use one of the
circuits in the following pdf for obtaining more current.
<http://tangentsoft.net/elec/vgrounds.html>

The other Howard

On 12/16/2012 3:06 PM, Pict wrote:
>
> In the process of building something. I have an OPA134 in unity gain
> feeding
> the signal on to 4 x OPA2134 also in unity gain. Both amps in the 2134 are
> utilised one for an in phase output and another for an inverted phase
> output. It is a distribution buffer for guitar or 4dBU line level sources.
>
> I see the quiescent current for these amps is 4mA each so I expect
> quiescent
> draw to be in the region of 36mA. Regarding signal current draw the
> OPA2134
> I presume will draw extremely minimal current from the input OPA134. The
> outputs of all the other 8 opamps are feeding the signal out through a DC
> decoupling 10uF NP connected to a 1K resistor and The highest load I
> expect
> is when then outputs are muted when I will tie the output side of the
> resistor to virtual ground
>
> I intended running the opamps from a +9/0/-9V split rail supply
> generated by
> a TLE2426. I see from the datasheet that it can sink/source 20mA.
>
> I am thinking that if the input signal is say 4 volts peak to peak and
> gain
> is unity everywhere the heaviest load is when the output is muted to
> virtual
> ground via the 1K resistor. I am thinking this should result in a peak
> current of 2/1000 =2mA, so I am not expecting the virtual ground output to
> have to supply more than 10mA.
>
> enough capacity. I've seen several posts that include the quiescent drive
> current in the TLE2426 load calculation but I was under the impression
> that
> that current is sourced directly from the power rails and has nothing
> to do
> with the virtual ground current. Confused as to how to calculate the
> determine whether the TLE2426 is up to the job without buffering or
> not. Can
> anyone shed some light for me please?
>
> Thanks,
> John
>
> [Non-text portions of this message have been removed]
>
>

[Non-text portions of this message have been removed]
• I am familiar with this page thanks, but don t have a suitable buffer opamp or transistors to hand. I can order them if need be but I d rather not if I don t
Message 3 of 10 , Dec 16, 2012
I am familiar with this page thanks, but don't have a suitable buffer opamp
or transistors to hand. I can order them if need be but I'd rather not if I
don't need to. I am not familiar with how to calculate max current required
for the supply. Is it only the current in the virtual ground that is limited
by the TLE2426 20mA limit on source/sink or do I have to factor in the rail
current as well? If it's only the virtual ground current was my guess of
around 2mA per opamp reasonable given the stated output loads and input
level?

Regards,
John

From: Howard Hansen <hrhan@...>
Date: Sun, 16 Dec 2012 17:46:51 -0600
To: <Electronics_101@yahoogroups.com>
Subject: Re: [Electronics_101] Is a TLE2426 enough to drive 9 opamps in
this scenario

supply +/- current for your design you may want to use one of the
circuits in the following pdf for obtaining more current.
<http://tangentsoft.net/elec/vgrounds.html>

The other Howard

On 12/16/2012 3:06 PM, Pict wrote:
>
> In the process of building something. I have an OPA134 in unity gain
> feeding
> the signal on to 4 x OPA2134 also in unity gain. Both amps in the 2134 are
> utilised one for an in phase output and another for an inverted phase
> output. It is a distribution buffer for guitar or 4dBU line level sources.
>
> I see the quiescent current for these amps is 4mA each so I expect
> quiescent
> draw to be in the region of 36mA. Regarding signal current draw the
> OPA2134
> I presume will draw extremely minimal current from the input OPA134. The
> outputs of all the other 8 opamps are feeding the signal out through a DC
> decoupling 10uF NP connected to a 1K resistor and The highest load I
> expect
> is when then outputs are muted when I will tie the output side of the
> resistor to virtual ground
>
> I intended running the opamps from a +9/0/-9V split rail supply
> generated by
> a TLE2426. I see from the datasheet that it can sink/source 20mA.
>
> I am thinking that if the input signal is say 4 volts peak to peak and
> gain
> is unity everywhere the heaviest load is when the output is muted to
> virtual
> ground via the 1K resistor. I am thinking this should result in a peak
> current of 2/1000 =2mA, so I am not expecting the virtual ground output to
> have to supply more than 10mA.
>
> enough capacity. I've seen several posts that include the quiescent drive
> current in the TLE2426 load calculation but I was under the impression
> that
> that current is sourced directly from the power rails and has nothing
> to do
> with the virtual ground current. Confused as to how to calculate the
> determine whether the TLE2426 is up to the job without buffering or
> not. Can
> anyone shed some light for me please?
>
> Thanks,
> John
>
> [Non-text portions of this message have been removed]
>
>

[Non-text portions of this message have been removed]

[Non-text portions of this message have been removed]
• Your second paragraph appears to be missing some text: The outputs of all the other 8 opamps are feeding the signal out through a DC decoupling 10uF NP
Message 4 of 10 , Dec 16, 2012
Your second paragraph appears to be missing some text:

"The outputs of all the other 8 opamps are feeding the signal out through a DC
decoupling 10uF NP connected to a 1K resistor and The highest load I expect
is when then outputs are muted when I will tie the output side of the
resistor to virtual ground"

...so I don't fully understand your circuit. I especially don't
understand what you are saying about muting. Why do you need to tie an
_output_ to ground?

Am I correct in assuming that the OUT pin of the tle2426 is tied to
the circuit's "ground"? And your power supply is floating (no internal
connection to ground) with 18V across its two output wires? (For
example it might be a couple 9V batteries in series, or a "wall wart"
power supply. BTW, beware that some wall warts can introduce
significant noise into your system - but that's another topic.) If

If all your op amps are driving high impedance loads, then the tle2426
doesn't have to supply any significant current at all.

The quiescent supply current of the op amps is irrelevant to this
discussion because the current flows out the + terminal of the power
supply, into the V+ terminals of the op amps, out the V- terminals,
and into the - terminal of the power supply. None of that current goes
in or out of the tle2426's OUT pin.

The resistors that create the inverting buffers also don't create
create a load for the tle2426. Trace the current paths and you'll see.
But make sure the input op amp can supply enough current to those
resistors. The "typical" output drive of the opa134 is +-35 mA. Who
knows what typical means. I'd keep it under 20 mA to be safe and avoid
possible distortion.

Some situations that _would_ cause a large current in or out of the tle2426 are:

- A shorted output. An opa134's "typical" output short circuit current
is +-40 mA. Assuming the short is to ground, the current would return
through the tle2426's OUT pin and exceed its current drive capacity.

- Is the typical load on each output 600 ohms? If each of the 4
positive outputs are loaded with 600 ohms, and the negative outputs
are unloaded, then the outputs are driving a total load of 600/4=150
ohms. An input that reached 3V at some instant would then cause 20 mA
total to flow into the loads and return through the tle2426's OUT pin.

Hope that helps...

- Jan
• From: Jan Kok Reply-To: Date: Sun, 16 Dec 2012 23:34:55 -0700 To:
Message 5 of 10 , Dec 17, 2012
From: Jan Kok <jan.kok.5y@...>
Date: Sun, 16 Dec 2012 23:34:55 -0700
To: <Electronics_101@yahoogroups.com>
Subject: Re: [Electronics_101] Is a TLE2426 enough to drive 9 opamps in
this scenario

<snip>
"...so I don't fully understand your circuit. I especially don't
understand what you are saying about muting. Why do you need to tie an
_output_ to ground?"
If I mute an output I wanted to have the output tied to ground because the
output is still connected to an amp via screened cable TS or TRS lead so I
want to make sure I silence the amp by tying its input to signal ground so
it isn't floating with a length of lead attached to it. Perhaps it would be
better to have the mute switch (spdt) disconnect the guitar amp input from
the OPA2134 output, and switch to connect the amp input to signal ground
while leaving the OPA2134 output floating. The switch is break before make
however so I thought it better to short the amp input to ground (connected
> grounded) rather than switch to ground (connected>floating>grounded).
There is 1K between the mute switch and the OPA2134 output which I don't
think should load down the typical amp input impedance significantly.

"Am I correct in assuming that the OUT pin of the tle2426 is tied to
the circuit's "ground"? And your power supply is floating (no internal
connection to ground) with 18V across its two output wires? (For
example it might be a couple 9V batteries in series, or a "wall wart"
power supply. BTW, beware that some wall warts can introduces
significant noise into your system - but that's another topic.) If

That's right. Power supply is floating ­ 2 9V batteries in series. TLE2426
output is signal ground.

"If all your op amps are driving high impedance loads, then the tle2426
doesn't have to supply any significant current at all."

All loads when outputs are active will be high impedance >20K to 1M

"The quiescent supply current of the op amps is irrelevant to this
discussion because the current flows out the + terminal of the power
supply, into the V+ terminals of the op amps, out the V- terminals,
and into the - terminal of the power supply. None of that current goes
in or out of the tle2426's OUT pin."

This is what I thought. Thanks for clarifying,

"The resistors that create the inverting buffers also don't create
create a load for the tle2426. Trace the current paths and you'll see.
But make sure the input op amp can supply enough current to those
resistors. The "typical" output drive of the opa134 is +-35 mA. Who
knows what typical means. I'd keep it under 20 mA to be safe and avoid
possible distortion."

The input opamp load will be very minimal I think. 10K resistor network in
the inverting buffers.

"Some situations that _would_ cause a large current in or out of the tle2426
are:

- A shorted output. An opa134's "typical" output short circuit current
is +-40 mA. Assuming the short is to ground, the current would return
through the tle2426's OUT pin and exceed its current drive capacity.

- Is the typical load on each output 600 ohms? If each of the 4
positive outputs are loaded with 600 ohms, and the negative outputs
are unloaded, then the outputs are driving a total load of 600/4=150
ohms. An input that reached 3V at some instant would then cause 20 mA
total to flow into the loads and return through the tle2426's OUT pin.""
The maximum load occurs when the opamps are muted effectively tying the
outputs to ground via 1K. The biggest threat to increased current draw seems
to be underestimating the input voltage. Re-reading the TLE2426 datasheet I
see the source/sink value of +-20mA is labelled 'Typical' (whatever that
means). Absolute Maximum is listed at +-80mA so I hope that should give
enough headroom to cope with transients without the virtual ground being
pulled off centre and/or meltdown.

"Hope that helps...
- Jan"

Yes I appreciate the pointers Jan. I have the confidence now to go ahead and
complete the build and see if there are any problems in use. I think it
should be OK but we'll see. Many thanks,
John

[Non-text portions of this message have been removed]
• I agree, one should never tie the output of an opamp to ground. Best is to use a MOSFET or BJT on the input and switch the transistor on to the ground node.
Message 6 of 10 , Dec 17, 2012
I agree, one should never tie the output of an opamp to ground. Best is
to use a MOSFET or BJT on the input and switch the transistor on to the
ground node. This would avoid the "pop" of a break-before-make circuit.

To put the original question simply, quiescent current is the current
drawn from the power supply rails to bias the internal circuitry of the
opamp. Otherwise, other currents involved are:

1. Output current drawn from the supply rails and is based on the load
and feedback current.
2. Current to the input nodes is based on the drive circuit and
resistance to the virtual node.

Derek Koonce
DDK Interactive Consulting Services

On 12/17/2012 11:05 AM, Pict wrote:
>
>
>
> From: Jan Kok <jan.kok.5y@... <mailto:jan.kok.5y%40gmail.com>>
> <mailto:Electronics_101%40yahoogroups.com>>
> Date: Sun, 16 Dec 2012 23:34:55 -0700
> To: <Electronics_101@yahoogroups.com
> <mailto:Electronics_101%40yahoogroups.com>>
> Subject: Re: [Electronics_101] Is a TLE2426 enough to drive 9 opamps in
> this scenario
>
> <snip>
> "...so I don't fully understand your circuit. I especially don't
> understand what you are saying about muting. Why do you need to tie an
> _output_ to ground?"
> If I mute an output I wanted to have the output tied to ground because the
> output is still connected to an amp via screened cable TS or TRS lead so I
> want to make sure I silence the amp by tying its input to signal ground so
> it isn't floating with a length of lead attached to it. Perhaps it
> would be
> better to have the mute switch (spdt) disconnect the guitar amp input from
> the OPA2134 output, and switch to connect the amp input to signal ground
> while leaving the OPA2134 output floating. The switch is break before make
> however so I thought it better to short the amp input to ground (connected
> > grounded) rather than switch to ground (connected>floating>grounded).
> There is 1K between the mute switch and the OPA2134 output which I don't
> think should load down the typical amp input impedance significantly.
>
> "Am I correct in assuming that the OUT pin of the tle2426 is tied to
> the circuit's "ground"? And your power supply is floating (no internal
> connection to ground) with 18V across its two output wires? (For
> example it might be a couple 9V batteries in series, or a "wall wart"
> power supply. BTW, beware that some wall warts can introduces
> significant noise into your system - but that's another topic.) If
>
> That's right. Power supply is floating ­ 2 9V batteries in series. TLE2426
> output is signal ground.
>
> "If all your op amps are driving high impedance loads, then the tle2426
> doesn't have to supply any significant current at all."
>
> All loads when outputs are active will be high impedance >20K to 1M
>
> "The quiescent supply current of the op amps is irrelevant to this
> discussion because the current flows out the + terminal of the power
> supply, into the V+ terminals of the op amps, out the V- terminals,
> and into the - terminal of the power supply. None of that current goes
> in or out of the tle2426's OUT pin."
>
> This is what I thought. Thanks for clarifying,
>
> "The resistors that create the inverting buffers also don't create
> create a load for the tle2426. Trace the current paths and you'll see.
> But make sure the input op amp can supply enough current to those
> resistors. The "typical" output drive of the opa134 is +-35 mA. Who
> knows what typical means. I'd keep it under 20 mA to be safe and avoid
> possible distortion."
>
> The input opamp load will be very minimal I think. 10K resistor network in
> the inverting buffers.
>
> "Some situations that _would_ cause a large current in or out of the
> tle2426
> are:
>
> - A shorted output. An opa134's "typical" output short circuit current
> is +-40 mA. Assuming the short is to ground, the current would return
> through the tle2426's OUT pin and exceed its current drive capacity.
>
> - Is the typical load on each output 600 ohms? If each of the 4
> positive outputs are loaded with 600 ohms, and the negative outputs
> are unloaded, then the outputs are driving a total load of 600/4=150
> ohms. An input that reached 3V at some instant would then cause 20 mA
> total to flow into the loads and return through the tle2426's OUT pin.""
> The maximum load occurs when the opamps are muted effectively tying the
> outputs to ground via 1K. The biggest threat to increased current draw
> seems
> to be underestimating the input voltage. Re-reading the TLE2426
> datasheet I
> see the source/sink value of +-20mA is labelled 'Typical' (whatever that
> means). Absolute Maximum is listed at +-80mA so I hope that should give
> enough headroom to cope with transients without the virtual ground being
> pulled off centre and/or meltdown.
>
> "Hope that helps...
> - Jan"
>
> Yes I appreciate the pointers Jan. I have the confidence now to go
> complete the build and see if there are any problems in use. I think it
> should be OK but we'll see. Many thanks,
> John
>
> [Non-text portions of this message have been removed]
>
>
• ... But as it is the output of the input buffer opamp directly feeds all four output buffers (the input buffer output is directly connected to the noninverting
Message 7 of 10 , Dec 17, 2012
On 17/12/2012 13:49, "Derek" <derek@...> wrote:

>I agree, one should never tie the output of an opamp to ground. Best is
>to use a MOSFET or BJT on the input and switch the transistor on to the
>ground node. This would avoid the "pop" of a break-before-make circuit.

But as it is the output of the input buffer opamp directly feeds all four
output buffers (the input buffer output is directly connected to the
noninverting input on all four output buffers). If I short the input of
one output buffer with a MOSFET I'll short them all to ground.

Regards,
John
• All four output amps should have some resistance isolation between them. If there is none, then there is a potential issue at that point. Or go back one more
Message 8 of 10 , Dec 17, 2012
All four output amps should have some resistance isolation between them.
If there is none, then there is a potential issue at that point. Or go
back one more stage and short the input of the input amp - this should
be isolated as well.

I may have missed something earlier. Did you post a schematic? Is this a
design you are modifying or in process? If you have to modify, then
think about a small daughter card with some type of buffer amp in place
so that the shorting of an input can be easily implemented.

Derek Koonce
DDK Interactive Consulting Services

On 12/17/2012 12:59 PM, Pict wrote:
>
>
>
> On 17/12/2012 13:49, "Derek" <derek@...
> <mailto:derek%40dkoonce.com>> wrote:
>
> >I agree, one should never tie the output of an opamp to ground. Best is
> >to use a MOSFET or BJT on the input and switch the transistor on to the
> >ground node. This would avoid the "pop" of a break-before-make circuit.
>
> But as it is the output of the input buffer opamp directly feeds all four
> output buffers (the input buffer output is directly connected to the
> noninverting input on all four output buffers). If I short the input of
> one output buffer with a MOSFET I'll short them all to ground.
>
> Regards,
> John
>
>

[Non-text portions of this message have been removed]
• I m out of time at the moment to work more on this but will post a schematic when I get a chance to devote some attention to it again (work taking over life
Message 9 of 10 , Dec 17, 2012
I'm out of time at the moment to work more on this but will post a schematic
when I get a chance to devote some attention to it again (work taking over
life again).

Regards,
John

From: Derek <derek@...>
Date: Mon, 17 Dec 2012 14:16:21 -0800
To: <Electronics_101@yahoogroups.com>
Subject: Re: [Electronics_101] Is a TLE2426 enough to drive 9 opamps in
this scenario

All four output amps should have some resistance isolation between them.
If there is none, then there is a potential issue at that point. Or go
back one more stage and short the input of the input amp - this should
be isolated as well.

I may have missed something earlier. Did you post a schematic? Is this a
design you are modifying or in process? If you have to modify, then
think about a small daughter card with some type of buffer amp in place
so that the shorting of an input can be easily implemented.

Derek Koonce
DDK Interactive Consulting Services

On 12/17/2012 12:59 PM, Pict wrote:
>
>
>
> On 17/12/2012 13:49, "Derek" <derek@... <mailto:derek%40dkoonce.com>
> <mailto:derek%40dkoonce.com>> wrote:
>
> >I agree, one should never tie the output of an opamp to ground. Best is
> >to use a MOSFET or BJT on the input and switch the transistor on to the
> >ground node. This would avoid the "pop" of a break-before-make circuit.
>
> But as it is the output of the input buffer opamp directly feeds all four
> output buffers (the input buffer output is directly connected to the
> noninverting input on all four output buffers). If I short the input of
> one output buffer with a MOSFET I'll short them all to ground.
>
> Regards,
> John
>
>

[Non-text portions of this message have been removed]

[Non-text portions of this message have been removed]
• ... So it sounds like there are 4 output sections, each section producing an in-phase and an inverted copy of the input. And it sounds like you want to mute
Message 10 of 10 , Dec 17, 2012
> <snip>
> "...so I don't fully understand your circuit. I especially don't
> understand what you are saying about muting. Why do you need to tie an
> _output_ to ground?"
> If I mute an output I wanted to have the output tied to ground because the
> output is still connected to an amp via screened cable TS or TRS lead so I
> want to make sure I silence the amp by tying its input to signal ground so
> it isn't floating with a length of lead attached to it. Perhaps it would be
> better to have the mute switch (spdt) disconnect the guitar amp input from
> the OPA2134 output, and switch to connect the amp input to signal ground
> while leaving the OPA2134 output floating. The switch is break before make
> however so I thought it better to short the amp input to ground (connected
>> grounded) rather than switch to ground (connected>floating>grounded).
> There is 1K between the mute switch and the OPA2134 output which I don't
> think should load down the typical amp input impedance significantly.

So it sounds like there are 4 output sections, each section producing
an in-phase and an inverted copy of the input. And it sounds like you
want to mute each of the 4 sections independently, i.e. you want 4
mute switches, one for each output.

First, I wonder why you need mute switches for the outputs, since
there is usually a mute switch or at least a volume control on each
input to most mixers, amplifiers, etc. So mute switches on the outputs
seems redundant and might sometimes cause confusion: "Why aren't we
getting signal from... Oh, the output of the distribution amp was
muted."

But anyway, if that's what you want to do, I'd suggest having the
output of the opa134 go to 4 SPST mute switches. The other side of
each mute switch would go to the + input of the non-inverting buffer
and to the 10K resistor that goes to the - input of the inverting
buffer. When the switch is open, there is no current through the input
10K resistor, and the inverting buffer acts to bring its - input to
ground. The + input of the noninverting buffer is tied to virtual
ground through the input 10K resistor. So you just need one SPST
switch per section.

By the way, the clicks and pops you get from switching high impedance
inputs is due to those inputs getting charged up from stray leakage
currents. When the switch is transitions from open to closed, there is
a large, abrupt voltage step as the charge is discharged through the
switch into the low impedance signal source. A simple fix is to tie
the high impedance input to ground through a large (say 100K or 1Meg)
resistor. When the switch is open, the resistor drains off stray
leakage current and keeps the input at ground. With the mute switch
circuit described in the previous paragraph, a separate drain resistor
is not needed because the 10K input resistor and the 10K feedback
resistors serve that function.

Here's another reason why trying to ground the output of an op amp
doesn't make much sense and probably won't work well if at all. Note
that the closed-loop output impedance of the opa134 is 0.01 ohm! In a
struggle between 0.01 ohm and 1K ohm, which do you think will win? :-)

Cheers,
- Jan
Your message has been successfully submitted and would be delivered to recipients shortly.