596uA input bias current on LM6132?

[[email protected]]

what the heck. i have an LM6132BIM. the second opamp's inputs and
output are left floating. the first opamp is used. its output is
connected to pin 2, its inverting input, with 200k. its inverting input
is also connected to ground via 33k. the positive input (pin 3) is
connected to about 3.5V from a bench supply. in series with the 3.5V is
my DMM set to uA. it measures 596uA!! How is this possible? There is no
other current path except the one from the bench supply to the positive
input and the voltage does not exceed the opamps supply rails. The
datasheet says this opamp has a 110nA nominal input bias current. This
is causing havoc in my circuit because a voltage divider connected to
the positive input has its voltage dragged down by the huge bias
current. WHAT THE HECK IS GOING ON

 

[martin griffith]

On 9 Nov 2006 15:43:45 -0800, in sci.electronics.design

what the heck. i have an LM6132BIM. the second opamp's inputs and
output are left floating. the first opamp is used. its output is
connected to pin 2, its inverting input, with 200k. its inverting input
is also connected to ground via 33k. the positive input (pin 3) is
connected to about 3.5V from a bench supply. in series with the 3.5V is
my DMM set to uA. it measures 596uA!! How is this possible? There is no
other current path except the one from the bench supply to the positive
input and the voltage does not exceed the opamps supply rails. The
datasheet says this opamp has a 110nA nominal input bias current. This
is causing havoc in my circuit because a voltage divider connected to
the positive input has its voltage dragged down by the huge bias
current.

WHAT THE HECK IS GOING ON

You aren't providing enough white space in your message for those with
less than perfect eyesight


martin

 

[Jim Thompson]

what the heck. i have an LM6132BIM. the second opamp's inputs and
output are left floating. the first opamp is used. its output is
connected to pin 2, its inverting input, with 200k. its inverting input
is also connected to ground via 33k. the positive input (pin 3) is
connected to about 3.5V from a bench supply. in series with the 3.5V is
my DMM set to uA. it measures 596uA!! How is this possible? There is no
other current path except the one from the bench supply to the positive
input and the voltage does not exceed the opamps supply rails. The
datasheet says this opamp has a 110nA nominal input bias current. This
is causing havoc in my circuit because a voltage divider connected to
the positive input has its voltage dragged down by the huge bias
current. WHAT THE HECK IS GOING ON

Dead? Are you SURE that Vin < Vdd ??

...Jim Thompson

 

[[email protected]]

I'm on my 5th part now, so I really doubt the opamp is dead. I have
measured the input voltage at the chip with a dmm at the chips ground
and input pins. I am really getting confused here. Am I killing parts
somehow? I ground myself on the soldering iron before soldering each
chip in. I have also built the circuit in-the-air several times now all
with the same result.

 

[Jim Thompson]

I'm on my 5th part now, so I really doubt the opamp is dead. I have
measured the input voltage at the chip with a dmm at the chips ground
and input pins. I am really getting confused here. Am I killing parts
somehow? I ground myself on the soldering iron before soldering each
chip in. I have also built the circuit in-the-air several times now all
with the same result.

[snip]

Try posting your schematic on a.b.s.e or post a link to your
schematic. Can't really tell what your problem is without a
schematic.

...Jim Thompson

 

[[email protected]]

Okay its in pdf format at

www.phonestockviewer.com/ibias.pdf

Current into the non-inverting input terminal is 596uA!

I'm on my 5th part now, so I really doubt the opamp is dead. I have
measured the input voltage at the chip with a dmm at the chips ground
and input pins. I am really getting confused here. Am I killing parts
somehow? I ground myself on the soldering iron before soldering each
chip in. I have also built the circuit in-the-air several times now all
with the same result.

[snip]

Try posting your schematic on a.b.s.e or post a link to your
schematic. Can't really tell what your problem is without a
schematic.

...Jim Thompson

 

[Fred Bloggs]

:

what the heck. i have an LM6132BIM. the second opamp's inputs and
output are left floating. the first opamp is used. its output is
connected to pin 2, its inverting input, with 200k. its inverting input
is also connected to ground via 33k. the positive input (pin 3) is
connected to about 3.5V from a bench supply. in series with the 3.5V is
my DMM set to uA. it measures 596uA!! How is this possible? There is no
other current path except the one from the bench supply to the positive
input and the voltage does not exceed the opamps supply rails. The
datasheet says this opamp has a 110nA nominal input bias current. This
is causing havoc in my circuit because a voltage divider connected to
the positive input has its voltage dragged down by the huge bias
current. WHAT THE HECK IS GOING ON

You don't know how to identify the chip's pins.

 

[Jim Thompson]

Okay its in pdf format at

www.phonestockviewer.com/ibias.pdf

Current into the non-inverting input terminal is 596uA!

Nope! Output is railed trying to force inverting input node to +3.5V
asked for by non-inverting input. Where are you measuring? My guess
is that inverting input pin is at ~604mV.

I'm on my 5th part now, so I really doubt the opamp is dead. I have
measured the input voltage at the chip with a dmm at the chips ground
and input pins. I am really getting confused here. Am I killing parts
somehow? I ground myself on the soldering iron before soldering each
chip in. I have also built the circuit in-the-air several times now all
with the same result.

[snip]

Try posting your schematic on a.b.s.e or post a link to your
schematic. Can't really tell what your problem is without a
schematic.

...Jim Thompson

...Jim Thompson

 

[legg]

Okay its in pdf format at

www.phonestockviewer.com/ibias.pdf

Current into the non-inverting input terminal is 596uA!

You couldn't get 1/2 ma into that terminal with a sledge hammer.

Check your pinout and use a series resistor to connect the ni input,
not a yard of loopy meter leads.

Bypass the supply terminal. Don't apply low impedance input signals
when power is not on.

Then decide what you want the ouput to be, within the supply rail's
capabilities.

RL

 

[[email protected]]

You couldn't get 1/2 ma into that terminal with a sledge hammer.

Check your pinout and use a series resistor to connect the ni input,
not a yard of loopy meter leads.

Bypass the supply terminal. Don't apply low impedance input signals
when power is not on.

Then decide what you want the ouput to be, within the supply rail's
capabilities.

RL

Believe me, I'm not a newbie at this, and I have tried all of those
things. I put 5V onto a 10k resistor and the other leg of the resistor
has a scope probe. As soon as I touch the resistor to the ni/+ input of
the opamp, the voltage drops from 5 to 3 or 4 .Something is WACK here.
Am I going crazy? I'm going to try a totally different opamp just to
make sure I am still on planet earth

 

[Fred Bartoli]

[email protected] a écrit :

Believe me, I'm not a newbie at this, and I have tried all of those
things. I put 5V onto a 10k resistor and the other leg of the resistor
has a scope probe. As soon as I touch the resistor to the ni/+ input of
the opamp, the voltage drops from 5 to 3 or 4 .Something is WACK here.
Am I going crazy? I'm going to try a totally different opamp just to
make sure I am still on planet earth

If you're not a newbie, then look at
http://cache.national.com/ds/LM/LM6132.pdf#page=11
figure 2.
Now you can explain us why it so. If you can't then... I guess you'll
have to extend somewhat your definition of newbie

If that current is a pb to you, then you have no choice but to change
the opamp model.

 

[colin]

what the heck. i have an LM6132BIM. the second opamp's inputs and
output are left floating. the first opamp is used. its output is
connected to pin 2, its inverting input, with 200k. its inverting input
is also connected to ground via 33k. the positive input (pin 3) is
connected to about 3.5V from a bench supply. in series with the 3.5V is
my DMM set to uA. it measures 596uA!! How is this possible? There is no
other current path except the one from the bench supply to the positive
input and the voltage does not exceed the opamps supply rails. The
datasheet says this opamp has a 110nA nominal input bias current. This
is causing havoc in my circuit because a voltage divider connected to
the positive input has its voltage dragged down by the huge bias
current. WHAT THE HECK IS GOING ON

Sounds like the input protection circuitry is kicking in
youve probably got 0.6v or so accros the input because the output
cant drive the (-) input high enough the feedback network as its limited by
the supply.

Colin =^.^=

 

[Winfield Hill]

[email protected] wrote...

I'm on my 5th part now, so I really doubt the opamp is dead. I have
measured the input voltage at the chip with a dmm at the chips ground
and input pins. I am really getting confused here. Am I killing parts
somehow? I ground myself on the soldering iron before soldering each
chip in. I have also built the circuit in-the-air several times now
all with the same result.

You need to find a different type of opamp. BTW, this is an example
of why it's a good reason for IC manufacturers to include schematics
with their chips. NSC doesn't give you a full schematic, but they
do describe a special feature of the input stage, see figure 2, as
Fred Bartoli suggested. If your source impedance is low enough, you
can drive a great deal of current into the + input if it's more than
+/-0.6V away from - input. This is an intentional LM6132 feature,
meant to increase the slew rate for large input-signal errors, as
shown in figure 1. The current to accomplish this comes from the
input pin. We're sorry, but you need to choose a different part.

 

[Jim Thompson]

Believe me, I'm not a newbie at this, and I have tried all of those
things. I put 5V onto a 10k resistor and the other leg of the resistor
has a scope probe. As soon as I touch the resistor to the ni/+ input of
the opamp, the voltage drops from 5 to 3 or 4 .Something is WACK here.
Am I going crazy? I'm going to try a totally different opamp just to
make sure I am still on planet earth

Try learning to read first... study the replies to your post, study
the schematic, apply MATH, etc ;-)

...Jim Thompson

 

[[email protected]]

Try learning to read first... study the replies to your post, study
the schematic, apply MATH, etc ;-)

...Jim Thompson

I thought thats where figure 2 was leading. Thanks for the help
everyone!

 

[[email protected]]

You need to find a different type of opamp. BTW, this is an example
of why it's a good reason for IC manufacturers to include schematics
with their chips. NSC doesn't give you a full schematic, but they
do describe a special feature of the input stage, see figure 2, as
Fred Bartoli suggested. If your source impedance is low enough, you
can drive a great deal of current into the + input if it's more than
+/-0.6V away from - input. This is an intentional LM6132 feature,
meant to increase the slew rate for large input-signal errors, as
shown in figure 1. The current to accomplish this comes from the
input pin. We're sorry, but you need to choose a different part.

Thanks Win!

FWIW, the spice model in circuitmaker does not show this behavior,
which made things even more confusing. My fault for not reading the
datasheet in more detail.

Asa

 

[Jim Thompson]

Thanks Win!

FWIW, the spice model in circuitmaker does not show this behavior,
which made things even more confusing. My fault for not reading the
datasheet in more detail.

Asa

It's Figure 2 in the data sheet.

You also missed that your configuration, as posted, would rail the
OpAmp EVEN if it didn't include cross-couple slew-boost devices.

I mentioned this before. I suspect you don't understand the
fundamental rule of OpAmps... while in their linear (fed back) region,
the loop forces Vin- EQUAL TO Vin+

...Jim Thompson

 

[Ancient_Hacker]

what the heck. i have an LM6132BIM. the second opamp's inputs and
output are left floating. the first opamp is used. its output is
connected to pin 2, its inverting input, with 200k. its inverting input
is also connected to ground via 33k. the positive input (pin 3) is
connected to about 3.5V from a bench supply. in series with the 3.5V is
my DMM set to uA. it measures 596uA!! How is this possible?

I'll let you figure this out. Hints:


Your resistors are programming the op amp to act like an amplifier with
a voltage gain of 200/33, or a voltage gain of about 6.

You're feeding 3.5 volts to the input.

Sooo the poor op amp is trying to put out _____ volts..

But your Vcc is only ___ volts.

So the output is only rising to ____ volts.

So the voltage fed back thru the 200K resistor is at pin 2, only ____
volts.

So the voltage difference between the input pins is ____ volts.

An op amp is usually designed to be operated with the inputs within ___
microvolts of each other.

If you bias an op-amp at ___ times the expected input difference, the
input bias current specs will be ______ .


A kind of Socratic dialogue, I'm pretty sure you can fill in the
blanks with a little thought.

 

[Jim Thompson]

I'll let you figure this out. Hints:


Your resistors are programming the op amp to act like an amplifier with
a voltage gain of 200/33, or a voltage gain of about 6.

You're feeding 3.5 volts to the input.

Sooo the poor op amp is trying to put out _____ volts..

But your Vcc is only ___ volts.

So the output is only rising to ____ volts.

So the voltage fed back thru the 200K resistor is at pin 2, only ____
volts.

So the voltage difference between the input pins is ____ volts.

An op amp is usually designed to be operated with the inputs within ___
microvolts of each other.

If you bias an op-amp at ___ times the expected input difference, the
input bias current specs will be ______ .


A kind of Socratic dialogue, I'm pretty sure you can fill in the
blanks with a little thought.

There you go...lead 'em by the nose ;-)

...Jim Thompson

 

[[email protected]]

thanks. this isnt how the opamp is actually used in the circuit. this
is how i wired it up to isolate the bias current paths.