Forums

blowing up OPA277's

Started by George Herold November 4, 2016
On Friday, November 4, 2016 at 3:02:56 PM UTC-4, George Herold wrote:
> On Friday, November 4, 2016 at 2:55:09 PM UTC-4, bloggs.fred...@gmail.com wrote: > > On Friday, November 4, 2016 at 2:28:18 PM UTC-4, George Herold wrote: > > > Hi all, I'm prototyping this square root circuit. (I had a thread here maybe a year ago.) The circuit works fine in spice, but turning it on and off a few times.. (or something) is causing the opamp protection diodes to fail. (All of a sudden the inputs are spitting out DC voltages.) > > > > > > A circuit fragment looks like this, > > > > > > > > > Vin----RRRRR-----+----more of the same... > > > | > > > R1 > > > | > > > +--|>|---+ > > > | |\ | > > > +-|-\ | > > > | >--+ > > > V ref--------------|+/ > > > |/ > > > > > > I've found the OPA277's to be a little "delicate" in the past, is there > > > something bad about the diode in the feedback path? > > > > > > I'm going to try a different opamp type. > > > > > > Oh, I also use the similar OPA227... faster more Vos. > > > In the spec sheet for the opa227 they recommend a minimum resistance > > > of 500 ohms in the feedback path. > > > > > > George H. > > > > Can't be that: > > > > 7.3.3 Input Protection > > The inputs of the OPAx277 series are protected with 1-kΩ series input resistors and diode clamps. The inputs > > can withstand ±30-V differential inputs without damage. The protection diodes conduct current when the inputs > > are over-driven. This may disturb the slewing behavior of unity-gain follower applications, but will not damage the > > operational amplifier. > > > > What the heck does inputs spitting out DC voltages mean? > > The circuit seems to work once. Then I add the next bit to it. > Power on and something's wrong. The reference voltage is set by a resistor string > and the voltages along the string are now not right... once they went negative at > one opamp input... another time it was a positive over voltage. > > I unhooked the resistor string, and measured voltage ~-9V .. it's weird. > > I ripped it up and am now using opa134's.. everything is fine so far. > > George H. > (gotta run.)
This can't be happening at DC, have you scoped it out in the time domain?
On Friday, November 4, 2016 at 12:02:56 PM UTC-7, George Herold wrote:
> The circuit seems to work once. Then I add the next bit to it. > Power on and something's wrong.
Agreed with Phil, this sounds like a power supply problem. Hang a scope on your supply rails and watch for bad stuff happening at turnon or turnoff? Another common cause of "works once" is large undamped LC filters that ring way above the bus voltage. -- john, KE5FX
On Mon, 07 Nov 2016 05:22:12 +1000, David Eather <eather@tpg.com.au> wrote:

> On Sat, 05 Nov 2016 04:28:14 +1000, George Herold > <gherold@teachspin.com> wrote: > >> Hi all, I'm prototyping this square root circuit. (I had a thread here >> maybe a year ago.) The circuit works fine in spice, but turning it on >> and off a few times.. (or something) is causing the opamp protection >> diodes to fail. (All of a sudden the inputs are spitting out DC >> voltages.) >> >> A circuit fragment looks like this, >> >> >> Vin----RRRRR-----+----more of the same... >> | >> R1 >> | >> +--|>|---+ >> | |\ | >> +-|-\ | >> | >--+ >> V ref--------------|+/ >> |/ >> >> I've found the OPA277's to be a little "delicate" in the past, is there >> something bad about the diode in the feedback path? >> >> I'm going to try a different opamp type. >> >> Oh, I also use the similar OPA227... faster more Vos. >> In the spec sheet for the opa227 they recommend a minimum resistance >> of 500 ohms in the feedback path. >> >> George H. > > the diode means the OpAmp has to be stable at unity gain
"John Miles, KE5FX"  wrote in message 
news:7ccea0a6-ade3-428d-8689-4a69f91dc721@googlegroups.com...

> On Friday, November 4, 2016 at 12:02:56 PM UTC-7, George Herold wrote: >> The circuit seems to work once. Then I add the next bit to it. >> Power on and something's wrong.
> Agreed with Phil, this sounds like a power supply problem. Hang a scope > on your supply rails and watch for bad stuff happening at turnon or > turnoff?
> Another common cause of "works once" is large undamped LC filters that > ring way above the bus voltage.
I have heard that some cheap variable lab supplies exhibit a high voltage transient at turn-on. I checked mine at 12V and did not see anything. If in doubt, a fairly large low ESR capacitor, as well as perhaps a small inductor, and maybe a TVS diode, would eliminate this possibility. Paul
On Friday, November 4, 2016 at 10:19:11 PM UTC-4, bloggs.fred...@gmail.com wrote:
> On Friday, November 4, 2016 at 3:02:56 PM UTC-4, George Herold wrote: > > On Friday, November 4, 2016 at 2:55:09 PM UTC-4, bloggs.fred...@gmail.com wrote: > > > On Friday, November 4, 2016 at 2:28:18 PM UTC-4, George Herold wrote: > > > > Hi all, I'm prototyping this square root circuit. (I had a thread here maybe a year ago.) The circuit works fine in spice, but turning it on and off a few times.. (or something) is causing the opamp protection diodes to fail. (All of a sudden the inputs are spitting out DC voltages.) > > > > > > > > A circuit fragment looks like this, > > > > > > > > > > > > Vin----RRRRR-----+----more of the same... > > > > | > > > > R1 > > > > | > > > > +--|>|---+ > > > > | |\ | > > > > +-|-\ | > > > > | >--+ > > > > V ref--------------|+/ > > > > |/ > > > > > > > > I've found the OPA277's to be a little "delicate" in the past, is there > > > > something bad about the diode in the feedback path? > > > > > > > > I'm going to try a different opamp type. > > > > > > > > Oh, I also use the similar OPA227... faster more Vos. > > > > In the spec sheet for the opa227 they recommend a minimum resistance > > > > of 500 ohms in the feedback path. > > > > > > > > George H. > > > > > > Can't be that: > > > > > > 7.3.3 Input Protection > > > The inputs of the OPAx277 series are protected with 1-k&Omega; series input resistors and diode clamps. The inputs > > > can withstand &plusmn;30-V differential inputs without damage. The protection diodes conduct current when the inputs > > > are over-driven. This may disturb the slewing behavior of unity-gain follower applications, but will not damage the > > > operational amplifier. > > > > > > What the heck does inputs spitting out DC voltages mean? > > > > The circuit seems to work once. Then I add the next bit to it. > > Power on and something's wrong. The reference voltage is set by a resistor string > > and the voltages along the string are now not right... once they went negative at > > one opamp input... another time it was a positive over voltage. > > > > I unhooked the resistor string, and measured voltage ~-9V .. it's weird. > > > > I ripped it up and am now using opa134's.. everything is fine so far. > > > > George H. > > (gotta run.) > > This can't be happening at DC, have you scoped it out in the time domain?
Not closely... only after it failed. I'll look at the power supply today, it's an old HP triple output 6235A... I've been using it forever without any issues. thanks to everyone else with their ideas. George H.
On Monday, November 7, 2016 at 6:42:55 AM UTC-7, George Herold wrote:
> On Friday, November 4, 2016 at 10:19:11 PM UTC-4, bloggs.fred...@gmail.com wrote: > > On Friday, November 4, 2016 at 3:02:56 PM UTC-4, George Herold wrote: > > > On Friday, November 4, 2016 at 2:55:09 PM UTC-4, bloggs.fred...@gmail.com wrote: > > > > On Friday, November 4, 2016 at 2:28:18 PM UTC-4, George Herold wrote: > > > > > Hi all, I'm prototyping this square root circuit. (I had a thread here maybe a year ago.) The circuit works fine in spice, but turning it on and off a few times.. (or something) is causing the opamp protection diodes to fail. (All of a sudden the inputs are spitting out DC voltages.) > > > > > > > > > > A circuit fragment looks like this, > > > > > > > > > > > > > > > Vin----RRRRR-----+----more of the same... > > > > > | > > > > > R1 > > > > > | > > > > > +--|>|---+ > > > > > | |\ | > > > > > +-|-\ | > > > > > | >--+ > > > > > V ref--------------|+/ > > > > > |/ > > > > > > > > > > I've found the OPA277's to be a little "delicate" in the past, is there > > > > > something bad about the diode in the feedback path? > > > > > > > > > > I'm going to try a different opamp type. > > > > > > > > > > Oh, I also use the similar OPA227... faster more Vos. > > > > > In the spec sheet for the opa227 they recommend a minimum resistance > > > > > of 500 ohms in the feedback path. > > > > > > > > > > George H. > > > > > > > > Can't be that: > > > > > > > > 7.3.3 Input Protection > > > > The inputs of the OPAx277 series are protected with 1-k&Omega; series input resistors and diode clamps. The inputs > > > > can withstand &plusmn;30-V differential inputs without damage. The protection diodes conduct current when the inputs > > > > are over-driven. This may disturb the slewing behavior of unity-gain follower applications, but will not damage the > > > > operational amplifier. > > > > > > > > What the heck does inputs spitting out DC voltages mean? > > > > > > The circuit seems to work once. Then I add the next bit to it. > > > Power on and something's wrong. The reference voltage is set by a resistor string > > > and the voltages along the string are now not right... once they went negative at > > > one opamp input... another time it was a positive over voltage. > > > > > > I unhooked the resistor string, and measured voltage ~-9V .. it's weird. > > > > > > I ripped it up and am now using opa134's.. everything is fine so far. > > > > > > George H. > > > (gotta run.) > > > > This can't be happening at DC, have you scoped it out in the time domain? > > Not closely... only after it failed. > I'll look at the power supply today, it's an old HP triple output 6235A... > I've been using it forever without any issues. > > thanks to everyone else with their ideas. > > George H.
I had a very similar problem a few years ago...it turned out to be the HP 6218C I was powering the circuit with. Transient when powering up.
On Monday, November 7, 2016 at 12:08:09 PM UTC-5, DemonicTubes wrote:
> On Monday, November 7, 2016 at 6:42:55 AM UTC-7, George Herold wrote: > > On Friday, November 4, 2016 at 10:19:11 PM UTC-4, bloggs.fred...@gmail.com wrote: > > > On Friday, November 4, 2016 at 3:02:56 PM UTC-4, George Herold wrote: > > > > On Friday, November 4, 2016 at 2:55:09 PM UTC-4, bloggs.fred...@gmail.com wrote: > > > > > On Friday, November 4, 2016 at 2:28:18 PM UTC-4, George Herold wrote: > > > > > > Hi all, I'm prototyping this square root circuit. (I had a thread here maybe a year ago.) The circuit works fine in spice, but turning it on and off a few times.. (or something) is causing the opamp protection diodes to fail. (All of a sudden the inputs are spitting out DC voltages.) > > > > > > > > > > > > A circuit fragment looks like this, > > > > > > > > > > > > > > > > > > Vin----RRRRR-----+----more of the same... > > > > > > | > > > > > > R1 > > > > > > | > > > > > > +--|>|---+ > > > > > > | |\ | > > > > > > +-|-\ | > > > > > > | >--+ > > > > > > V ref--------------|+/ > > > > > > |/ > > > > > > > > > > > > I've found the OPA277's to be a little "delicate" in the past, is there > > > > > > something bad about the diode in the feedback path? > > > > > > > > > > > > I'm going to try a different opamp type. > > > > > > > > > > > > Oh, I also use the similar OPA227... faster more Vos. > > > > > > In the spec sheet for the opa227 they recommend a minimum resistance > > > > > > of 500 ohms in the feedback path. > > > > > > > > > > > > George H. > > > > > > > > > > Can't be that: > > > > > > > > > > 7.3.3 Input Protection > > > > > The inputs of the OPAx277 series are protected with 1-k&Omega; series input resistors and diode clamps. The inputs > > > > > can withstand &plusmn;30-V differential inputs without damage. The protection diodes conduct current when the inputs > > > > > are over-driven. This may disturb the slewing behavior of unity-gain follower applications, but will not damage the > > > > > operational amplifier. > > > > > > > > > > What the heck does inputs spitting out DC voltages mean? > > > > > > > > The circuit seems to work once. Then I add the next bit to it. > > > > Power on and something's wrong. The reference voltage is set by a resistor string > > > > and the voltages along the string are now not right... once they went negative at > > > > one opamp input... another time it was a positive over voltage. > > > > > > > > I unhooked the resistor string, and measured voltage ~-9V .. it's weird. > > > > > > > > I ripped it up and am now using opa134's.. everything is fine so far. > > > > > > > > George H. > > > > (gotta run.) > > > > > > This can't be happening at DC, have you scoped it out in the time domain? > > > > Not closely... only after it failed. > > I'll look at the power supply today, it's an old HP triple output 6235A... > > I've been using it forever without any issues. > > > > thanks to everyone else with their ideas. > > > > George H. > > I had a very similar problem a few years ago...it turned out to be the HP 6218C I was powering the circuit with. Transient when powering up.
Well the power supply is perfectly well behaved in turn on and off. Circuit works fine with a different opamp. I'm moving on, but would like to hear from anyone using the oap227 (and cousins) who has had "issues". George H.
On Monday, 7 November 2016 18:35:46 UTC, George Herold  wrote:

> Well the power supply is perfectly well behaved in turn on and off. > Circuit works fine with a different opamp. > I'm moving on, but would like to hear from anyone using the oap227 (and cousins) > who has had "issues". > > George H.
I've used similar opamps and have issues... the 2 aren't really connected though :) NT