Forums

Opamp (output) over voltage protection

Started by George Herold June 23, 2017
Hi all, So I've got an opamp driving a remote FET.  The current incarnation of
the circuit has students able to change (screw up) the connections.  
There's a ~60V power supply available so I'd like to make it so that the 
opamp survives having it's output connected to 60 V.  Since it's driving a
Fet (slowly) there is not much current and I'm sticking in two series 
depletion Fets (LND150's) (input and output by the drains, G and S's all
shorted.)  I'm just wondering about this ~1.5 mA feeding back into the 
opamp (opa2134).  Should I put diodes to the power rails on the output?  
Or can the opamp output withstand the 1 mA?  

(I'll find out, but just looking for advice before I let the magic smoke
out. :^)

George H.  
On Friday, June 23, 2017 at 9:11:56 AM UTC-4, George Herold wrote:
> Hi all, So I've got an opamp driving a remote FET. The current incarnation of > the circuit has students able to change (screw up) the connections. > There's a ~60V power supply available so I'd like to make it so that the > opamp survives having it's output connected to 60 V. Since it's driving a > Fet (slowly) there is not much current and I'm sticking in two series > depletion Fets (LND150's) (input and output by the drains, G and S's all > shorted.) I'm just wondering about this ~1.5 mA feeding back into the > opamp (opa2134). Should I put diodes to the power rails on the output? > Or can the opamp output withstand the 1 mA? > > (I'll find out, but just looking for advice before I let the magic smoke > out. :^) > > George H.
With no diodes to power rail, the 'output' of the opamp went to ~1.5V above the supply rail. And swallowed the 1.5 mA. Seems OK. George H.
On 06/23/2017 09:51 AM, George Herold wrote:
> On Friday, June 23, 2017 at 9:11:56 AM UTC-4, George Herold wrote: >> Hi all, So I've got an opamp driving a remote FET. The current incarnation of >> the circuit has students able to change (screw up) the connections. >> There's a ~60V power supply available so I'd like to make it so that the >> opamp survives having it's output connected to 60 V. Since it's driving a >> Fet (slowly) there is not much current and I'm sticking in two series >> depletion Fets (LND150's) (input and output by the drains, G and S's all >> shorted.) I'm just wondering about this ~1.5 mA feeding back into the >> opamp (opa2134). Should I put diodes to the power rails on the output? >> Or can the opamp output withstand the 1 mA? >> >> (I'll find out, but just looking for advice before I let the magic smoke >> out. :^) >> >> George H. > > With no diodes to power rail, the 'output' of the opamp went to > ~1.5V above the supply rail. And swallowed the 1.5 mA. Seems OK. > > George H. >
Whats the supply voltage? If you're really paranoid you could buffer the output of the op amp with CMOS gates, 4 NANDs in parallel to source current into the load, 4 NORs to sink. Socket them so if worst comes to worst and they blow from an incorrect connection you'll be able to have the culprit swap them out
On Friday, June 23, 2017 at 11:04:03 AM UTC-4, bitrex wrote:
> On 06/23/2017 09:51 AM, George Herold wrote: > > On Friday, June 23, 2017 at 9:11:56 AM UTC-4, George Herold wrote: > >> Hi all, So I've got an opamp driving a remote FET. The current incarnation of > >> the circuit has students able to change (screw up) the connections. > >> There's a ~60V power supply available so I'd like to make it so that the > >> opamp survives having it's output connected to 60 V. Since it's driving a > >> Fet (slowly) there is not much current and I'm sticking in two series > >> depletion Fets (LND150's) (input and output by the drains, G and S's all > >> shorted.) I'm just wondering about this ~1.5 mA feeding back into the > >> opamp (opa2134). Should I put diodes to the power rails on the output? > >> Or can the opamp output withstand the 1 mA? > >> > >> (I'll find out, but just looking for advice before I let the magic smoke > >> out. :^) > >> > >> George H. > > > > With no diodes to power rail, the 'output' of the opamp went to > > ~1.5V above the supply rail. And swallowed the 1.5 mA. Seems OK. > > > > George H. > > > > Whats the supply voltage? If you're really paranoid you could buffer the
+/- 15V, but I tried it with the supply shut off too. Output went to +/- 2.5 V. I'm not really sure if this is true, but I figure if the IC is only dissipating (2.5V * 1.5 mA) ~4 mW nothing can be damaged.
> output of the op amp with CMOS gates, 4 NANDs in parallel to source > current into the load, 4 NORs to sink. > > Socket them so if worst comes to worst and they blow from an incorrect > connection you'll be able to have the culprit swap them out
I'm not sure I understand the Cmos idea. This is a linear circuit so I need to be able to run the Fet gate at different voltages. These LND150's are amazing. I'm sticking 'em all over the circuit*. It would be really cool if someone put them into a two terminal package. George H. *I've also got them buffering a current source feeding a diode temperature sensor. One (pair) to save the current source, another two to protect the drive and sense lines of the diode. Current source is 10/100 uA so a few k ohm in series doesn't hurt.
On 06/23/2017 12:05 PM, George Herold wrote:
> On Friday, June 23, 2017 at 11:04:03 AM UTC-4, bitrex wrote: >> On 06/23/2017 09:51 AM, George Herold wrote: >>> On Friday, June 23, 2017 at 9:11:56 AM UTC-4, George Herold wrote: >>>> Hi all, So I've got an opamp driving a remote FET. The current incarnation of >>>> the circuit has students able to change (screw up) the connections. >>>> There's a ~60V power supply available so I'd like to make it so that the >>>> opamp survives having it's output connected to 60 V. Since it's driving a >>>> Fet (slowly) there is not much current and I'm sticking in two series >>>> depletion Fets (LND150's) (input and output by the drains, G and S's all >>>> shorted.) I'm just wondering about this ~1.5 mA feeding back into the >>>> opamp (opa2134). Should I put diodes to the power rails on the output? >>>> Or can the opamp output withstand the 1 mA? >>>> >>>> (I'll find out, but just looking for advice before I let the magic smoke >>>> out. :^) >>>> >>>> George H. >>> >>> With no diodes to power rail, the 'output' of the opamp went to >>> ~1.5V above the supply rail. And swallowed the 1.5 mA. Seems OK. >>> >>> George H. >>> >> >> Whats the supply voltage? If you're really paranoid you could buffer the > > +/- 15V, but I tried it with the supply shut off too. > Output went to +/- 2.5 V. > > I'm not really sure if this is true, but I figure if the IC is only > dissipating (2.5V * 1.5 mA) ~4 mW nothing can be damaged. > >> output of the op amp with CMOS gates, 4 NANDs in parallel to source >> current into the load, 4 NORs to sink. >> >> Socket them so if worst comes to worst and they blow from an incorrect >> connection you'll be able to have the culprit swap them out > > I'm not sure I understand the Cmos idea. This is a linear circuit > so I need to be able to run the Fet gate at different voltages.
Ya, no problem! You can send the output of an op-amp into some paralleled CMOS gate inputs and then run the feedback resistor around from the paralleled outputs of the gates back to the input of the op-amp, now it'll work as a linear circuit. The propagation delay of most 40xx logic gates is in the single-digit ns and shouldn't affect loop stability at all. You get R-R output for "free", too. It's a thing. It's in an app note.
On 06/23/2017 12:20 PM, bitrex wrote:
> On 06/23/2017 12:05 PM, George Herold wrote: >> On Friday, June 23, 2017 at 11:04:03 AM UTC-4, bitrex wrote: >>> On 06/23/2017 09:51 AM, George Herold wrote: >>>> On Friday, June 23, 2017 at 9:11:56 AM UTC-4, George Herold wrote: >>>>> Hi all, So I've got an opamp driving a remote FET. The current >>>>> incarnation of >>>>> the circuit has students able to change (screw up) the connections. >>>>> There's a ~60V power supply available so I'd like to make it so >>>>> that the >>>>> opamp survives having it's output connected to 60 V. Since it's >>>>> driving a >>>>> Fet (slowly) there is not much current and I'm sticking in two series >>>>> depletion Fets (LND150's) (input and output by the drains, G and >>>>> S's all >>>>> shorted.) I'm just wondering about this ~1.5 mA feeding back into the >>>>> opamp (opa2134). Should I put diodes to the power rails on the >>>>> output? >>>>> Or can the opamp output withstand the 1 mA? >>>>> >>>>> (I'll find out, but just looking for advice before I let the magic >>>>> smoke >>>>> out. :^) >>>>> >>>>> George H. >>>> >>>> With no diodes to power rail, the 'output' of the opamp went to >>>> ~1.5V above the supply rail. And swallowed the 1.5 mA. Seems OK. >>>> >>>> George H. >>>> >>> >>> Whats the supply voltage? If you're really paranoid you could buffer the >> >> +/- 15V, but I tried it with the supply shut off too. >> Output went to +/- 2.5 V. >> >> I'm not really sure if this is true, but I figure if the IC is only >> dissipating (2.5V * 1.5 mA) ~4 mW nothing can be damaged. >> >>> output of the op amp with CMOS gates, 4 NANDs in parallel to source >>> current into the load, 4 NORs to sink. >>> >>> Socket them so if worst comes to worst and they blow from an incorrect >>> connection you'll be able to have the culprit swap them out >> >> I'm not sure I understand the Cmos idea. This is a linear circuit >> so I need to be able to run the Fet gate at different voltages. > > Ya, no problem! You can send the output of an op-amp into some > paralleled CMOS gate inputs and then run the feedback resistor around > from the paralleled outputs of the gates back to the input of the > op-amp, now it'll work as a linear circuit. The propagation delay of > most 40xx logic gates is in the single-digit ns and shouldn't affect > loop stability at all. > > You get R-R output for "free", too. It's a thing. It's in an app note. >
Sorry, 74HC may be in the single-digits, standard 4000 series will be longer but still probably irrelevant Anyway see page 3: <https://www.fairchildsemi.com/application-notes/AN/AN-88.pdf>
On Friday, June 23, 2017 at 12:25:46 PM UTC-4, bitrex wrote:
> On 06/23/2017 12:20 PM, bitrex wrote: > > On 06/23/2017 12:05 PM, George Herold wrote: > >> On Friday, June 23, 2017 at 11:04:03 AM UTC-4, bitrex wrote: > >>> On 06/23/2017 09:51 AM, George Herold wrote: > >>>> On Friday, June 23, 2017 at 9:11:56 AM UTC-4, George Herold wrote: > >>>>> Hi all, So I've got an opamp driving a remote FET. The current > >>>>> incarnation of > >>>>> the circuit has students able to change (screw up) the connections. > >>>>> There's a ~60V power supply available so I'd like to make it so > >>>>> that the > >>>>> opamp survives having it's output connected to 60 V. Since it's > >>>>> driving a > >>>>> Fet (slowly) there is not much current and I'm sticking in two series > >>>>> depletion Fets (LND150's) (input and output by the drains, G and > >>>>> S's all > >>>>> shorted.) I'm just wondering about this ~1.5 mA feeding back into the > >>>>> opamp (opa2134). Should I put diodes to the power rails on the > >>>>> output? > >>>>> Or can the opamp output withstand the 1 mA? > >>>>> > >>>>> (I'll find out, but just looking for advice before I let the magic > >>>>> smoke > >>>>> out. :^) > >>>>> > >>>>> George H. > >>>> > >>>> With no diodes to power rail, the 'output' of the opamp went to > >>>> ~1.5V above the supply rail. And swallowed the 1.5 mA. Seems OK. > >>>> > >>>> George H. > >>>> > >>> > >>> Whats the supply voltage? If you're really paranoid you could buffer the > >> > >> +/- 15V, but I tried it with the supply shut off too. > >> Output went to +/- 2.5 V. > >> > >> I'm not really sure if this is true, but I figure if the IC is only > >> dissipating (2.5V * 1.5 mA) ~4 mW nothing can be damaged. > >> > >>> output of the op amp with CMOS gates, 4 NANDs in parallel to source > >>> current into the load, 4 NORs to sink. > >>> > >>> Socket them so if worst comes to worst and they blow from an incorrect > >>> connection you'll be able to have the culprit swap them out > >> > >> I'm not sure I understand the Cmos idea. This is a linear circuit > >> so I need to be able to run the Fet gate at different voltages. > > > > Ya, no problem! You can send the output of an op-amp into some > > paralleled CMOS gate inputs and then run the feedback resistor around > > from the paralleled outputs of the gates back to the input of the > > op-amp, now it'll work as a linear circuit. The propagation delay of > > most 40xx logic gates is in the single-digit ns and shouldn't affect > > loop stability at all. > > > > You get R-R output for "free", too. It's a thing. It's in an app note. > > > > Sorry, 74HC may be in the single-digits, standard 4000 series will be > longer but still probably irrelevant > > Anyway see page 3: > > <https://www.fairchildsemi.com/application-notes/AN/AN-88.pdf>
OK, I guess these days if I needed R-R output I'd choose an opamp that does that. And how does a Cmos 'buffer' on the output save me from overvoltage? Are Cmos protected from having their output lifted to +/- 60V? George H.
On 06/23/2017 12:42 PM, George Herold wrote:

> OK, I guess these days if I needed R-R output I'd choose an opamp > that does that. And how does a Cmos 'buffer' on the output save > me from overvoltage? Are Cmos protected from having their output lifted to +/- 60V? > > George H. >
No, they might die and sacrifice themselves to protect the op-amp. So if it stops working then you know what happened. And you then direct the student to kindly remove the sacrificial gates from their sockets, make their way over to the bin holding the reel of replacements, and please pay attention in future and don't do that again, thanks! :-)
On Friday, June 23, 2017 at 12:58:02 PM UTC-4, bitrex wrote:
> On 06/23/2017 12:42 PM, George Herold wrote: > > > OK, I guess these days if I needed R-R output I'd choose an opamp > > that does that. And how does a Cmos 'buffer' on the output save > > me from overvoltage? Are Cmos protected from having their output lifted to +/- 60V? > > > > George H. > > > > No, they might die and sacrifice themselves to protect the op-amp. So if > it stops working then you know what happened. And you then direct the > student to kindly remove the sacrificial gates from their sockets, make > their way over to the bin holding the reel of replacements, and please > pay attention in future and don't do that again, thanks! :-)
Oh, OK. I've used opamp's (DIL 8) as fuses in a different instrument. Put 'em in a socket and send 'em spares. The difference between a $0.50 Cmos IC and a $2.00 opamp is unimportant. But in that instrument the opamp was right in the circuit they were building, so easy to test and replace. In this case the opamp is in a separate box. Not so simple. but hey, thanks for playing with me. :^) George H.
On Fri, 23 Jun 2017 09:42:34 -0700 (PDT), George Herold
<gherold@teachspin.com> wrote:

>On Friday, June 23, 2017 at 12:25:46 PM UTC-4, bitrex wrote: >> On 06/23/2017 12:20 PM, bitrex wrote: >> > On 06/23/2017 12:05 PM, George Herold wrote: >> >> On Friday, June 23, 2017 at 11:04:03 AM UTC-4, bitrex wrote: >> >>> On 06/23/2017 09:51 AM, George Herold wrote: >> >>>> On Friday, June 23, 2017 at 9:11:56 AM UTC-4, George Herold wrote: >> >>>>> Hi all, So I've got an opamp driving a remote FET. The current >> >>>>> incarnation of >> >>>>> the circuit has students able to change (screw up) the connections. >> >>>>> There's a ~60V power supply available so I'd like to make it so >> >>>>> that the >> >>>>> opamp survives having it's output connected to 60 V. Since it's >> >>>>> driving a >> >>>>> Fet (slowly) there is not much current and I'm sticking in two series >> >>>>> depletion Fets (LND150's) (input and output by the drains, G and >> >>>>> S's all >> >>>>> shorted.) I'm just wondering about this ~1.5 mA feeding back into the >> >>>>> opamp (opa2134). Should I put diodes to the power rails on the >> >>>>> output? >> >>>>> Or can the opamp output withstand the 1 mA? >> >>>>> >> >>>>> (I'll find out, but just looking for advice before I let the magic >> >>>>> smoke >> >>>>> out. :^) >> >>>>> >> >>>>> George H. >> >>>> >> >>>> With no diodes to power rail, the 'output' of the opamp went to >> >>>> ~1.5V above the supply rail. And swallowed the 1.5 mA. Seems OK. >> >>>> >> >>>> George H. >> >>>> >> >>> >> >>> Whats the supply voltage? If you're really paranoid you could buffer the >> >> >> >> +/- 15V, but I tried it with the supply shut off too. >> >> Output went to +/- 2.5 V. >> >> >> >> I'm not really sure if this is true, but I figure if the IC is only >> >> dissipating (2.5V * 1.5 mA) ~4 mW nothing can be damaged. >> >> >> >>> output of the op amp with CMOS gates, 4 NANDs in parallel to source >> >>> current into the load, 4 NORs to sink. >> >>> >> >>> Socket them so if worst comes to worst and they blow from an incorrect >> >>> connection you'll be able to have the culprit swap them out >> >> >> >> I'm not sure I understand the Cmos idea. This is a linear circuit >> >> so I need to be able to run the Fet gate at different voltages. >> > >> > Ya, no problem! You can send the output of an op-amp into some >> > paralleled CMOS gate inputs and then run the feedback resistor around >> > from the paralleled outputs of the gates back to the input of the >> > op-amp, now it'll work as a linear circuit. The propagation delay of >> > most 40xx logic gates is in the single-digit ns and shouldn't affect >> > loop stability at all. >> > >> > You get R-R output for "free", too. It's a thing. It's in an app note. >> > >> >> Sorry, 74HC may be in the single-digits, standard 4000 series will be >> longer but still probably irrelevant >> >> Anyway see page 3: >> >> <https://www.fairchildsemi.com/application-notes/AN/AN-88.pdf> > >OK, I guess these days if I needed R-R output I'd choose an opamp >that does that. And how does a Cmos 'buffer' on the output save >me from overvoltage? Are Cmos protected from having their output lifted to +/- 60V? > >George H.
Bitrex thinks so >:-} _Some_ 'HC' parts are designed for (tri-state) bus driver/receiver applications and the output can be driven above local rail voltage (VDD) without harm... but NOT below VSS. ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | STV, Queen Creek, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | Thinking outside the box... producing elegant solutions. "It is not in doing what you like, but in liking what you do that is the secret of happiness." -James Barrie