Forums

static precaution, fet opamp

Started by Unknown July 5, 2015
I have a minimal opamp circuit that uses a random very low current opamp, such as TL062 (but could be anything low current). This thing is going to be used in uncontrolled conditions in the 3rd world, and needs to be relatively reliable. 

The outside world goes to -in via 68k + 47nF. Question is what level of extra static protection do I need. What I've added so far is a diode from 0v to the outwide world input, and another diode from opamp -in to the half Vcc rail, which has 0.5M resistance and 0.22uF smoothing cap. (Its very high impedance for a reason.) The 2nd diode goes to half Vcc not Vcc since its smoothed, I don't want the diode capacitance coupling psu crap to the opamp input.

The opamp has no static protection on other pins yet, I daresay it'll need them doing too. I'm thinking 2 diodes + resistor on the output. Power input is protected by a reverse diode plus a very small reservoir cap.

Enough? Overkill? Completely wrong?


NT
On Sunday, 5 July 2015 21:46:59 UTC+1, tabb...@gmail.com  wrote:
> I have a minimal opamp circuit that uses a random very low current opamp, such as TL062 (but could be anything low current). This thing is going to be used in uncontrolled conditions in the 3rd world, and needs to be relatively reliable. > > The outside world goes to -in via 68k + 47nF. Question is what level of extra static protection do I need. What I've added so far is a diode from 0v to the outwide world input, and another diode from opamp -in to the half Vcc rail, which has 0.5M resistance and 0.22uF smoothing cap. (Its very high impedance for a reason.) The 2nd diode goes to half Vcc not Vcc since its smoothed, I don't want the diode capacitance coupling psu crap to the opamp input. > > The opamp has no static protection on other pins yet, I daresay it'll need them doing too. I'm thinking 2 diodes + resistor on the output. Power input is protected by a reverse diode plus a very small reservoir cap. > > Enough? Overkill? Completely wrong?
Just to clarify I basically am looking for the minimum configuration that'll do the job. Otherwise I'd stick diodes & resistors all over it and not worry. Thank you. NT
On Sunday, 5 July 2015 21:51:40 UTC+1, tabb...@gmail.com  wrote:
> On Sunday, 5 July 2015 21:46:59 UTC+1, tabb...@gmail.com wrote: > > I have a minimal opamp circuit that uses a random very low current opamp, such as TL062 (but could be anything low current). This thing is going to be used in uncontrolled conditions in the 3rd world, and needs to be relatively reliable. > > > > The outside world goes to -in via 68k + 47nF. Question is what level of extra static protection do I need. What I've added so far is a diode from 0v to the outwide world input, and another diode from opamp -in to the half Vcc rail, which has 0.5M resistance and 0.22uF smoothing cap. (Its very high impedance for a reason.) The 2nd diode goes to half Vcc not Vcc since its smoothed, I don't want the diode capacitance coupling psu crap to the opamp input. > > > > The opamp has no static protection on other pins yet, I daresay it'll need them doing too. I'm thinking 2 diodes + resistor on the output. Power input is protected by a reverse diode plus a very small reservoir cap. > > > > Enough? Overkill? Completely wrong? > > Just to clarify I basically am looking for the minimum configuration that'll do the job. Otherwise I'd stick diodes & resistors all over it and not worry. Thank you.
There isn't really any effective short cut is there. :) NT
What voltage range?

What supply range?

What devices are available (clamp diodes, zeners, TVS, galena crystals and 
other floor sweepings...?!)?

How much capacitance/resistance load is permissible?  What bandwidth? 
Distortion specs?

Also, presumably you have a feedback resistor to -in somewhere, or at least 
a bias resistor to ground.  Otherwise that input's flapping in the breeze 
and will eventually saturate one way or another...

Tim

-- 
Seven Transistor Labs
Electrical Engineering Consultation
Website: http://seventransistorlabs.com

<tabbypurr@gmail.com> wrote in message 
news:dc5e0ae1-e94a-409f-90fe-19a80a0f4901@googlegroups.com...
I have a minimal opamp circuit that uses a random very low current opamp, 
such as TL062 (but could be anything low current). This thing is going to be 
used in uncontrolled conditions in the 3rd world, and needs to be relatively 
reliable.

The outside world goes to -in via 68k + 47nF. Question is what level of 
extra static protection do I need. What I've added so far is a diode from 0v 
to the outwide world input, and another diode from opamp -in to the half Vcc 
rail, which has 0.5M resistance and 0.22uF smoothing cap. (Its very high 
impedance for a reason.) The 2nd diode goes to half Vcc not Vcc since its 
smoothed, I don't want the diode capacitance coupling psu crap to the opamp 
input.

The opamp has no static protection on other pins yet, I daresay it'll need 
them doing too. I'm thinking 2 diodes + resistor on the output. Power input 
is protected by a reverse diode plus a very small reservoir cap.

Enough? Overkill? Completely wrong?


NT 

On Sunday, July 5, 2015 at 1:46:59 PM UTC-7, tabb...@gmail.com wrote:
> I have a minimal opamp circuit that uses a random very low current opamp, such as TL062 (but could be anything low current). This thing is going to be used in uncontrolled conditions in the 3rd world, and needs to be relatively reliable. > > The outside world goes to -in via 68k + 47nF. Question is what level of extra static protection do I need.
The TL062 is a jFET input op amp; it's unlikely to take harm from the small charge of a static jolt, because it doesn't have a thin oxide that can be punctured. There's plenty of protection diodes, accidentally built in. Thermal overload can happen, though; you can protect it with a fuse or fusible resistor against that, or use back-to-back zeners. If you use protection diodes, and 220VAC is applied, there might be over half a watt dissipated in the 68k resistor; can it take that?
On Sunday, 5 July 2015 22:48:26 UTC+1, Tim Williams  wrote:
> nt wrote in message > news:dc5e0ae1-e94a-409f-90fe-19a80a0f4901@googlegroups.com...
> I have a minimal opamp circuit that uses a random very low current opamp, > such as TL062 (but could be anything low current). This thing is going to be > used in uncontrolled conditions in the 3rd world, and needs to be relatively > reliable. > > The outside world goes to -in via 68k + 47nF. Question is what level of > extra static protection do I need. What I've added so far is a diode from 0v > to the outwide world input, and another diode from opamp -in to the half Vcc > rail, which has 0.5M resistance and 0.22uF smoothing cap. (Its very high > impedance for a reason.) The 2nd diode goes to half Vcc not Vcc since its > smoothed, I don't want the diode capacitance coupling psu crap to the opamp > input. > > The opamp has no static protection on other pins yet, I daresay it'll need > them doing too. I'm thinking 2 diodes + resistor on the output. Power input > is protected by a reverse diode plus a very small reservoir cap. > > Enough? Overkill? Completely wrong?
> What voltage range? > > What supply range?
1 - 12v, typically 6-9v
> What devices are available (clamp diodes, zeners, TVS, galena crystals and > other floor sweepings...?!)?
the least possible cost devices from scrap electronics. I'm going so far with unspecced small diodes plus resistor.
> How much capacitance/resistance load is permissible?
its driving around 1mA load. I don't want to increase that current by much. Power consumption is the big deal on this one.
> What bandwidth?
audio
> Distortion specs?
'meh' is the official distortion spec
> Also, presumably you have a feedback resistor to -in somewhere, or at least > a bias resistor to ground. Otherwise that input's flapping in the breeze > and will eventually saturate one way or another...
sure, I'm not worried about it saturating though. NT
On Sunday, 5 July 2015 23:55:26 UTC+1, whit3rd  wrote:
> On Sunday, July 5, 2015 at 1:46:59 PM UTC-7, tabb...@gmail.com wrote:
> > I have a minimal opamp circuit that uses a random very low current opamp, such as TL062 (but could be anything low current). This thing is going to be used in uncontrolled conditions in the 3rd world, and needs to be relatively reliable. > > > > The outside world goes to -in via 68k + 47nF. Question is what level of extra static protection do I need. > > The TL062 is a jFET input op amp; it's unlikely to take harm from the small charge of a static jolt, > because it doesn't have a thin oxide that can be punctured. There's plenty of protection > diodes, accidentally built in.
I presume one can't count on either of those with random opamps.
> Thermal overload can happen, though; you can protect > it with a fuse or fusible resistor against that, or use back-to-back zeners.
I can't see that being a problem in this circuit.
> If you use protection diodes, and 220VAC is applied, there might be over half a watt > dissipated in the 68k resistor; can it take that?
It doesn't need to survive 220v. If it were applied, its a gamble as to whether the input cap would take it. 47nF so likely ceramic, so the odds aren't bad. If it did, everything else would hold up ok. NT
On Sunday, 5 July 2015 22:35:49 UTC+1, nt  wrote:
> On Sunday, 5 July 2015 21:51:40 UTC+1, nt wrote: > > On Sunday, 5 July 2015 21:46:59 UTC+1, nt wrote:
> > > I have a minimal opamp circuit that uses a random very low current opamp, such as TL062 (but could be anything low current). This thing is going to be used in uncontrolled conditions in the 3rd world, and needs to be relatively reliable. > > > > > > The outside world goes to -in via 68k + 47nF. Question is what level of extra static protection do I need. What I've added so far is a diode from 0v to the outwide world input, and another diode from opamp -in to the half Vcc rail, which has 0.5M resistance and 0.22uF smoothing cap. (Its very high impedance for a reason.) The 2nd diode goes to half Vcc not Vcc since its smoothed, I don't want the diode capacitance coupling psu crap to the opamp input. > > > > > > The opamp has no static protection on other pins yet, I daresay it'll need them doing too. I'm thinking 2 diodes + resistor on the output. Power input is protected by a reverse diode plus a very small reservoir cap. > > > > > > Enough? Overkill? Completely wrong? > > > > Just to clarify I basically am looking for the minimum configuration that'll do the job. Otherwise I'd stick diodes & resistors all over it and not worry. Thank you. > > There isn't really any effective short cut is there. :)
ok, on the output I've got 390R then diodes to 0v and Vcc. I think the answer to the next question is no: can I rely on those diodes to also protect against wrong battery polarity? They'd limit Vcc to about -2v. NT