Author Topic: For an opamp input, which network protection is preferable?  (Read 2956 times)

0 Members and 1 Guest are viewing this topic.

Offline schmitt triggerTopic starter

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
For an opamp input, which network protection is preferable?
« on: April 17, 2019, 06:17:12 pm »
I am in the process of designing a 10X preamplifier to extend the range of an analog meter.
The input will be +/- 300mV DC or AC pk-pk. Hopefully I can get a 100 Khz bandwidth.

However, I know that sh!t happens, and require to protect the input against mild overvoltages.

The standard protection network as shown in (A), series schottkys to V+ and V-, works fine. But schottkys have significant leakage, and because of the 1 Meg impedance, it could generate offset voltage errors.

Therefore I was thinking along the lines of inverse parallel diodes from input to ground, as shown in (B). But while the reverse-bias capacitance is well specified, the capacitance below the forward bias is not. Using my Peak Atlas LCR meter, a pair of antiparallel 1N4148 show about 50 pF, which could cause loading at higher frequencies.
I don't know how much I can trust this reading.

I had read somewhere that LEDs, covered to avoid photoelectric effects, show both ultra-low leakage and minimum capacitance. Again using the Peak Atlas, 3 mm green LEDs measures 15 pF.

So what do you think? What would be the best method?
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7678
  • Country: ca
Re: For an opamp input, which network protection is preferable?
« Reply #1 on: April 17, 2019, 06:48:42 pm »
I would just use low-leakage silicon diodes or diode-connected transistors, and assuming that clamp voltage is now down to say +/-1.3V, then the next resistor to limit op-amp input current.
OPA1611 limit is +/-0.5V past the rails and +/-10mA. Unless power could be off when input or ESD is present, so 2V/10mA is around 200R minimum.

I think your impedances are really high 90k, 505k I would use 1/10 of that or lower for lower noise and less susceptibility to moisture.

If I need higher clamp voltages I've used BAV99S quadruple package, for clamps to +/-2.5V at 150mA.
TVS diodes most have high capacitance, although some Ethernet/USB parts are also very good.
 
The following users thanked this post: schmitt trigger

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7043
  • Country: nl
Re: For an opamp input, which network protection is preferable?
« Reply #2 on: April 17, 2019, 07:19:27 pm »
However, I know that sh!t happens, and require to protect the input against mild overvoltages.

10 mA * 100 kOhm = 1 kV ... forget about it. Almost all active components have protection build in, so that level of series resistance is protection enough.
 
The following users thanked this post: nctnico, schmitt trigger

Offline schmitt triggerTopic starter

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: For an opamp input, which network protection is preferable?
« Reply #3 on: April 17, 2019, 08:29:39 pm »
Thanks Marco!

Now that you mention, it is perfectly obvious.

A case of not seeing the forest because the sight is fixed on the trees.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: For an opamp input, which network protection is preferable?
« Reply #4 on: April 17, 2019, 09:21:55 pm »
OPA1611? There will be up to a few mV of offset error from input bias currents and the rather high and unequal impedances of input networks.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7678
  • Country: ca
Re: For an opamp input, which network protection is preferable?
« Reply #5 on: April 17, 2019, 10:33:30 pm »
10 mA * 100 kOhm = 1 kV ... forget about it. Almost all active components have protection build in, so that level of series resistance is protection enough.

I notice the OPA1611 already has the diodes built-in. It does have 60-250nA input bias current, it's bi-polar not FET input.
ESD events are easily over 1kV and just kill the op-amp, or arc across an input resistor and kill the op-amp.
The input resistor should be good for an ESD hit, or a spark-gap on the PC board traces.

 

Offline schmitt triggerTopic starter

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: For an opamp input, which network protection is preferable?
« Reply #6 on: April 18, 2019, 12:06:46 am »
OPA1611? There will be up to a few mV of offset error from input bias currents and the rather high and unequal impedances of input networks.

Indeed;
I am considering lowering and matching them.
 

Offline schmitt triggerTopic starter

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: For an opamp input, which network protection is preferable?
« Reply #7 on: April 18, 2019, 12:29:53 am »
Giving this problem further thoughts;

Opamp performance parameters are always tradeoffs.
After Magic’s comments, I have come to the realization that I cannot have low voltage offset AND at least 1Meg input impedance AND overload protection AND wide bandwidth.

Traditionally, this would mean a two-amp solution. Each tailored for a specific performance parameter?
 

Offline profdc9

  • Frequent Contributor
  • **
  • Posts: 334
  • Country: us
Re: For an opamp input, which network protection is preferable?
« Reply #8 on: April 18, 2019, 12:49:05 am »
You may not need protection diodes at the input of your op-amp, but as a note:

A 1N4148 probably has on the order of 1 pF of capacitance, not 50 pF.  It is a fairly low capacitance diode and will not load the input significantly.  If you want +/-1 V limiting you can use two in series, or four total.

But this protection may be already built into your op-amp and would be redundant and only add additional capacitance without adding any benefit.

 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: For an opamp input, which network protection is preferable?
« Reply #9 on: April 18, 2019, 01:33:31 am »
What's the expected source impedance?  A meg or less?

What's the desired accuracy -- how much DC and noise error is tolerable?

What's the temperature range?

BAT54 is out, due to leakage even at room temperature, but PN diodes (BAV99, etc.) are open, at least as long as the required accuracy isn't stellar.

LEDs are indeed quite good, or they should be; the downside is, you're dependent on unspecified characteristics.  (By the way, don't use blue LEDs!  They are more ESD sensitive than red.)

Note that the leading 10k resistor is likely to arc over in case of ESD, so choose a diode or TVS capable of handling full surge current, whatever that may be.

If you need terribly high accuracy, consider using a buffer stage instead, and bootstrapping the ESD diodes.  Then followed with the requisite gain stage.

Bootstrapped ESD diodes are wired as shown (typically with the antiparallel pair), but instead of a ground return, they connect to the buffer's output, through a small resistor.  Supply clamp or TVS diodes then limit the buffered voltage (after the series resistor), so that incoming ESD goes through two diodes in series (the antiparallel pair, and the clamp), and not through the buffer's input or output structures (thanks to the series resistors).

Zeners or even GDTs may also be of interest, despite the low operating input voltage.  A 12V zener (antiseries), placed where the LEDs are shown, will have quite low leakage at signal voltages (<6V say?), and the peak clamping voltage (say ~18V, plus ESL?) won't deliver too much current into the op-amp input pins, thanks to the second series resistor.  (But mind that the injected current may well upset the amp's internal workings, leading to a long recovery tail, or hopefully not latchup, but you never know I guess.)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: schmitt trigger

Offline schmitt triggerTopic starter

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: For an opamp input, which network protection is preferable?
« Reply #10 on: April 18, 2019, 02:12:02 am »
Wow!!
Excellent advice Tim. I find the idea of bootstrapping the ESD diodes very clever.

For the 10k input resistor, I had already considered using multiple devices in series, both to increase power rating and the voltage withstanding capabilities.

As I mentioned in my original post, the purpose of this little circuit is to extend the low level range, and increase the input impedance of an analog meter (full scale = 2.5 volts AC, 9kohm/volt).
As such I don't require terribly high accuracy (0.5% would be excellent) nor ultra-high impedance (although 1 Meg would be nice to achieve).
The meter is reasonably accurate to 100 Khz, so I would also like that bandwidth, too.

Initially, I had wanted the amp to also be useful at DC, but if the offset proves to be a chore, I can always AC couple the output.
 

Offline iMo

  • Super Contributor
  • ***
  • Posts: 5570
  • Country: va
Re: For an opamp input, which network protection is preferable?
« Reply #11 on: April 18, 2019, 07:21:29 am »
Do you need the diodes actually? With 100k input resistor and say 5mA opapm's clamping diodes max current the input voltage could be around 1000Vpp.
Readers discretion is advised..
 
The following users thanked this post: Someone

Offline schmitt triggerTopic starter

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: For an opamp input, which network protection is preferable?
« Reply #12 on: April 18, 2019, 02:53:34 pm »
As I mentioned to Marco, I wasn't seeing the forest because I was focused on a tree.

But this discussion so far has been very enlightening, and I've learned a pair of new tricks which I wasn't aware before.
 

Offline Kalvin

  • Super Contributor
  • ***
  • Posts: 2145
  • Country: fi
  • Embedded SW/HW.
Re: For an opamp input, which network protection is preferable?
« Reply #13 on: April 18, 2019, 03:27:54 pm »
LEDs and diodes will produce some current (read:offset voltage into op amps inputs) when exposed to light, so use a good enclosure which will block light 100%.
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7453
  • Country: pl
Re: For an opamp input, which network protection is preferable?
« Reply #14 on: April 18, 2019, 05:42:19 pm »
I cannot have low voltage offset AND at least 1Meg input impedance AND overload protection AND wide bandwidth.
Perhaps some of TI's e-trim opamps? Picoamps of input current, okay input capacitance for a FET, microvolts of offset and offset thermal drift, reasonable noise, good CMRR up to 3V below VCC or so, rated for ±10mA input overload, a few MHz of GBW, not terribly expensive too. OPA191, OPA192, OPA376 are types I know.
 

Offline schmitt triggerTopic starter

  • Super Contributor
  • ***
  • Posts: 2431
  • Country: mx
Re: For an opamp input, which network protection is preferable?
« Reply #15 on: April 18, 2019, 08:34:26 pm »
To all of you, thanks for your valuable suggestions.

While I was researching the topic, I found this very interesting e-book from TI. Similar in tone to Analog's fabled "Analogue Dialogue" booklets.  :-+

While the better seasoned engineers will know most of the information, most of us will find very helpful tips to improve our circuits. To me it was chapter 28: The unused op amp: what to do? I always grounded the inputs, but that is probably not the best strategy.

http://www.ti.com/lit/ml/slyt701/slyt701.pdf
« Last Edit: April 18, 2019, 08:43:23 pm by schmitt trigger »
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: For an opamp input, which network protection is preferable?
« Reply #16 on: April 19, 2019, 10:14:59 am »
The standard protection network as shown in (A), series schottkys to V+ and V-, works fine. But schottkys have significant leakage, and because of the 1 Meg impedance, it could generate offset voltage errors.

Use low capacitance low leakage diodes like 2N3904 base-emitter (or base-collector for higher voltage) junctions and a lower input series resistance.  Bypass the input series resistance with 1000 picofarads to swamp the capacitance of the diodes and amplifier.

Quote
Therefore I was thinking along the lines of inverse parallel diodes from input to ground, as shown in (B). But while the reverse-bias capacitance is well specified, the capacitance below the forward bias is not. Using my Peak Atlas LCR meter, a pair of antiparallel 1N4148 show about 50 pF, which could cause loading at higher frequencies.

I don't know how much I can trust this reading.

LCR meters generally have trouble making this measurement without help.

Diode datasheets usually have a capacitance versus voltage graph.

Use two diodes in series to halve the capacitance.

Higher reverse voltages lower the capacitance.

Quote
I had read somewhere that LEDs, covered to avoid photoelectric effects, show both ultra-low leakage and minimum capacitance. Again using the Peak Atlas, 3 mm green LEDs measures 15 pF.

I do not know about the capacitance but LEDs often do make good low leakage diodes.
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 8995
  • Country: us
  • Retired, now restoring antique test equipment
Re: For an opamp input, which network protection is preferable?
« Reply #17 on: April 19, 2019, 05:21:53 pm »
Diode capacitance at or very near zero can be high.  Instead of returning each diode to ground, you can reverse bias it by connecting to another diode that is forward-biased with a small current; when the protection diode is forward-biased, its current adds to the quiescent current of the other diode and the clamp voltage is two diode drops.
I have had good luck with c-b junctions of low-noise audio transistors such as 2N2484 (npn) or 2N4250 (pnp).
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf