Signal Input Protection Diodes

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Newmarket

Well-known member
Joined
Oct 10, 2016
Messages
2,031
Location
Brighton Sussex UK
Hi all. Chewing over Signal Input Protection diodes here.
Looking at BAT85 type Schottkys vs 1N914 etc but in SMT.
Recall using jfets in diode config previously for ultra low leakage. also possibility to use BJT B-C diode connection.
Leakage Current / Capacitance  comparisons ? Thoughts / experience to help me sift through the datasheets.
Thanks.
 
What kind if input are you trying to protect? Many modern ICs have protection built in.

THAT corp has a good white paper about clamping mic preamp inputs that can suffer amps of surge current from charged phantom blocking capacitors.

JR
 
In this particular case it's a 'standard' ac coupled Non-Inverting OpAmp stage. Nominally NE5532 but could be changed to any other 'standard' pinout Dual Op Amp - so LM4562 / OPA2134 etc etc.
Yes there'll often be diodes built into the inputs and a bit of input resistance before them will handle most things.
But I'm aiming for something 'universal' so belts and braces...

The THAT notes are good stuff. I'll search out the paper you mention.

Thanks.
 
As long as you are not using crazy high impedance values, simple small signal clamp diodes to + and - rails on any line exposed to the outside world should get it done. Modern op amps may have protection build it.

I recall talking with an engineer at an IC company about tweaking one of their standard parts to do a different trick and they declined because even though they could make the mod with a relatively inexpensive metalization layer, the protection circuitry was far more involved and  the design change would leave one input pin unprotected.

JR
 
JohnRoberts said:
As long as you are not using crazy high impedance values, simple small signal clamp diodes to + and - rails on any line exposed to the outside world should get it done. Modern op amps may have protection build it.
JR

Yes - that's the simple diode scheme I'm looking at.
I recall using BAT85 and similar Schottky types on ADC inputs and possibly jfets in diode config for low leakage. But these wouldn't have been facing the 'outside world'.
For my own stuff I don't usually have additional protection beyond say 100R resistor that is primarily there as part of a passive RC low pass for rf.
But as I'm thinking of running off a pcb (tired of fiddling about with strpboard  ;D) I thought I might allow for it.
So really just wondering what views people have on best types re leakage / capacitance etc. Same considerations as when used for output protection I guess where phantom issues more likely.
I haven't had chance to look up the THAT stuff yet.
 
L´Andratté said:
aren´t ne5532/34 protected internally?

I think you might be correct though not sure.
IIRC they do have diodes between the inputs (so don't try and use them as comparators). Not sure about protection for input voltages beyond the rails.
But in any case I was really asking about protection diodes in general - could be any sort of input stage - and wondering what people here thought wrt leakage / reverse linearity etc...
 
I use SMBJ15CA which are 15V TVS diodes to protect inputs (and outputs) from high voltage spikes.

And I use ES1D where a conventional diode might be used like on regulator boards.

They're both DO-214A surface mount packages which makes for a cleaner build I think. Normally I wouldn't use SMD for any signal lines but for some things I don't think it adds distortion (unlike smaller SMD resistors which I believe do introduce distortion).

Also, for completeness, I use 74279215 inductors and 100p caps on pins 2,3 to shunt RF to pin 1 but they have to be very close to the pins. These are tiny 0805 SMD but they are good for 3A. Arguably I could go with larger and reduce the current to get down into the 10's of MHz range.

So I have inductor -> shunt caps -> TVS diodes on all inputs and outputs (minus inputs with phantom power because the TVS diodes would just short the 48V supply).
 
JohnRoberts said:
... simple small signal clamp diodes to + and - rails on any line exposed to the outside world should get it done ...
better to the + and - supply pins of the opamp to protect, so input voltage will not exceed the opamps supply voltage. Depending on opamp used, the often seen resistors between supply rails and the opamp for current limiting or as part of a R/C filter will cause a voltage drop that the clamped input to the rails otherwise might exceed.
 
squarewave said:
I use SMBJ15CA which are 15V TVS diodes to protect inputs (and outputs) from high voltage spikes.

And I use ES1D where a conventional diode might be used like on regulator boards.

They're both DO-214A surface mount packages which makes for a cleaner build I think. Normally I wouldn't use SMD for any signal lines but for some things I don't think it adds distortion (unlike smaller SMD resistors which I believe do introduce distortion).

Also, for completeness, I use 74279215 inductors and 100p caps on pins 2,3 to shunt RF to pin 1 but they have to be very close to the pins. These are tiny 0805 SMD but they are good for 3A. Arguably I could go with larger and reduce the current to get down into the 10's of MHz range.

So I have inductor -> shunt caps -> TVS diodes on all inputs and outputs (minus inputs with phantom power because the TVS diodes would just short the 48V supply).

Yes - that's interesting re TVS type devices esp if you are thinking about high voltage transients.
Any thoughts on their capacitance esp voltage coefficient ?

I can't agree about SMD resistors and distortion  - unless talking about 'thick film' types and not metal film types (MELF). I guess thin film types would be okay too but I haven't used them.
All else being equal physically larger resistors (ie the actual resistor itself - not attached leads / packaging) will be less noisy than a physically smaller resistor but the effect is really marginal - I've never been able to measure it and it's not something I really consider in my design work (non audio precision industrial instrumentation stuff)
 
Harpo said:
better to the + and - supply pins of the opamp to protect, so input voltage will not exceed the opamps supply voltage. Depending on opamp used, the often seen resistors between supply rails and the opamp for current limiting or as part of a R/C filter will cause a voltage drop that the clamped input to the rails otherwise might exceed.

That's correct if there are resistors in the voltage lines.
However, it can become impractical if you have resistors for each op amp package and several op amp packages. You then need to have diodes at every op amp to be 'correct' as each op amp will, in general, have different voltage drops associated with it.
The idea, from my point of view is to have the protection at one point only - where the signal enters the circuit.

Looking at the maths the if we take the current to be, say, 10mA and there is a 10 Ohm resistor then the drop is only 100mV.
Let's say we use a standard diode, it might clamp the overvoltage to, say, +0.5V over the rail. So this gives +0.6V wrt the Op Amp.
Not much difference and using a small signal Schottky diode would reduce this to, say,+0.4V.
All figures indicative only but in the right ballpark.
 
Ahh - I remembered what I was thinking about with diode connected FETs and low leakage.
Some moons ago I designed a very (I mean VERY) sensitive I to V converter for surface science instrumentation.
Input bias of the op amp (OPA128) is spec'd at 75fA !
Obviously anything else which diverted current  from the feedback resistor was an issue.

Anyway, not directly an audio application but interesting nevertheless.
Basically jfet - 2N4117A - diode connected.
Link to TI (Burr Brown) note below.

http://www.ti.com/lit/an/sboa058/sboa058.pdf
 
Newmarket said:
Ahh - I remembered what I was thinking about with diode connected FETs and low leakage.
Some moons ago I designed a very (I mean VERY) sensitive I to V converter for surface science instrumentation.
Input bias of the op amp (OPA128) is spec'd at 75fA !
Obviously anything else which diverted current  from the feedback resistor was an issue.

Anyway, not directly an audio application but interesting nevertheless.
Basically jfet - 2N4117A - diode connected.
Link to TI (Burr Brown) note below.

http://www.ti.com/lit/an/sboa058/sboa058.pdf
Yes back to back (anti-parallel) diodes to ground  from the - input is simple and effective with inverting amps, BUT you said non-inverting in your original post.

Also using JFET gate diodes as low leakage diodes is another old trick if the cost is acceptable.

JR
 
Newmarket said:
Yes - that's interesting re TVS type devices esp if you are thinking about high voltage transients.
Any thoughts on their capacitance esp voltage coefficient ?
Ick! They have super high capacitance. I measured 625p. Strangely most of the SMBJ15CA datasheets don't mention anything about capacitance and I chose the part because it was used in the schematic for an eval board for some fancy audio chip. Only the OnSemi datasheet has a graph of capacitance vs reverse voltage. I guess modern designs assume everything is being driven by low Z. The source Z would have to be in the 5K ohms range to start seeing distortion though. I suppose I can keep using them for the moment. It seems there are TVS diodes with much lower capacitance (like CDSOD323-T15C which is only ~3pF).

Attaching diodes to supply rails is a PITA. It would be nice if I could find a simple 2 terminal part.

Thanks for pointing this out. I would like to interface with high Z gear.
 
squarewave said:
Thanks for pointing this out. I would like to interface with high Z gear.
I don't think the damage is caused by low z gear... more likely static hits, and/or energized grounds (more likely in live SR than studio). Where there is the occasional inexperienced customer feeding an amplifier output into a line input (the early CS power amps would actually allow that, but most modern gear not so much). 

Last century while working at Peavey we did a review of where the majority of IC failures occurred, when we were considering removing sockets from the PCB fab (old through hole days).  It turned out the vast majority of IC failures occurred at inputs and outputs where ICs touched the outside world.

Some ICs were literally vaporized, probably from energized grounds.

JR
 
squarewave said:
Attaching diodes to supply rails is a PITA. It would be nice if I could find a simple 2 terminal part.

You need three terminals: positive rail, negative rail (which might be ground) and signal.
 
Andy Peters said:
You need three terminals: positive rail, negative rail (which might be ground) and signal.
I'm not sure I understand. What's wrong with a 2 terminal bidirectional TVS diode to shunt signal to chassis ground?
 
abbey road d enfer said:
What about two 15V Zeners in series back-to-back?
I think a bi-directional TVS diode is essentially like back-to-back zeners. They have a reverse breakdown voltage.

What about this part?:

CDSOD323-T15C.png


CDSOD323-T15C

Reverse Breakdown: 16.7V
Junction Capacitance: ~3pF

It's a bidirectional TVS diode but I guess the extra regular diodes in series make it low capacitance? I don't know anything about this part. I just found it on Mouser looking for low capacitance TVS diodes.

Diverting into power supply rails doesn't make as much sense to me. I think it makes the most sense to place protection diodes right next to pins (between 2 and 1 and 3 and 1) and then use the XLR sockets that have the metal spike that connects pin 1 to the chassis so that shocks or whatever transients will be shunted directly into the chassis without getting that far inside the enclosure.
 

Latest posts

Back
Top