Gate stoppers are for FETs, grid stoppers are for tubes. Stability, RF/EMI and all that jazz.
I have no idea why someone came up with the notion that opamps need gate stoppers.
I agree with you it's a bit non-sensical.
But I'm doing this only 50 years, so I may have missed a lot 😎
Jan
I have no idea why someone came up with the notion that opamps need gate stoppers.
I agree with you it's a bit non-sensical.
But I'm doing this only 50 years, so I may have missed a lot 😎
Jan
Last edited:
JRA needed an RC damper at the input of an OPA1656 phono amplifier to get it stable, https://www.diyaudio.com/community/...split-from-opa1656-thread.377331/post-6785883
I think JRA's problem was related to the overall feedback loop rather than to local parasitic Colpitt-like oscillations that you normally use gate, base or grid stoppers for.
I think JRA's problem was related to the overall feedback loop rather than to local parasitic Colpitt-like oscillations that you normally use gate, base or grid stoppers for.
Ha, ha. Too many threads here end like a ‘broken telephone game’ where initial message is changed with every next post.
- lineup asked only do we need input resistors with CMOS opamps, like is usually seen with FET ones. Their purpose was not named.
- Osvaldo da Banfield named them stoppers, assuming their function.
- I pointed out that we need to limit input current because of ESD protection diodes.
- tomchr elaborated on ESD protection.
- JMFahey expanded with example of protection for hard working conditions.
I own a portable shortwave radio that is infamous for its input JFET being blown out by ESD to the antenna.
Even the high-gain bipolar transistors used in audio have a base current rating of only 10mA.
I would say that ESD protection is a good idea for external inputs.
Internally, series resistance is not needed.
Ed
Even the high-gain bipolar transistors used in audio have a base current rating of only 10mA.
I would say that ESD protection is a good idea for external inputs.
Internally, series resistance is not needed.
Ed
I insist on using stoppers only when needed. I never use them from regulated PSU switching or linear to complex Phantastron circuits unless it is strictly necessary.
Before mass production starts, a couple of samples are subjected to worse abuse than that:
During latch-up testing, usually 100 mA is forced through each input for a couple of milliseconds while the chip is at the maximum allowable temperature.
A current of 1.33 A that decays exponentially with a 150 ns time constant is forced through the inputs and outputs with the chip at room temperature during human body model ESD testing.
A couple of amperes flow during a few nanoseconds during charged device model ESD testing.
Does anyone still use machine model? I haven't seen any machine model requirements in the last 15 years.
That's a good question. It's possible that MM has fallen out of favour and CDM taken over. I've been out of that game for a while and haven't kept up.
Tom
Tom
Did the thread not start saying "CMOS OpAmps gate stoppers?"
The ones I've seen and used are to prevent spurious oscillation (at RF frequencies and causing RFI) that I've seen happen in a power MOSFET circuit, due to input capacitance and I presume spurious inductance and such. I saw this in a data sheet decades ago and added 470 ohm resistors right at the gate terminals, and that fixed it. I presume the name came from stopping the device from oscillating. These had nothing to do with preventing gate damage, as others in the thread were discussing for other uses of series resistors. With static inputs that are likely to cause damage, 470 ohms would likely not be enough to prevent it.
Looking only at the content of first post, and forgetting the title, I supported my ‘broken telephone game’ claim by example. Did I not? 😆
We are her for hobby and every word is not carefully measured with preparations adequate for a dissertation.
Most resistors at the opamp inputs are there for other reasons than as stoppers. I’m fine with one resistor providing several functions.
We are her for hobby and every word is not carefully measured with preparations adequate for a dissertation.
Most resistors at the opamp inputs are there for other reasons than as stoppers. I’m fine with one resistor providing several functions.
And especially: never take any measurements. Even a single measurement can ruin the audiophile design.
The measured performance also improves when you leave out RF filtering and external ESD protection, as long as there is no RF interference (or ESD) in the measurement set-up and the circuit stays stable: flatter frequency response, less phase shift, less noise...
The problem is the Miller Effect. In a single MOSFET it toghether to parasitic inductance and large di/dt may cause ringing of such value to destroy the thin gate isolation. But OPAMPS has very little Miller capacitance, very little current in them to cause them oscillate or breack down of input gates.
Utopia/lab conditions don't guarantee normal expected operation at home which is IMHO what really counts. Still many here seem to like such designs and their peculiarities.
