• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Sensitivity of opamps to air coupled EM fields, especially of the LM4562/LME497X0 family

JensH

Member
Joined
Jan 5, 2020
Messages
38
Likes
52
What in the DECT phone causes the 100Hz? I just did a quick test with my Panasonic KX-TG9541B phone...
It is the transmission from the antennas. DECT is a TDMA (Time Division Multiple Access), so the transmitter is only transmitting in short bursts, then shuts down and after a while the unit will receive from the other end. The base station will transmit every 10 ms. When a call has been established the base station will receiver from the handset 5 ms after the transmit burst. Semiconductors (e.g. op-amps) will in many cases rectify the modulated RF signal and generate noise at 100 Hz plus harmonics.

Was a call (connection) set up between the base station and the handset when you made the test? If no call is active, the handset will only receive and thereby it will not create interference. (There are a few exceptions to this, but most of thte time the handset will not transmit anything when idle).

The OPA1612 is more robust against EMI than the LME49720. On some of the newr parts TI has released measurements of the EMI immunity, e.g. the OPA209 family.
 
OP
pma

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
I have also found sensitivity of some circuits to timing signal bursts (with low frequency repetition) that are distributed into the home mains line. Don't you speak about the same thing?
 

JensH

Member
Joined
Jan 5, 2020
Messages
38
Likes
52
I am not quite sure what you mean by "timing signal bursts"?
 
OP
pma

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague

JensH

Member
Joined
Jan 5, 2020
Messages
38
Likes
52
At those low frequencies I would expect the power supply to deal with it and deliver a clean supply. It will not generate any significant airborne radio transmission.
I was thinking of radio signals e.g. at 1.9 GHz (DECT), 2.4 GHz (Wi-Fi) or 5.x GHz (Wi-Fi). Especially TDMA systems are often a problem, switching the transmitter ON and OFF many times per second. When rectified by a nonlinear circuit, it will generate a pulsating voltage level, which corresponds to audible noise.
 
OP
pma

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
At those low frequencies I would expect the power supply to deal with it and deliver a clean supply. It will not generate any significant airborne radio transmission.
I was thinking of radio signals e.g. at 1.9 GHz (DECT), 2.4 GHz (Wi-Fi) or 5.x GHz (Wi-Fi). Especially TDMA systems are often a problem, switching the transmitter ON and OFF many times per second. When rectified by a nonlinear circuit, it will generate a pulsating voltage level, which corresponds to audible noise.

I see. But the signal I speak about has low repetitive frequency as mentioned, but the signal itself are high frequency bursts distributed through the mains network. Maybe I will find some old measurements. In Czech it is called HDO, however I cannot find a direct English equivalent.

https://cs.wikipedia.org/wiki/HDO
 

JensH

Member
Joined
Jan 5, 2020
Messages
38
Likes
52
As far as I can see (using Google translate) the highest carrier frequency is 1060 Hz in the Czech Republic, so not really RF. In terms of e.g the LME49720 I would not worry.

But I could see some potential issues if the carrier propagates through the power supply. With the frequencies involved they will be more annoying than the usual 50 Hz, but of course the amplitudes are lower. I have no experience with this.
 
OP
pma

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
As far as I can see (using Google translate) the highest carrier frequency is 1060 Hz in the Czech Republic, so not really RF.

The carrier frequency is low, but the bursts have high frequency. This is the record

1578255457729.png


But let's leave it for now. Specifically this was not about the LME49720. LME49720 catches EM field even if it is battery supplied, as I explained in post #1.
 

stereo coffee

Active Member
Audio Company
Joined
Jan 2, 2019
Messages
143
Likes
44
Having read presently about 1/10th of Hendrik Lorentz first paper , it may not be traditionally thought of EM fields, but light. I would suggest to conduct some tests with the DUT in the dark, to discern any differences.
 
Last edited:

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,749
Location
Alfred, NY
Having read presently about 1/10th of Hendrik Lorentz first paper , it may not be EM fields but light. I would suggest to conduct some tests with the DUT in the dark, to discern any differences.

Here's your sign.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
Having read presently about 1/10th of Hendrik Lorentz first paper , it may not be EM fields but light. I would suggest to conduct some tests with the DUT in the dark, to discern any differences.
Modern tiny packages (usually BGA) with exposed wafer edges can be sensitive to light. SOIC (and, shudder, DIP) should be encapsulated well enough not to have any such issues.
 

Arpiben

Member
Joined
Dec 1, 2019
Messages
38
Likes
37
The carrier frequency is low, but the bursts have high frequency. This is the record

View attachment 44734

But let's leave it for now. Specifically this was not about the LME49720. LME49720 catches EM field even if it is battery supplied, as I explained in post #1.

Later you may have a look at Power Line Communication standards like IEEE 1901 /HomePlug AV,GP,AV2....( Home networking through power lines ).
I have no specific knowledge about it but it seems they may also radiate.
 

ajawamnet

Active Member
Joined
Aug 9, 2019
Messages
288
Likes
460
It is the transmission from the antennas. DECT is a TDMA (Time Division Multiple Access), so the transmitter is only transmitting in short bursts, then shuts down and after a while the unit will receive from the other end. The base station will transmit every 10 ms. When a call has been established the base station will receiver from the handset 5 ms after the transmit burst. Semiconductors (e.g. op-amps) will in many cases rectify the modulated RF signal and generate noise at 100 Hz plus harmonics.

Was a call (connection) set up between the base station and the handset when you made the test? If no call is active, the handset will only receive and thereby it will not create interference. (There are a few exceptions to this, but most of thte time the handset will not transmit anything when idle).

The OPA1612 is more robust against EMI than the LME49720. On some of the newr parts TI has released measurements of the EMI immunity, e.g. the OPA209 family.

Yep - tried calling an outside line as well as calling the second line on the same phone (makes Jimi Hendrix noises in conf mode). I have an old Bell Princess on the wall and tried that too. I have a Teltone phone line simulator (4 lines ) that I got way back when I was doing this - http://www.ajawamnet.com/amnet/index.html.
I couldn't get the QA401 to show anything at 10 or 100 Hz... even with the cover off and the open test lead. Doesn't look like he's doing anything special on the front end ... dunno...
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,581
Likes
38,283
Location
Gold Coast, Queensland, Australia
Sorry I could not find a proper term. It is a remote control via power line, here they use 183.33 Hz, 216.66 Hz, 283.33 Hz, 760 Hz and 1060 Hz.

It's called ripple control over here. Used for all sorts of stuff ranging from our hot water heater peak demand control, soon to be air-conditioning control at peak times and also some remote operations of distribution infrastructure. I am seeing "smart" (non wireless or cellular) control units being installed on pole mounted street transformers- my guess is they are coded and it gives the power distributor the capability to individually switch on or off subsections of the grid, even on a street by street basis- all over the line.

The ripple control pulses are easily heard on some HiFi gear if you listen around 5:30pm on weekdays (just before peak demand) and run for around 20-30 seconds. If you happen to be using a CRO or DSO when they activate it's a real mess. I'll try to catch some if I get a chance.
 
Last edited:

ajawamnet

Active Member
Joined
Aug 9, 2019
Messages
288
Likes
460
Here's your sign.
Most semi's are photosensitive... From https://en.wikipedia.org/wiki/Photodiode
"Unwanted photodiode effects
Any p–n junction, if illuminated, is potentially a photodiode. Semiconductor devices such as diodes, transistors and ICs contain p–n junctions, and will not function correctly if they are illuminated by unwanted electromagnetic radiation (light) of wavelength suitable to produce a photocurrent;[10][11] this is avoided by encapsulating devices in opaque housings. If these housings are not completely opaque to high-energy radiation (ultraviolet, X-rays, gamma rays), diodes, transistors and ICs can malfunction[12] due to induced photo-currents. Background radiation from the packaging is also significant.[13] Radiation hardening mitigates these effects."


They can also generate photons via hot-carrier emission - from IBM:
The paper : https://pdfs.semanticscholar.org/2ed9/16ec785f3aeb716cbe757e6daa894cf27bda.pdf
The patent: https://patents.google.com/patent/US6469529
news article from years ago : https://www.chicagotribune.com/news/ct-xpm-2000-03-06-0003060049-story.html

Another paper - http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.555.404&rep=rep1&type=pdf

the video that was published years ago - this is the clock buffer, repeaters and distribution to latches of the IBM S/390* processor - this is explained in the paper linked above on page 17 of the PDF...
 
Last edited:

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,749
Location
Alfred, NY
"Opaque housings." Note that if you illuminate the innards of any actual piece of engineered audio gear, it has zero effect. I even did this to a forward biased LED in response to some online fretting about noise pickup (I used the LED to bias a phono stage). And even the light from a xenon strobe had no effect.

So this is as irrelevant to actual audio equipment as the "test circuit" used here.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
There's nothing wrong with Pavel's test setup. It's just what's being tested and what conclusions are drawn from it that may be the source of disagreement. Remember, the test circuit (which he was very explicit about) is not at all representative of anything that would be used in an audio gain stage. So the results are correct for that circuit, but are irrelevant to circuits used in engineered audio equipment.
Sorry I wasn't being clear.

When I say "wrong" I include "inappropriate and irrelevant". However there may well be issues with the test set up that exacerbate the problems shown, board layouts or other. Also, as I mentioned right at the start, there is no quantification of the EM fields, or demonstration of cause and effect.

This site is about audio, it's not about "can I create pathological conditions that demonstrate a problem".

This sort of exercise just creates misleading FUD, and seems a repeat of the amplifier load thread.
 
Last edited:

stereo coffee

Active Member
Audio Company
Joined
Jan 2, 2019
Messages
143
Likes
44
Most semi's are photosensitive... From https://en.wikipedia.org/wiki/Photodiode
"Unwanted photodiode effects
Any p–n junction, if illuminated, is potentially a photodiode. Semiconductor devices such as diodes, transistors and ICs contain p–n junctions, and will not function correctly if they are illuminated by unwanted electromagnetic radiation (light) of wavelength suitable to produce a photocurrent;[10][11] this is avoided by encapsulating devices in opaque housings. If these housings are not completely opaque to high-energy radiation (ultraviolet, X-rays, gamma rays), diodes, transistors and ICs can malfunction[12] due to induced photo-currents. Background radiation from the packaging is also significant.[13] Radiation hardening mitigates these effects."


They can also generate photons via hot-carrier emission - from IBM:
The paper : https://pdfs.semanticscholar.org/2ed9/16ec785f3aeb716cbe757e6daa894cf27bda.pdf
The patent: https://patents.google.com/patent/US6469529
news article from years ago : https://www.chicagotribune.com/news/ct-xpm-2000-03-06-0003060049-story.html

Another paper - http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.555.404&rep=rep1&type=pdf

the video that was published years ago - this is the clock buffer, repeaters and distribution to latches of the IBM S/390* processor - this is explained in the paper linked above on page 17 of the PDF...

If photons can be generated, can the same photons emitting also then create a pathway allowing external light to enter at that same moment. ?

In SIY's argument, perhaps "any piece of actual engineered audio gear" was not the type that was emitting photons, or measurements conducted at that time were not specifically looking for photon emission and as a result of those emissions, any effects being caused.
 

stereo coffee

Active Member
Audio Company
Joined
Jan 2, 2019
Messages
143
Likes
44
For those who may be interested, -the prologue of Lorenz paper below translated from its Dutch language
https://en.wikipedia.org/wiki/Hendrik_Lorentz

https://ilorentz.org/history/proefschriften/sources/Lorentz_1875.pdf

"With a theoretical investigation of the movement symptoms der electricity came maxwell 1) the remarkable result,that there is a large agreement between sum some electrical movements and the aether vibrations that one to explain the light phenomena. He found,
that in a dielectric body there are vibrating movements of the elec can take place in the same way and with propagate the same speed as the light vibrations. The correspond The mood is so great that Maxwell gave rise to the under that in reality the light would consist of such electrical vibrations, a hypothesis, which he for some cases to the experience and therefore to a certain extent found.


Regardless of maxwell and others, lorenz is also 2) came to such results. Later, Helmholtz 3) joined the compilation of the movement equations of electricity the most general expression used for the inducing action of may incorporate power elements while undergoing maxwell and lorenz special assumptions about that operation were made. Helmholtz now found that also with that more alga for the consideration of the subject obtained by maxwell result essentially remains. He also commented to a circumstance that was not discussed by maxwell. He says namely 1) of the agreement found by them:This analogy is not very much in any other important relationship before hands which sir maxwell has not touched. You have the mecha niche state of the light ether some media so far see the fixed nel as ti Body equal set. This assumption results but for the border of two more transparent Media other boundary conditions, as a man,around the refraction and reflection of light to explain this limit, so here
theorists optics an unsolved contradiction files. The theory of electricians However, oscillations do not only result in the inside a similar insulating medium, but also on the border of two such media, the same laws of reproduction, the refraction and reflection of the waves, who we Light that find, provided that one can use either the magnetic or the dielectric polarity capability of both media the same and the latter is very large.

Of the designated alternative it depends on whether the electrical or magnetic oscillations of a polarized beam in the plane of polarization. "
It was this comment from Helmholtz that gave me cause to investigate to what extent the phenomena of reflection and refraction of light give rise to Maxwell's theory preferable to the previously adopted undulation theory.


In the first chapter I have the. difficulties discussed, which the latter yields in this regard. After then in the second chapter the movement equations of electricity I have the hypo in the last four chapters thesis by maxwell, or the electromagnetic theory of the light. In addition, the
reflection and refraction of light by non-conductive, isotropic substances, the optical properties of crystals, the theory of total reflection and finally the reflection of the light metals.

As the above-mentioned words from Helmholtz show,is a part of the results that I have obtained, already before
distracted by him. However, I do not know that he is more has written the quoted about this subject.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
I did see these at low frequency when I was measuring various opamps in o2 one and a half year ago. I thought the circuit was being unstable. The level was around -130dbV. Which is quite high actually for something that's not supposed to be in a circuit.
The truth to this is: The tests perform under different conditions in different environments. It happens sometime, other times it's not there. Just as proposed previously. Do I think this is a big issue? Not really. But big enough for me to not using it for high performance application or somewhere that can be affected by interference. The reason is probably the pin layout of the silicon. As many other opamps do not suffer from this, I guess they must have done something different from what's being normally done. Can also be something to do with the process.
So lm4562 is NOT a bad chip, BUT this IS something to consider when designing using lm4562.
 
Top Bottom