LME49720 DECT Cordless Phone EMI Susceptibility

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
Post Reply
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

LME49720 DECT Cordless Phone EMI Susceptibility

Post by mediatechnology »

Myself and some contributors at DIY Audio have noticed that the LME49720 and its variants are prone to EMI from DECT cordless phone base stations 100 Hz polling.

The following test circuit is low impedance (1-2 KΩ range) unity gain circuitry 8 feet from a DECT base station. Also nearby were a 4G phone, the DECT handsets, a WiFi access point and wireless weather instruments. The circuit board temporarily sits in a grounded metal tin. I removed the lid to illustrate the effect. The line frequency in these parts is 60 Hz.

To improve the FFT of the LME49720, only the DECT base station needed to be turned off.

The comparisons are similar when the LME49720 is compared to NE5532s, NJM2114s and other popular duals.
No other dual op amp I tested exhibited this rectification.

Image
EMI Susceptibility of OPA2134 vs LME49720 No DECT Base Station

The LME49720 is slightly quieter in the above owing to its lower voltage noise.

When the DECT 6.0 base station is turned on all hell breaks loose in the LME49720 channel...

Image
EMI Susceptibility of OPA2134 vs LME49720 DECT Base Station Powered

DECT transmits a "dummy bearer" signal to the headset every 10 ms.
The LME49720 exhibits rectification of the 100 Hz pulsed carrier.

I wanted to get this documented.
Your milage may vary.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: LME49720 DECT Cordless Phone EMI Susceptibility

Post by JR. »

I haven't been paying close attention to the modern uber op amps but that symptom looks like input stage rectification. Perhaps not the LTP but some new improved associated circuitry.

Back in the good old days, they would publish pseudo schematics of the innards. The 45 page data sheet at TI had a kitchen sink of application notes but no innards.

Perhaps some input stage bias current cancellation (guess), but something with diode junctions.

Does a simple LPF on input lines help?

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: LME49720 DECT Cordless Phone EMI Susceptibility

Post by mediatechnology »

JR. wrote: Sun Feb 18, 2018 3:04 pm Perhaps some input stage bias current cancellation (guess), but something with diode junctions.

Does a simple LPF on input lines help?

JR
An input LPF does not help.
I've wondered if it is either the bias-current cancellation or ESD diodes.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: LME49720 DECT Cordless Phone EMI Susceptibility

Post by JR. »

ESD diodes and body diodes are probably biased off the rails so unlikely to be involved at normal operating conditions.

Input LTP tricks OTOH could be in the mix.... :oops:

Some of these uber puppies have specs too good to believe and require gimmick circuits to even measure... :roll: Can't say I really trust that (it could be gamed). I had some conversations off list with Sam Groner and he did some tests of input stage CMRR that were pretty respectable (probably not that exact op amp, just one of the sundry ubers).

JR
Cancel the "cancel culture", do not support mob hatred.
diffuse371
Posts: 10
Joined: Thu Mar 10, 2016 6:48 am

Re: LME49720 DECT Cordless Phone EMI Susceptibility

Post by diffuse371 »

I wonder how frequency sensitive this is? It would be interesting to see how this plot changes with a different spectral pattern, it may just be that the internal geometry of the opamp has some structural resonance at 100Hz and it's harmonics. Whenever I see the die footprints for many modern ICs they are all symmetric squares.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: LME49720 DECT Cordless Phone EMI Susceptibility

Post by JR. »

isn't 100Hz just the repetition rate? Perhaps modulating the RFI at that lower F (like IMD?) , but it's probably the RF causing distress (another guess).

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: LME49720 DECT Cordless Phone EMI Susceptibility

Post by mediatechnology »

100 Hz is the repetition rate of the 1.9 GHz DECT 6 carrier in the idle link "dummy bearer" mode.

According to a post on GroupDIY ESD diodes were found to be resonant in another manufactures design and they had to re-do the mask.

I think its resonant based on a combination of lead frame/lead wire inductance, junction capacitance and a junction. At that frequency the diode junctions may not necessarily point in the same direction the schematic would indicate. Which junction it is is academic given its inability to be fixed externally.

As a practical matter its tuned to DECT 6 and nothing myself or a handful of other people have tried - other than shielding the package - fixes it.
It is not sensitive - to a great extent - to cellular or WiFi signals.
I don't have a network analyzer to check it.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: LME49720 DECT Cordless Phone EMI Susceptibility

Post by JR. »

Semiconductor physics in IC design is already hard enough without RF concerns. Somewhat above my frequency range of expertise. I expect some stand alone diodes could suffer a similar fate?

JR

PS: It is funny when reading old texts about wiring interference that modern RF was not even in their crystal balls.
Cancel the "cancel culture", do not support mob hatred.
Post Reply