ADS1293: Lead Off detection without RLD and with leakage resistors

Hi,

Could you explain and clarify a bit more about -

"configuration which gives me signal which could after postprocessing with digital filters is usable"?

--------------------------------------------------------------

Regarding to "struggling with LOD. It just doesn't work." -

Are you using 8.3.18 Digital AC Lead-Off Detect?

Can you explain more about "It just doesn't work." and "but I just can't get it."?

 ---------------------------------------------------------

Have you tried LOD with an ECG simulator first? It's suggested to start with ECG simulator and figure out and come up with proper settings before moving on to the other source(s).

Do you configure the "flex routing switch" properly? and try different current amplitude and frequency(ACDIV_LOD and ACDIV_FACTOR) - Φ = 50/(4 × K × (ACDIV_LOD + 1)) kHz equation (12) at page 25.

And, note that in datasheet page 27 -

"

The lead connectivity can be determined in the digital domain by applying an FFT to the digital data and by measuring the amplitude of the tone at the AC LOD excitation frequency.

It should be noted that the digital AC LOD can only determine the series connectivity of the two leads attached to the inputs of a differential channel, and hence
the connectivity of the individual input pins can only be determined by the DC or the analog AC LOD.

"

--------------------------

Thanks

Hi,

>> Could you explain and clarify a bit more about -

>> "configuration which gives me signal which could after postprocessing with digital filters is usable"?

Since I am going without RL electrode (without RLD) there are basically 2 ways to deal with it. Either you 

put RLD out to some other electrode (LL or RA or LA) directly or to use large leakage resistors from RLD_out

to all 3 electrodes. First method  gave me unusable signal (50Hz noise and no ECG signal) at the electrode where

I returned RLD out, so I tried with second solution and got wandering ECG signal with 50Hz noise. If I apply some software

methods to deal with wandering signal and digital 50Hz notch filter I can get ECG signal that I need.

>> Have you tried LOD with an ECG simulator first? It's suggested to start with ECG simulator and figure out and

>> come up with proper settings before moving on to the other source(s).

I have tried DC LOD with ECG simulator and it works OK, but I didn't try Digital AC LOD with it.

All 3 of my electrodes are metal plates with little bit different shape and size (not my decision), so impedance of each of

them is slightly different. That is why I had to put some "compensation" components on signal line.

>>  Do you configure the "flex routing switch" properly? and try different current amplitude and frequency

>> (ACDIV_LOD and ACDIV_FACTOR) - Φ = 50/(4 × K × (ACDIV_LOD + 1)) kHz equation (12) at page 25.

I hope I did.

>> The lead connectivity can be determined in the digital domain by applying an FFT to the digital data and by measuring

>> the amplitude of the tone at the AC LOD excitation frequency.

>> It should be noted that the digital AC LOD can only determine the series connectivity of the two leads attached to the inputs

>> of a differential channel, and hence the connectivity of the individual input pins can only be determined by the DC or the analog AC LOD.

I do understand that with Digital AC LOD, I can't have exact info about which electrode is off and that only 2 by 2 electrodes could be tested.

The thing which is not quite clear to me is that do I have to do FFT of ECG output data and measure amplitude of AC LOD excitation frequency or this is done by ADS1293.

Hi,

Are you using 8.3.18 Digital AC Lead-Off Detect?

If yes, then Developer/Designers need to "apply an FFT to the digital data and by measuring the amplitude of the tone at the AC LOD excitation frequency" to determine the lead connectivity.

It's suggested to test the desired LOD method using the evaluation kit/board(EVM) with an ECG simulator first. Then, figure out and come up with proper settings or do FFT before moving on to other source(s).

Thanks