ADS1298 - Lead-off Detection

**TI EDIT February 20th 2019; FOR GENERAL ADS129x ASSISTANCE, PLEASE SEE NEW BIOPOTENTIAL FAQ LOCATED HERE**

Hi Horizon,

I'm trying to put together a better description of the lead off detection based on DC versus AC modes of the ADS1298.  I'll try to have something posted for you on our ADS1298 FAQ page later today or early tomorrow.

Thank you Tom, will be looking forward to it.

Thank you for the neat description Tom. You did tickle my brain and now I have a couple of more questions!

Limitation of AC Lead-off Detection

Lets say, while in AC lead-off detection mode, my RA electrode is dislodged. This would cause the excitation signal to appear on both Lead I and Lead II since Lead I = LA - RA and Lead II = LL - RA. The situation will be similar when either or both of LA and LL are also left floating. Hence it would be difficult if not impossible to judge which all leads have come out. Similarly, the removal of any combination of LA, RA or LL would cause the excitation signal to appear on all the chest leads! Theoretically it may be possible to identify the electrodes that have been pulled our by measuring the p-p voltage of the excitation signal that has been extracted from each lead. However this voltage is also dependent on the electrode-skin impedance which cannot be guaranteed to remain a constant for even the same subject. Hence AC lead of may be usable only to warn the removal of any electrode.

Q: Is this limitation true or can AC method me used to precisely identify which all electrodes are pulled out (except RL electrode of course)?

Using DC Methods while Acquiring

With the feared limitation of AC lead-off detection, how good are the DC methods (current or pull-up.down) while acquisition is active? I remember this from somewhere that the DC methods cannot be used while acquisition is active. Please see the piece of schematic attached (derived from Application Report - SPRAB36A – July 2009 - ECG Implementation on the TMS320VC5505 DSP Medical Development Kit (MDK)).

Due to continuous feedback, the RL drive establishes a potential of 0 (theoretically) at the point where the electrodes are placed. Since both comparator and ADC have relatively huge input impedance, the pull-up resistor and on-board resistance can be viewed as part of a voltage divider network. Hence when the electrode is connected, the voltage at the mid-point of these two resistors would be 5mV for AVDD=2.5V and given value for resistors. This falls well within the common mode as well as differential mode of the ADC. Now when the electrode is off, there is no particular current path except the infinitesimally small leakage currents, and close to full AVDD appears at both comparator and ADC inputs.

Q: Is this theory true in case of ADS1298 with DC pull-up/down or current source methods of lead-off detection so as to use them while acquiring, or are these modes available only when the ADC is not acquiring?

Mohith,

We have customers currently using AC or DC lead off methods to continuously determine which leads are off.  There are differences between the lead off detection of the discrete ADS1258 solution and the fully integrated solution found in the ADS1298.   Have you by chance gotten one of the ADS1298ECGFE-PDK setups and a simulator?   We'll try to put together a little tutorial explaining the process...

Tom,

I haven't had a chance to get dirty with the PDK but I am making my own test board with help from PDK schematics. Should have the results in a couple of weeks. Currently my plan is to go ahead with current source mode for the better input impedance it gives. Since it doesn't require any external components to switch between the three modes, I am hoping to study in detail about all three using the same board.

Hi I have one question, I am using the following configuration.

All channels are in input shorted, 

And I am using LEAD OFF detection

LOFF=0x13 DC pull up/down

CONFIG4= 0x02

LOFF_SENSP=0XFF

LOFF_SENSN=0XFF

So all channels are preaperd for lead off detection, all channels are disconnected.

The question is the following the status output will be something like this 1100 + 11111111 + 1111111 GPIO, and I dont have that output, tha values changes so much.

I think maybe is the noise at the input.

Another Qestion  LEAD OFF detection  is only for used in electrode mode? 

Thanks a lot.

Hi Oscar,

We are in the final draft stages of writing an app note explaining AC/DC lead off and uses with the ADS1298. In the meantime, I recommend giving this post a read -

http://e2e.ti.com/support/data_converters/precision_data_converters/f/73/t/124326.aspx#464964

It is important that the RLD electrode is connected during lead off testing. Otherwise, the result will be a poorly defined common mode which may cause DC lead off to behave incorrectly.

To answer your last question - DC LEAD OFF detection used with the comparators is run independent of the channel mux. Meaning that you can have the mux for all the channels set to test signal and the comparator bits in the STATUS BYTE will still update depending on the status of the inputs. I guess your answer would be 'No, LEAD OFF works regardless of what mode the input channels are set to"

-Tony