ADS1292R: lead-off dectection insensitive

Bruce li

Part Number: ADS1292R
Other Parts Discussed in Thread: ADS1292

Dear TI Experts,

My customer is evaluating ADS1292R in a new project, they can hardly detect lead-off, could you kindly help to review below HW and SW and give some advice for this case? Thanks a lot.

FW initiation configuration：

ADS1292_SendCommand(ADS1292_CMD_SDATAC);

ADS1292_WriteRegister(ADS1292_REG_CONFIG1, 0x01); // continuous conversion mode,sample rate to 250sps

ADS1292_WriteRegister(ADS1292_REG_CONFIG2, 0xe0);//Lead-off comparators enabled

ADS1292_WriteRegister(ADS1292_REG_LOFF, 0xd0);//75% - 25%

ADS1292_WriteRegister(ADS1292_REG_CH1SET, 0x60);//pga = 12

ADS1292_WriteRegister(ADS1292_REG_CH2SET, 0x60);//pga = 12

ADS1292_WriteRegister(ADS1292_REG_RLD_SENS, 0x2f);

ADS1292_WriteRegister(ADS1292_REG_LOFF_SENS, 0x0F); // Turn on both P- and N-side of all channels for lead-off sensing

ADS1292_WriteRegister(ADS1292_REG_LOFF_STAT, 0x00);

ADS1292_WriteRegister(ADS1292_REG_RESP1, 0x02);

ADS1292_WriteRegister(ADS1292_REG_RESP2, 0x83);

ADS1292_WriteRegister(ADS1292_REG_GPIO, 0x00);

ADS1292_SendCommand(ADS1292_CMD_OFFSETCAL);

nrf_delay_ms(200);

Key Register configuration

Reg	value
01	0x01
02	0xe0
03	0xd0
04	0x60
05	0x60
06	0x2f
07	0x0f
08	0x00
09	0x02
0a	0x83

When INT comes, can hardly detect lead-off from the ADC data

Channel data(24 status bits + 24 bits × 2 channels) = 72 bits

read 3 byte status register (1100 + LOFF_STAT[4:0] + GPIO[1:0] + 13 '0's)

LOFF_STAT[4:0] is RLD_STAT,IN2N_OFF,IN2P_OFF,IN1N_OFF,IN1P_OFF

over 7 years ago

0 Tom Hendrick over 7 years ago

TI__Guru**** 181621 points

Hi Bruce,

Sure thing! We can take a look over the ADS1292R register details and schematic snips for you. In the meantime though, can you let us know what "hardly detect lead-off" means here? Is it slow to respond or does that mean that don't get the comparator function to work at all?

0 Bruce li over 7 years ago in reply to Tom Hendrick

Genius 5580 points

Dear Tom,
Thanks for your help, in most cases, my customer don't get the lead-off function to work at all. Occasionally, it works.

0 Ryan Andrews over 7 years ago in reply to Bruce li

TI__Guru 66816 points

Hi Bruce,

Your electrode inputs appear to be AC-coupled. DC lead-off detection will not work.

If the RLD output is being used to bias the subject, then you can replace the series capacitors with 0-ohm resistors on the ECG channels. You will need to keep Channel 1 AC-coupled if it is used for respiration measurements.

Best Regards,

0 Bruce li over 7 years ago in reply to Ryan Andrews

Genius 5580 points

Dear Ryan,
Thanks a lot for your support. Yes, at first time, the series capacitors caused DC Lead-off mode detaction failed. Follow your advice, customer tried to change to the AC Lead-off mode , however , it failed again, below are the result and the configration , could you kindly help to check again?

LOFF_STAT[4:0] is RLD_STAT,IN2N_OFF,IN2P_OFF,IN1N_OFF,IN1P_OFF
Reg value
01 0x01
02 0xe0
03 0x11
04 0x60
05 0x60
06 0x3f
07 0x3f
08 0x00
09 0x02
0a 0x83

0 Ryan Andrews over 7 years ago in reply to Bruce li

TI__Guru 66816 points

Bruce,

The comparators are used for DC lead-off detection only. AC lead-off is determined by post-processing the raw data and analyzing the magnitude of the signal at the lead-off current frequency.

Please review the AC Lead-Off Section on page 53 of the data sheet.

In addition, you may find the following application note helpful as well to learn the differences between the two methods:

http://www.ti.com/lit/an/sbaa196a/sbaa196a.pdf

There is also some useful training material about ECG applications on our training portal. The following video is part of a 6-part series detailing ECG applications and typical features.

Best Regards,

0 Bruce li over 7 years ago in reply to Ryan Andrews

Genius 5580 points

Dear Ryan,
Thanks for your support. It's very helpful.

0 Aaronzhong over 7 years ago in reply to Ryan Andrews

Prodigy 10 points

Dear TI Experts,

We change to the AC Lead-off mode,electrode 1 connected to REF_EXT,electrode 2 connected to FP2_EXT.Measure the channel 2,it was found that the success of the test was related to the contact time between electrode 1 and skin,Describe the following:

1.At first,electrode 1 and electrode 2 maintain good contact with the skin.At this point,the LOFF_STAT[4:0] value is normal whether the electrode 1 or electrode 2 falls off,or both electrodes fall off simultaneously.

Electrode 1 falls:LOFF_STAT[4:0]=01000;

Electrode 2 falls:LOFF_STAT[4:0]=00100;

Electrode 1 and electrode 2 fall off simultaneously:LOFF_STAT[4:0]=10000;

2. But, if the contact time between electrode 1 and skin is longer than 15 seconds, AC Lead-off results will fail, and the LOFF_STAT [4:0] value will remain "10000". Unless the electrode 1 is kept away from the human skin for about 10 seconds at this time, and then the electrode 1 is re-maintained in good contact with the human body, it will return to the sucess state of detection. This abnormality is very regular, as long as the electrode 1 keeps in good contact with the body for more than 15 seconds, this will happen. But electrode 2 will not have this abnormal condition.

Based on the above phenomenon,we want to know:

1.If the comparator can only be used in DC Lead-off mode, why is LOFF_STAT [4:0] valid when we set it to AC Lead-off mode?

2.In the AC Lead-off mode, although the value of LOFF_STAT[4:0] is detected effectively. But why does electrode 1 keep in good contact with the skin once more than 15 seconds will appear abnormal detection of this rule.

Reg value
01 0x01
02 0xe0
03 0x11
04 0x60
05 0x60
06 0x3f
07 0x3f
08 0x00
09 0x02
0a 0x83

Best Regards,

Data converters

Data converters forum

ADS1292R: lead-off dectection insensitive