Questions about ADS1292 (2 Electrodes use, # of Channels, RLD)

Hi Denys,

Do you have a Win7 machine that you can install the software on?  There have been some mixed success stories here on the Precision Data Converter Forum with folks being able to get the software to work on Win10.

Here are the footprints you asked for...

ADS1292_PCBFP.zip

Hi Tom,

thanks for the footprints!

After playing with the ADS1292 Device Kit for the past few days, I have the following questions:

1) Does ELA channel work as a real Left Leg Electrode? In the schematics, it’s just connected directly to the Left Arm Electrode channel in series with a 51K resistor. Could you please explain to me the idea behind this connection?

2) I have tested ADS1292 with and without RLD with Ag/AgCl Electrodes and it gives really good results. In fact, there is not much difference in performance if I don’t use RLD somehow. The problem is that I don’t use Ag/AgCl Electrodes in my design. I have to use EPIC sensors from Plessey semiconductors, yes EPIC is the name… here is the datasheet of what I am trying to connect to the ADS1292 Device Kit – http://www.plesseysemiconductors.com/doc/?id=291501.

These are capacitive sensors and should deliver really good performance even if measured through a T-Shirt. In fact, when I am using Plesseys’ evaluation board, the results are really good even through a T-Shirt. However, if I connect these sensors to the ADS1292 Device Kit, the results are terrible. In fact, I can only see noise. Here is a screen shot:

I have also tested these Plessey sensors with similar chip from Analog Devices ADAS1000 and it worked pretty well. So, I am assuming that it also should work with the TI chip. Also, when I set Channel Input to: “Input Short” the results do become better:

Of course not enough to be considered an ECG. 

Also, while using the sensors from Plessey the Lead of Status turns on and off all the time on all channels, i.e. it shows that there is no constant contact although the electrodes are attached to my body very tightly. 

My questions are:

-          Can ADS1292 work with this kind of capacitive sensors, i.e. such as Plessey EPIC Electrodes (www.plesseysemiconductors.com/.../

-          Why does it work with Short Input Settings better? (even if it’s still bad)

-          Could you suggest me some directions on what should I set up in the registers to get max performance for my type of electrodes/sensors?

Kind Regards,

Den

Hi Den,

Unfortunately we've not used that particular sensor here in our lab. I suspect the issue might be to biasing though, when you used the other ECG chip, was that on an evaluation board like our ADS1292 or was it your own wiring/hardware? One thing about the input short test is that the inputs are shorted internally to the ADS1292 - it does not have the means to actually 'short' your leads together, so nothing on the outside of the chip comes into play. Send along your register settings and a schematic for the EPIC electrodes if you happen to have one.

Hi Tom,

Unfortunately, I don’t have schematics and the manufacturer will not disclose them to me.

I am thinking that the problem is with biasing as well. The sensors I am using are a type of “active electrodes”, i.e. they have an op. amp. right next to the sensors’ output. Therefore, they are biased with +/-2.5 Vdd/Vss. The output of the sensor can cover a range from +2.1V to -2.1V.

As I understand, the ADS1292 dev. kit has differential input configuration, which is described in sbas502b.pdf page 20, i.e. each INP and INN inputs can swing from CM + Vref/2 to CM - Vref/2, where CM is equal to (INP + INN)/2.  To allow for such a large swing from EPIC sensors I’ve set Ref. Voltage to 4V and PGA to 1. The result was the same as when I shorted the input, i.e. I could see something resembling an ECG signal, yet it was still too noisy. (Note, with 2.4V reference voltage I could see only noise).

Regarding the ECG chip from Analog Devices: yes, I also used a Dev. Kit -www.analog.com/.../UG-426.pdf. I also don’t understand why the chip from Analog Devices shows better results while it’s input range is smaller, i.e 1.3 V.

Here are my register settings:

I know I’m asking a lot of questions, but if you could answer them for me, it will greatly help me in the understanding of ADS1292, therefore:

Questions to the register settings:

• Since I am using capacitive sensors, I am assuming that AC Lead-off detection is more suitable than DC Lead-off detection?
• Does Lead-off detection influence an ECG signal quality?
• How do you decide on the PGA chop frequency? I just went with the lowest frequency, i.e. fMOD / 16.
• Could you please give me advice on the right register settings for respiration measurement on channel 1? Somehow even with normal electrodes, i.e. Ag/AgCl, my respiration measurement is incorrect. Jumping from 30 to 70 and rarely to 19-20, which is close to the truth.
• Could you advise me on the right settings for the RLD in RLDSENS register? Should I have it turned on for both channels or only for channel 2, where I am measuring an ECG signal? Should both positive and negative inputs be on or only one of them?
• Since each Left and Right Electrode go to channel 1 and channel 2(page 51, slau384a.pdf), I am assuming that LOFF1N + LOFF1P set to 1, should be enough to detect if the electrodes are properly connected? 
• Why does respiration detection make the signal a lot noisier? (tested with Ag/AgCL Electrodes, with EPIC electrodes the signal is too bad)
• Which Respiration control frequency should I choose and why?

  Questions to the Dev. Kit Board Schematics:

• The way channel 1 is terminated in ADS1292R Dev Kit, page 51 of slau384a.pdf, I am assuming that I can only measure Respiration on it?

 Kind Regards,

Den

Hi Stella,

I did get some of the active electrodes, but I was not able to implement RLD with them.  Channel 1 can be used for ECG or respiration, so that should not be a problem.

Hi Tom,

thank you so much for the quick reply. At this point I would like to ask you if you where anyway able to obtain a clean signal, or in your opinion it's not really possible to obtain a good quality signal with this kind of electrodes interfacing the ADS1292R. 

Thank you for your help.

Regards

Stella

Hi Stella,

Yes, you can get decent signals from the electrodes.  I found the package to be a little difficult to prototype with, but otherwise they are easy to use.

Hi Stella,

I don't remember off the top of my head how I had the registers configured for the EPIC sensors.  I'm out of the office this week but can take a look at the configuration again once I am back in Dallas.

Hi Tom,

thanks a lot, it would be really helpful if you could take a look at my configuration (posted above), or share the one you used to obtain good quality raw data.

Best regards,

Stella